-------
CORPORATION
^
0
C
•f—
-fj
c
o
O
^
^f
s^
«^
r-
^C
Q
LU
i
3
_j
HH
§
LU
CC
^
C^
ry
U.
Q
LU
^™
_J
=5
O
^
O
co
0
1— 1
r-
OL
^r
2
•s^
o
CO
CO
Q
Z
^>"
o
Q_
^s^
0.
rO
CO
•
CN
1
ro
LU
i
rf)
^c
H-
O
0.
0
CO
CQ
1ST
O
Q.
•^
rO
m
L.
<§
C
0
*,_
10
0)
•r-
E
LU
0
O
t
3
0
CO
m
ro >
1 O
0 rH
CN *fr rH r-l rH CN
^H^Hr-IO OCCCCX
• • • «x «zz
OOOO OQQCN
00 •
• r— 1
m
0
c
o
000r-00 00
CCCUCC CC
ooo>,oo oo
zzzozzu zz
in
U) 0
^^ f*-
c u
0 C
C rO
0 I-
0 CL-0
in E
3 O « O
>*- O U) O) D)i—
0^0) c C ra U
t- in > -r- •?- -p ,—
0 ro "O Ji O ro
r— i- 0 rO O -P -P
rO "O ^~ in 0 ro 1 O
a. O s_ x: -Q in -p
i- £i •> 0 o i in
o>u in r— tn -P«^
•f" c -P ^0 O S— fO +•* 3
Ss^a"0^ 0roi>
3 0 C
-Q 1 1 1 P T- r— r-
W U) r- 0 0
c to 0 O in to
0 a> t- in 0 0
Q.
0 ro
^
p- "O
3 0
Q. P
-C
0)
o 0
a. 3
.
c
o
•r-
ra
1—
a.
o
CL
O
•P
3
«J
VO
O^
r— 1
C
o
•o
0
01
ra
0
O)
!_
0
>
(O
T3
0
-p
.c
0)
•^
0
u
•
£_
o
4->
u
0
pH.
1—
o
o
f.—
to
o
•r-
c
JC
8
E
• •
o
•a
•
"O
0
p
{_
o
a.
0
±)
o
c
(O
•p
ro
Q
• •
CC
Q
0
3-7
-------
RADIAN
Another "indicator" of POM emissions, often used in reporting coke oven
emissions, is benzene soluble organics (BSD). Although most POM compounds
(including BaP) are soluble in benzene, BSD is not necessarily equivalent to
POM because it includes compounds other than POM (6). It has been established
that BaP is generally about 1 percent of BSD in coke oven emissions (7). How-
ever, since this correlation applies only to coke ovens and since BSO includes
compounds that are not POM, the correlation was not considered useful for
developing total POM estimates.
More recently, techniques have evolved for sampling and analysis of poly-
nuclear aromatic hydrocarbons (PAH). These compounds are a subset of total
POM as indicated by Table 3-1. Measurements of PAH do not include such POM
groups as aza arenes,.imino arenes, carbonyl and dicarbonyl arenes, hydroxy
carbonyl arenes, oxa-and thia-arenes, and polychlorinated polycyclic com-
pounds. However, PAH does include most of the major POM compounds and, as
such, PAH emission factors were considered to be appropriate for use in
developing total POM estimates where no other data were available.
Some total POM emission factors have been reported in the literature,
although most of these data were collected in two or three studies in the late
60's and early 70's. These initial studies have been cited repeatedly in more
recent reports, but not much new total POM emission data have been collected.
The general emission test method used in collecting the reported data,
the compounds included in the measurement(s), and the key characteristics of
the source(s) tested have been reported here when such information was readily
available in the literature sources used in this study.
3.1.2 National Emission Estimates
The most uncertain aspect of developing national emission estimates for a
source category is the assumption that an emission factor based on limited
te.st data for a narrow range of sources can be considered representative of
3-8
-------
RADIAN
the entire source category. In reality* POM emissions from specific sources
will vary with:
o raw materials used,
o process design and operating conditions, and
o emission control technology applied.
Most of the emission factor data used was reported to have been obtained
from "representative sources" with respect to raw materials and process design
and operating conditions. In some cases, emission factors for a weighted
average source population had been developed in the literature. Variations in
raw material and process conditions tend to be averaged out over the source
category population and thus should not interfere greatly with the development
of useful national emission estimates.
The impact of emission control technology application is more difficult
to assess. Emission factors reported in the literature vary with respect to
whether they were measured before or after the emission control device.
Additionally, there is a significant lack of data regarding the effect of
technologies designed to control other pollutants (particulate, NOX, S02» CO,
or hydrocarbons) on POM emissions. Finally, "controlled" emission factors in
the literature may not be representative of the emission control technologies
currently used in the source category.
During the course of developing national emission estimates, applicable
Federal air pollution regulations (NSPS, NESHAPs) were briefly reviewed to
determine the emission controls required for sources subject to such regula-
tions. Also, readily available published information on the current use of
emission controls in the source category was assembled. Engineering judg-
ments were made in some cases to use emission factors representative of the
application of a particular technology to the entire source category. Despite
these efforts, however, the national impact of current non-POM emission con-
trol technology requirements on POM emissions is difficult to assess (even
3-9
-------
RADIAN
CORPORATION
qualitatively) and is likely to be a major factor in the usefulness of the
national POM estimates developed in this study.
National production or fuel consumption estimates inherently contain
some inaccuracies associated with the collection and documentation of such
data. For the following source categories the data used to develop national
estimates were considered particularly susceptible to inaccuracies:
o tonnages of material burned in forest fires and other open
burning,
o tonnages in burning coal refuse piles, outcrops, and
abandoned mines,
o residential wood consumption, and
o the amounts of refuse combusted in municipal, industrial,
and commercial incinerators.
3.2 QUALITY OF REPORTED EMISSIONS DATA
Much of the POM emitted is associated with small particles entrained in
the exhaust gases from the source (i.e. plant or process equipment). There-
fore, high small particle collection efficiency is an important consideration
in the accuracy of the sampling apparatus used in measuring POM emissions.
EPA Method 5, which has been adopted as a standard method for measuring par-
ticulate matter emissions, is the method most commonly employed to measure POM
from stationary emission sources.
Some POM compounds from certain sources are emitted as vapor. The vapor-
phase POM emissions are not totally captured by the typical Method 5 particu-
late sampling apparatus used to measure POM. Depending on the temperature of
the source stream and the sampling apparatus temperature, undetected emissions
of POM may be significant (8).
Another problem with typical POM measurements is the potential for loss
of POM already trapped on the sampling train filter. Such losses are reported
3-10
-------
CORPORATION
to increase with increasing gas velocity and increasing gas temperature. POM
losses can also occur through chemical rearrangement of the collected sub-
stances on the filter surface. One source has reported that POM emission
measurements based on conventional Method 5 particulate sampling techniques
are low by a factor of 2 to 200 (8). Modified Method 5 techniques using
Impingers and solvent-filled bubble trains have been shown to substantially
reduce POM losses. Advanced techniques using sorbent resins to capture vapor
phase POM can result in much more accurate results* but these techniques do
not appear to generally have been used in collecting the available data
reported in much of the published literature.
A variety of analytical methods have been and are being used to analyze
the collected samples for POM. Apparently* agreement between POM concentra-
tions obtained with different analytical techniques can be "expected to be no
more than an order of magnitude (9)."
In summary* the uncertainty and variability associated with POM sampling
and analysis techniques cast substantial doubt upon the accuracy of POM emis-
sion factors reported in the literature. Virtually all of the literature
sources reviewed have included such a caveat in cautioning readers about
usefulness of the data reported.
3.3 REFERENCES
1. Merrill, Ray (IERL-RTP), Process Measurements Branch. Telephone conver-
sation with Mary Kelly (Radian Corp.). March 16. 1983.
2. National Academy of Sciences. Particulate Polvcvclic Organic Matter.
Washington, DC, pp. 4-12, 1972.
3. White, J. B., and R. R. Vanderslice (Research Triangle Institute, Inc.).
POM Source and Ambient Concentration Data; Review and Analysis. EPA-
600/7-80-044, U.S. Environmental Protection Agency, Research Triangle
Park, NC, p. 52, March 1980.
3-11
-------
4. Haro, J. T. (Energy and Environmental Analysis, Inc.) Preliminary
Assessment of the Sources, Control and Population Exposure to Airborne
Polvcvclic Organic Matter (POM) as Indicated by BaP. Final Report,
Prepared for U.S. Environmental Protection Agency, Research Triangle Par,
NC, pp. 22-35, November 10, 1978.
5. U.S. Environmental Protection Agency. Scientific and Technical Assess-
ment Report on Participate Polycyclic Organic Matter. EPA-600/75-001,
Washington, DC, March 1975.
6. Emission Standards and Engineering Division. Coke Oven Emissions from
Bv-Product Coke Production - Background Information for Proposed
Standards. Draft Report, U.S. Environmental Protection Agency, Research
Triangle Park, NC, pp. 3-17, March 1981.
7. Reference 6, p. 3-19.
8. Reference 3, Chapters 2 and 4.
9. Reference 3, p. 27.
3-12
-------
CORPORATION
4.0 NATURAL SOURCES OF POM
Natural sources of POM emissions are basically natural combustion sources
such as forest fires and volcanoes. Data on total POM emissions from natural
sources are scarce. The only readily available emission factors located were
from laboratory tests of burning pine needles. These emission factors and
corresponding national estimates are discussed below.
Table 4-1 shows POM emission factors reported for the laboratory forest
fires (1). As indicated, the emission factors vary by up to three orders of
magnitude depending on the type of fire. Other variables that significantly
impact emissions from forest fires are:
o the type of vegetation burned,
o burning conditions, and
o weather conditions.
In 1980, the U.S. Forest Service estimates that 1.8 x 1010 m2 (4.43
million acres) were burned in wildfires (2). This figure includes all types
of wildfires, from grasslands to forests. At an average of 2.3 kg/m^ (10.4
tons/acre burned), this amounts to approximately 41.2 million metric tons/yr
(45.3 million tons/yr) of vegetation burned in 1980 (2).
The average intermediate POM emission factor for all types of fires
included in the lab study results is 20 mg/kg burned. Multiplying this
emission factor by the total estimate tonnage burned in 1980 yields a rough
estimate of national POM emissions from wildfires of:
4-1
-------
RADIAN
CORPORATION
TABLE 4-1. EMISSION FACTORS FOR LABORATORY FOREST FIRES (l)a
Type of Forest Fire
Heading (flaming)
Heading (smoldering)
Heading (overall)
Backing (overall)
Minimum
3.46
21.8
7.63
10.2
POM Emission Factor^ (ma/ka
Intermediate
5.3
26
14
36
burned)
Maxi mum
8.39
31.5
22.8
172
aTests involved burning pine needles in a controlled environment burning room.
A modified "hi-vol" sampler was used to collect particulate samples. Samples
were extracted with methylene chloride, separated by liquid chromotography
and analyzed by GC/MS.
^Compounds measured: anthracene, phenanthrene, methyl anthracene,
fluoranthene, pyrene, methyl pyrene, benzo(c)phenanthrene, chrysene,
benzo(a)anthracene, methyl crysene, benzofluoranthenes, benzo(a)pyrene,
benzo(e)pyrene, perylene, methylbenzopyrenes, Indeno (l,2,3-c,d)pyrene, and
benzo(g,h,i)perylene.
4-2
-------
RADIAN
[41.2 million metric tons A [ ;
\ yr / V
20 am \ I metric ton
yr J \. metric ton/ \10 gm /
= 824 metric tons/yr (906 tons/yr)
It should be noted that this estimate is very rough for two reasons:
o Very limited emission factor data were available, and
those data were from laboratory fires burning only one
type of vegetation.
o Estimates of acreage and tonnage burned in wildfires are
uncertain and widely varied.
4.1 REFERENCES
1. Energy and Environmental Analysis, Inc. Preliminary Assessment of the
Sources, Control and Population Exposure to Airborne Polycvclic Organic
Matter (POM) as Indicated bv Benzo (A) Pvrene (BaP). Final Report,
Prepared for U.S. Environmental Protection Agency, Research Triangle
Park, North Carolina, p. 29, November 10, 1978.
2. Yamate, George (IIT Research Institute). Emissions Inventory from
Forest Wildfires, Forest Managed Burns, and Agricultural Burns,
EPA-450/3-74-062, U.S. Environmental Protection Agency, Research Triangle
Park, North Carolina, p. 7, November 1974.
4-3
-------
CORPORATION
5.0 MAN-MADE SOURCES AND EMISSION ESTIMATES
5.1 INTRODUCTION
This section presents the development of national total POM emission
estimates for the source categories shown in Table 5-1. The following sec-
tions discuss each of the following items for each source category:
o a brief source category or process description with
identification of POM emission points and the key factors
influencing emissions,
o emission control methods (for compounds other than POM)
currently used in the source category, their effect on POM
emissions and a brief discussion of applicable NSPS,
NESHAPs, and SIP air pollution regulations,
o geographical locations of sources,
o available emission factor data,
o national emission estimates for a baseline year, and
o readily identifiable trends in technology, source category
growth, or emission regulations that are likely to
influence POM emissions.
The baseline year for emission estimates was chosen as 1980. In addition
to providing relatively recent information, this year required a minimum of
extrapolation of production and fuel consumption data used to calculate
national emissions.
A summary of national emission estimates for the various source cate-
gories is contained in Section 2.0.
5-1
-------
RADIAN
coniKMumOM
TABLE 5-1. POM SOURCE CATEGORIES EXAMINED
Section No.
Category
5.2
5.3
5.4
5.5
5.6
o
o
0
5.7
5.8
5.9
5.10
5.11
5.12
0
0
0
o
o
0
Burning Coal Refuse and Other Open Burning
Combustion of Solid* Liquid, and Gaseous Fuels for Heat
and Power Generation
- Utility coal, oil, and gas combustion
- Industrial coal, oil, and gas combustion
- Industrial wood combustion
- Commercial/Institutional coal, oil, and gas combustion
- Residential coal, oil, and gas combustion
- Residential wood combustion
Coke Production
Iron and Steel Processes
Asphalt Production
- Hot Mix for Paving
- Saturated Felt for Roofing
Catalytic Cracking in Petroleum Production
Combustion of Municipal, Industrial, and Commercial
Wastes
Carbon Black Production
Wood Charcoal Production
Vehicle Disposal
Mobile Sources
- Gasoline autos
- Diesel autos
- Diesel trucks
- Tire wear
5-2
-------
RADIAN
5.2 BURNING COAL REFUSE AND OTHER OPEN BURNING
The section discusses three sources of POM emissions:
o burning coal refuse piles* outcrops* and mines
o agricultural open burning* and
o prescribed open burning (forest fires).
Unprescribed burning of leaves, grass, and other materials was not con-
sidered because of (1) the difficulty of obtaining meaningful data on emission
factors and tonnage burned and (2) the widespread prohibition of unprescribed
open burning (1).
5.2.1 Source Category Description
5.2.1.1 Process Description
Coal Refuse Piles, Outcrops, and Mines — Waste material separated from
coal is often piled into banks near coal mines or coal preparation plants. The
waste material is referred to as coal refuse, gob, culm, or reject material.
Indiscriminant dumping and poor maintenance of refuse piles are two practices
that can result in spontaneous combustion of refuse piles (2). Spontaneous
combustion of coal can also result in fires in abandoned mines, outcrops, and
impoundments. Emissions of POM from these sources are influenced by oxygen
concentration, type of coal and refuse, relative humidity of the ambient air,
and moisture content and temperature of the burning material (3).
Agricultural Open Burning — Open burning is performed in some rural
areas as a means of controlling agricultural wastes. The burning is carried
out in open drums or baskets, large-scale open dumps, or pits (4). The rela-
tively low burning temperatures and inefficient combustion typical of agri-
cultural open burning make it a potential source of POM. Emissions will vary
widely with the type of waste burned and burning conditions.
5-3
-------
RADIAN
coftpaiunoM
Prescribed Open Burning — Prescribed open burning of forests is prac-
ticed to reduce the chances of wildfires. Prescribed burning is usually well-
controlled and occurs at a lower intensity than wildfires. However, combus-
tion tends to be incomplete due to the high moisture content and varying
composition of the materials burned (5). Thus, prescribed burning is a source
of POM emissions. Emissions of POM will vary widely with the material being
burned, burning conditions, and weather conditions.
5.2.1.2 Emission Controls/Regulations
Coal Refuse Piles, Outcrops, and Mines — Various techniques exist to
control particulate and gaseous emissions from burning coal refuse piles,
outcrops, and abandoned mines. These techniques are based on cutting off the
oxygen source to extinguish the fire and on preventing the fire from spread-
ing. The control techniques for coal refuse piles include quenching the pile
with water and blanketing it with an incombustible material. Control tech-
niques for outcrops and abandoned mines include the use of fire barriers, sur-
face sealing, and flushing void spaces with water or an incombustible material
(6).
No data were available on the use of these control techniques for exist-
ing fires. However, in 1975 the U.S. Department of the Interior promulgated
regulations that require coal companies to dispose of coal refuse in a manner
that prevents or minimizes the chances for spontaneous combustion (6).
Agricultural Open Burning — Most states have regulations that prohibit
open burning. However, agricultural burning is not specifically restricted in
any of the states (4). Some states do require the farmer to obtain a permit
and also give local air pollution control authorities discretion over when
burns can occur (4).
-------
Prescribed Open Burning — POM emissions from prescribed open burning
can be reduced somewhat by attempting to maintain well-controlled and effi-
cient burning. No data were available, however* to characterize POM emissions
as a function of burning conditions.
5.2.1.3 Source Locations
Coal Refuse Piles* Outcrops* and Mines — Locations of these sources
are, of course, linked to mining locations. Kentucky, West Virginia, and
Pennsylvania accounted for about 63 percent of burning coal refuse and im-
poundments in 1972. Montana, Wyoming, Colorado, and New Mexico accounted for
about 66 percent of burning abandoned mines and outcrops. Other states in
which these sources are located include Alabama, Ohio, and Virginia (2).
Agricultural Open Burning — Limited data indicate that agricultural
burning (in terms of total acres burned) is prevalent in California, Florida,
Georgia, Hawaii, Kansas, Louisiana, Mississippi, North Carolina, Oregon, and
Washington all of which have significant levels of agricultural activity (4).
Prescribed Burning — Data on the acreage of prescribed burns by state
were not located. However, based on state-specific particulate estimates from
prescribed burning, the states where prescribed burning is used to a rela-
tively larger extent (than in other states) are Florida, Georgia, Idaho,
Montana, North Carolina, South Carolina, Tennessee, Texas, Washington, and
West Virginia (7).
5.2.2 Emission Factors
Coal Refuse Piles, Outcrops, and Mines — Preliminary sampling of
emissions from a burning coal refuse pile was conducted by Monsanto Research
Corp. (2). Particulate matter collected from a representative coal refuse
pile using "hi-vol" sampling equipment was analyzed for POM. A total POM
emission rate of 0.019 mg/m^-hour of burning coal refuse was reported. Based
on an average density of 1.5 metric tons/m^ for refuse piles, this translates
a-b
-------
into 0.013 mg/hour-metric ton of refuse burned. The POM compounds measured
are shown in Table 5-2.
Agricultural Open Burning — No POM emission factor data were available
for agricultural open burning.
Prescribed Burning - No POM emission data were available specifically
for prescribed burning. Therefore* the emission factors presented in Section
4.0 for laboratory forest fires were used to develop the estimates.
5.2.3 National Emissions Estimates
Coal Refuse Piles* Outcrops* and Mines — Estimates of the amounts of
burning coal refuse in piles* outcrops* and mines are difficult to develop.
Reference 2 reports that in 1968 there were about 250 x 106 metric tons (275
million tons) of coal refuse in piles* but no estimates were given for
impoundments* outcrops, or abandoned mines. A representative coal refuse pile
was defined as having the following characteristics (2):
o volume: 1.7 x
o dry density: 1.5 metric tons/m3
o percent of pile burning: 21
The total number of active piles in 1972 was estimated at 206 (2). Based
on "typical pile" characteristics and the number of piles the estimated amount
of burning coal refuse in piles is:
(206 piles)(1.7 x lO^/pile) (1.5 metric tons/m3) ( .21)
= 110 million metric tons (121 million tons)
This is less than half the total tonnage estimate given in Reference 2 which
was based on a Bureau of Mines estimate of "refuse material contained in coal
piles." Because not all the material is likely to be burning, the 110 million
metric ton figure calculated above is more suitable for use in calculating
5-6
-------
TABLE 5-2. POM COMPOUNDS IDENTIFIED IN PARTICULATE EMISSIONS FROM BURNING
COAL REFUSE PILES (2)
Dibenzothlophene
Anth racene/phenanth rene
Methylanthracenes/phenanthrenes
9-Methy1anth racene
Fluoranthene
Pyrene
Benzo(c)phenanthrene
Ch rysene/benz (a)anth racene
Dimethylbenzoanthracenes
Benzo(k or b)fluoranthene
Benzo(a)pyrene/Benzo(e)pyrene/pyrene
3-Methylcholanthrene
Dibenz(a»h or a»c)anthracene
Indeno(l,2,3-c,d)pyrene
9H-Dibenzo(c»g)carbazole
Dibenzo(a>h or a,i)pyrene
5-7
-------
RADIAN
national emission estimates from coal refuse piles. Due to a lack of more
recent data* the number of piles was assumed to be the same in 1980 as for
1972.
Thus, national POM emissions from coal refuse piles were estimated to be:
(-
\metr
013 mg \ /HO million metric tons\ /8760 hr\ /metric ton1
letric ton/h/ V / \ yr / V 106 gm
12.2 metric tons/yr (13.4 tons/yr)
Estimates of national POM emissions from burning outcrops and abandoned
mines could not be developed since no emission factor or "tonnage burning"
data were available.
Prescribed Burning — Limited data indicate that in 1976 approximately
15 million metric tons (17 million tons) of forest vegetation was consumed in
prescribed burns (1). The acreage was assumed to be the same for 1980.
Multiplying this figure by the 20 gm/metric ton average emission factor for
wild fires presented in Section 4 results in a national POM estimate of 300
metric tons/yr (330 tons/yr).
5.2.4 Trends Influencing POM Emissions
Coal Refuse Piles, Outcrops, and Mines — Proper enforcement of
existing regulations requiring preventative disposal measures and extinguish-
ing existing fires could substantially reduce POM emissions from coal
refuse piles. However, indications are that extinguishing existing fires is
difficult and costly and, as a result, has proceeded somewhat slowly (8).
Also, in 1971 there were about 40 burning piles and about 160 inactive piles
for which information on owners was not available. Debate over responsibility
for extinguishing existing burning piles of unknown ownership could preclude
quick action on these sources.
5-3
-------
RADIAN
Agricultural Open Burning — No Information was located on trends in
agricultural open burning. It can be hypothesized that more diligent enforce-
ment of open burning regulations and public awareness of obvious pollution
problems will tend to limit agricultural open burning to current or lower
levels.
Prescribed Open Burning — Because of its use as a wild fire prevention
technique* the level of prescribed open burning is not likely to change in the
near future. In the long term, continued research on fire control techniques
may provide for some reduction in POM from prescribed burning.
5.3 COMBUSTION OF SOLID, LIQUID AND GASEOUS FUELS FOR HEAT AND POWER
GENERATION
This section examines total POM emissions from the following combustion
categories:
o utility coal, oil, and gas combustion,
o industrial coal, oil, and gas combustion,
o industrial wood combustion,
o commercial/institutional coal, oil, and gas combustion,
o residential coal, oil, and gas combustion, and
o residential wood combustion.
Combustion of municipal solid waste and industial waste is covered in Section
5.8. Only combustion in boilers and residential furnaces, stoves and fire-
places is included in this study. Process heat sources were not covered due
to a lack of readily available POM emission factor data. However, coal- and
oil-fired process heaters are a potential source of POM.
A general principal applicable to all these combustion sources is that
more efficient combustion reduces POM formation (9). Therefore, as discussed
in detail later in this section, POM emission factors (POM emitted per unit
5-9
-------
RADIAN
heat input of fuel) are significantly less for utility combustion sources than
for less efficient residential stoves.
The following factors act to increase POM formation in combustion sources
(10):
o high carbon to hydrogen ratio and high concentrations of
oxygen and aromatics in the fuel,
o low temperatures in the combustion and post-combustion
zone,
o short residence time of combustion gases in the combus-
tion chamber,
o inefficient fuel/air mixing and lower air/fuel ratios,
o high frequency of start-up and shut-down, and
o larger (solid fuel) feed size.
POM emission rates from combustion sources are highly variable and, as indi-
cated by the factors listed above, are tied directly to fuel type and the
design and operation of the combustor.
5.3.1 Source Category Description
5.3.1.1 Process Description
POM compounds are formed in the gaseous phase in the combustion zone. As
the combustion off-gases cool, some POM compounds condense onto particles
present in the gas stream. POM is more likely to condense onto the smaller
particles in the gas stream because of the larger surface area-to-volume
ratios of small particles (11). At normal flue gas Stack temperatures of
about -150eC (300°F), significant amounts of POM reportedly exists as vapor
(9,11).
5-10
-------
RADIAN
COKMMUmOM
Utility Combustion
Utility coal-fired power plants burn crushed or pulverized coal to
generate steam> which is in turn used to generate electric power. The common
utility boiler types are:
o pulverized dry bottom (vertically-* front wall-* or tan-
gential! y-fi red ),
o pulverized wet bottom (opposed-fired),
o cyclone (crushed coal), and
o spreader stokers (generally used at smaller, older
utilities),
Pulverized dry bottom boilers are the most commonly used, accounting for about
76 percent of bituminous coal consumption by utilities in 1978 (12). Reported
POM emission factors vary somewhat between these boiler types (13).
In 1980, utilities firing coal, oil, and gas together accounted for
approximately 69 percent of the installed nameplate capacity of electric
utilities (14). On a heat input basis, the 1980 distribution of utility
consumption of these fuels is shown in Table 5-3.
The figure for coal includes consumption of lignite and anthracite.
Industrial Combustion
Boilers are used in industry primarily to generate process steam and to
provide for space heating. Some industrial boilers are also used for elec-
tricity generation. Industrial boilers are widely used in the manufacturing,
processing, mining, and refining sectors.
Coal-fired industrial boilers are generally watertube designs. Firing
mechanisms include pulverized coal and stoker (spreader, underfeed and over-
feed stoker). Most industrial boilers are front-wall fired (16). Large
5-11
-------
TABLE 5-3. BREAKDOWN OF FOSSIL FUEL USE IN COMBUSTION SOURCE CATEGORIES
1981 DATA
Percent of Total Fossil Fuel Consumed in
Cateaorv Accounted for bvc:
Category
Utility Boilers
Industrial Boilers
Commerci al /Insti tuti onal
Goal3
66
18
3
Oil"
13
17
17
Natural Gas
21
65
80
Reference
15
17
18
Boilers
Residential Furnaces
and Stoves
21
78
21
aCoal figure includes lignite and anthracite.
bOil consumed in utility and industrial boilers is primarily No. 6 residual
oil; distillate oil is primarily burned in residential furnaces.
cHeat input basis.
5-12
-------
RADIAN
industrial boilers are commonly pulverized coal-fired» while underfeed and
overfeed stokers are usually smaller units. Spreader stokers are found across
the entire industrial boiler size range.
Oil and natural gas are also burned in industrial boilers. Both water
tube and firetube designs are common for combustion of oil and gas.
The estimated 1980 distribution of coal, oil, and gas consumption in
industrial boilers is shown in Table 5-3.
These data indicate that coal accounts for a much smaller percentage of
boiler fossil fuel consumption in the industrial boiler category than in
utility boilers.
Dry or wet wood can be burned in stoker-fired industrial boilers. Often
these boilers are equipped with multi cyclones (mechanical collectors) which
are used to capture large* partially burned particulate for reinjection to the
boiler.
Commercial/Institutional Combustion
Boilers and furnaces at commercial and institutional facilities are used
primarily to provide space heat. The commercial/institutional category is
defined to include such facilities as hospitals, schools, office buildings,
and apartment buildings. The boilers are generally smaller firetube and cast
iron designs. Coal-fired units are small stokers. The 1980 distribution of
coal, oil, and gas consumed in this sector is shown in Table 5-3.
These data indicate that coal consumption is low relative to oil and gas
use in the commercial/institutional category.
5-13
-------
Residential Combustion
Goal-* oil-, and gas-fired furnaces* coal- and wood-fired stoves and
fireplaces are all used to heat homes. Combustion of coal and wood in domes-
tic stoves is a slow* low temperature* inefficient process (19). As discussed
above* inefficient combustion generally leads to higher POM emissions on a
heat input basis.
Residential coal-fired furnaces are usually underfeed or hand-stoked
units; oil-fired home furnaces use pressure or vaporization to atomize the
fuel; and air is premixed with the gas before the burner in residential gas
furnaces (20). Some types of wood stoves are more efficient than others due
to differences in sealing of the chamber and control of the intake and exhaust
systems.
The estimated 1980 distribution of residential coal* oil* and gas con-
sumption is shown in Table 5-3.
Wood consumption in residential units is discussed in Section 5.3.3.
5.3.1.2 Emission Controls/Regulations
This subsection provides a brief overview of the particulate* S02» and
NOX emission controls typically applied to utility, industrial, commercial/
institutional, and residential combustion sources. Available qualitative
information on the indirect effect of these emission control technologies on
POM emissions is also presented. Little quantitative data were located, but
some qualitative assessments can be made. In addition, a simplified discus-
sion of air emission regulations applicable to these combustion sources is
provided. Table 5-4 summarizes the controls and regulations for the various
combustion categories being considered. A detailed analysis of the current
use of emission controls on combustion sources and applicable regulatory
requirements is beyond the scope of this study.
5-14
-------
COMPORimOl
en
0
1
CO
z
o
HH
1-
to
g
0.
01
2
i—
CD
^
a
£
2
8
§
CO
CO
LiJ
O
1
CO
^
Ul
UJ
$
•P
I
1
z
Ul
^- o
.a f~
CJ (O
r— 3
a. co
GL (D
o
CD
_l Ul
Ul
r— CD
0 J3
C
O
o
(A
-P
0
C
Ul
X
LU
Ul
£. o
03 -P
CJ fl
r— 3
a. co
Q. CD
< tr
Ul
O
t.
C-Q
O "O
o CD
U)
u
CD
CL
CD
O
1
CO
<*-
to
QL
CO
z
OS
OS
i-H
CL
CO
I- LU
O
Ul U
L. c
CD CD
-P •*-
r- CJ
•*- 4-
Cj CD
JD 0)
Ll_ .C
i-H
r-
in co
o_ a.
HH CO
CO Z
Ul
L.
CD
3
t_
CJ
Ul
•P t-
a> CD
•o •»-
U1
CL CJ
CO •*-
LU t-
U l*-
3p I-
ftj Q
•P f—
^ O
1 °
**-
CO
a.
to
z
r-
^
Ul
•P
in
at1
ca
u.
CD
CO
Q_
CO
Z
J^
OS
i-H
Ul
Q_
HH
CO
§
•o o
CD
23
£
O 0
o o
c c
3 ra u)
^ §
.. Q.4J
t-
to
a.
C/l
z
01
<-H
85
LU
t.
O
°
in
g
— i
CO
Q_
to
j-4
r-
in
a.
HH
CO
Ul
8;
UJ
0
-o
CD
L.
•P
C
O
CJ
C
Q
0
4_
CO
Q_
CO
Z
CD
r*.
^
o -a
CD
L. r-
3 *—
4- O
r— L.
3 4->
Ul C
O
§ C
—1 3
CD
CO
Q_
CO
Z
<— 1
OS O)
Ul
Q_
HH
to
•o
•o o
CD r- t-
1- CD U
•P CJ C
C C 3
O fl
O T-
C r— «
•^ CL. CN
«t d
(C
-P
in ui
(O Z
>— i
to
Q_
CO
z
O)
c
•a
c
Cu
Ul Ul
to co
Ul
I. » o c -P
(. r- 4J (0 Ul
O -Q (O •»- X
i~l- r- Ul
in in »
LU CJ t. t- Ul
CD CD i—
ri- 4-> » Q. ?
..,_,_ CN g CD
S3 O i- OI O **-
5^ o M- cnf u -^
^
'fl
o
o
CO
Q_
z
O)
c
•o
c
CD
Q-
£
CO
"o
.c
Ul
1
'o
L.
•P
c
o
CJ
c
Ul
Q_
HH
CO
•a
^
o
L.
-P
c
o
CJ
c
a
o
CO
CL
CO
z
O)
c
•o
c
CD
CL
Ul
0-
HH
CO
0
CD
CJ
c.
a
CL
0
in
a.
HH
CO
t.
o
-a
CD
O r—
•P "3
C
O CD
O U
C C
•• a.
CN E
SIS
CO
a.
CO
z
CO
c
•o
c
CD
CL
Ul
Q_
CO
Ul
L.
CD
3
t_
O
Ul
•o
CD V
•— CD
"o *
L. t.
•P O
C
O in
CJ CL
C CO
13 LU
01
U)
Q_
HH
CO
ui in
t- u
-0 (DO
CD .a -P
r- ^ (J
r- 3 CD
O S- <—
t- tj »—
•p in o
c o
~^ y
U) C
Q_ fl
CN co jz
O( LU CJ
c/3 CD
^n yi ^
CL) £LJ O
T3
in o
-------
CD
3
C
C
O
o
«o
CO
Qj
CO
1—
CO
8
ce
£
CO
g
v-t
h-
3
UJ
a:
1
V)
1
o
1
i
o
CO
(/I
1— 1
£
LU
U_
O
^
3
CO
1
Ul
UJ
_J
§
i-
Ul
+J
c
=
3
z
Ul
•P
_c
o
c
•V
X
UJ
tn
r- CD
0 £1
L.
•P
C
O
o
Ul
II
•0 +*
O r- •— •— i—
U I- O O O O
•r- +J L. t_ L. L.
(O C C C C
JT CD o i— o o o
o "o cj m y y o
a> T- c o c c c
Z u) Z3 U 3 3 3
•^•> •"—>• "^"i
Ul (/> Ul CD
Q_ Q_ OL C
CO CO CO CO Z
Ul
o
u
CD "O "O
r- CD 0)
O "O r— r—
y CD O C
i— t_ t_
*— .— -P -P
«J O C C
O I- O O
— -P U 0 Tg
!i o 5 5 i—
£ O r—
U c •* •* O
CD ID CN CN L.
S O 6j -P
CO COJ C
^O Sj, to| Q_| CE| Z)
•P
^
•P
Ul
c
n- *0 f- T3
> CD ^
Ul E 01 -r- -PC
0 -^ U >» DQ O
a o >a r—
O -P a. i— c *o
U. » U iQ fl O CD
— - 4J at u u •*• in
X Q. 0 i— fl •
4J 4J ^- 3 QO 0T-U)
c a. r- 01 E ui u
u^Jzc " ""* o U*)CD£
,— T3f— ® -P • '-C — > O
i— CD S -
CD -^ -^ in j= S Ul CO CN O i^ ^ 4J (D J 3 TJ 4-
"a) * HJ 3 xi a -^.to-pc
OOino- 01 « -r- -P T)
f— X 3 r— *O CU1 *0cDoi, * — i CD O C -^ 3 L.>-H 3 r- O O O O >» Ul
2) 0 * 01 f— O t. O > $ r— fa'ro
ra-pyo -^-1-1 o 4- ougf— ^
OOC-CO) O 4- E Oi-rOC. O) -P (-O4- O£m
a.f- ui -p >*i-4 -P c N- co o f- -P JT
Q.T3 Ul r- Ol -P •»- -^- (J CN UlLn+J-^
flCCO-^ •»- +J O r- 3 R) O-PC CN 3 * Ul
-^ot-x n- u> fl « i— o a. coco — +j L.
f— L. "ui c +3oioy*O'— +jy*in ^^CD-P
i— O 01 O 4- 33 O « CL.U1 i Ul 01 H3
iO4-i-yO • < « t_ t_ 4- O CDCDt- ZCOt.
T-E ^-* t. o O O CNyo.e •»- o. IB
c'*oa.+J (D4-CD x: •• CDO r-r— I-CD«
•P -P O "O t- fl Ol**- -P^-t- f- O 4- i- 4-
OCOLCO CDCidf-EEUl E O*Oco O U CD e O
4->C "*? Ol "» "O •»- i— »— t- (— fl-O >. o •»- oo UICD
•^•py o o 3 o i— i- CD i- o o in iou*o N
tn 01 «j H- M4-L.>r-(nuii- ui r~- L. a. OCD-P-
* E E Q) CJ^CDUlUl i— "^ ^^ «^- "OOr— UC31*—
"O t— +J tn ^^ <~ C •*" E E Ul E C 4— T- O4-O i- (D 1- •«• 3 3 i- 3 3+J-O C-»-inJ3I_
OOU1®(- ^ t. CD 3 -P -P O -P -P CO CD f-l_ CD
«-PO -PgrcD c c >, c cO4~ ns 3
•*- »— O4- W ft— t — i — i — r— r— rtj •f-'PCNO
U13trtJ-P COOOr--r--r--P -^ -^E t--PQCD
•*- E» m CN «CN CDW1i--ac
« r— t- t- ^ CD 3 +J *-l • • »— O «OO U1Z 3
,— f- o cD-Puicoca«r-ioc • ^-i^^to O cow
34-4-r->-P-»ui «-3 "N. c ZLUO
•P*r-CE CCC-r-^j'-v^g) ,-, •xsOlfl COCO
I- L. O 01 OCDOU.I— *^.O1 O)-^ "^s. O1C UCCD1-O
fl ^ -r- -P 4-* •^- O i— Ol C C i— Ol C Z »•— -^OEOOQ. O^1 UlO t->i
•a y r-34*-^ mCN^-fl en cNro4- t-r— (D»—
CL-^- -PCOCN CO Ul^-(DCD-«-rtJ
CNCO •o-o-ot_4-a.i^»«" CL •• •- c o~o-P4-
OUJOCDCUflOCO CNCO CN OJ3 flV
CO £ -P X3 Q. Z -PS O Z S O i-« U1CDOO
•. yc CD XQ-CO OL co UJ-OCD-^I--^C
>% 01 C CD CD COr-4S C\ U10O1 Oil.
*— SOQ.-P-cr«. r* f -p f- < xi >i
TJ XI U "O CD 4- Ol-Ci— T-)
5-16
-------
RADIAN
CORPORATION
Unfortunately, without detailed information on the application of emis-
sion control technologies and quantitative data describing indirect effect of
particulate, SC>2» or NOX controls on POM emissions, the accuracy of national
emission estimates developed in Section 5.3.3 is questionable. However, an
effort has been made to use published emission factors that were judged to be
fairly representative of the current population of combustion sources, includ-
ing the type of emission controls used. The estimated accuracy of the
national emission estimates 1s discussed further in Section 5.3.3,
Utility Combustion
Utility coal-fired boilers constructed in the last 10 to 12 years are
required to use control measures to limit S02» particulate matter, and NOX
emissions. Older coal-fired utilities are generally equipped only with parti-
culate control devices. POM emissions are likely to be affected to some
degree by particulate, S02» and NOX control systems.
Some POM compounds condense onto particulate matter at normal flue gas
temperatures encountered in utility boilers. As discussed above, the com-
pounds tend to condense on smaller (fine) particles. Therefore, par-
ticulate emission control devices that are efficient collectors of fine par-
ticles will provide a significant degree of control of the POM associated with
particulate matter. Fabric filters and ESPs are high efficiency particulate
control techniques applied to the current population of utility boilers.
(Fabric filters are used only on relatively new utility boilers.) The fine
particle collection efficiencies of ESPs currently in use will often depend on
the age of the boiler. Older boilers are generally subject to less stringent
air emission regulations, and ESPs applied to these units are likely to exhi-
bit lower fine particle collection efficiencies.
Wet scrubbers and multicylcones are also used on some existing utility
boilers for particulate control. Low-energy wet scrubbers (low pressure drop)
have lower fine particulate collection efficiencies than high-energy venturi-
type scrubbers. However, it should be noted that wet scrubbers may be more
5-17
-------
CORPORjmOM
effective in reducing POM emissions than "dry" control systems because those
POM compounds existing as vapors would be condensed and collected as the gas
is saturated in the scrubber.
Multicyclones, which may still be in use as the sole particul ate control
device on a few old utility boilers, are not efficient collectors of fine
particles, and are therefore not expected to significantly reduce POM emis-
sions.
The most common S02 control technology currently used on utility coal-
fired boilers is 1ime/limestone flue gas desulfurization (FGD). This techni-
que employs a wet scrubber, which is often proceeded by an ESP. The ESP
collects the particulate before the flue gas enters the FGD system. Wet
FGD/ESP systems, while providing for control of POM condensed on particulate
at the entrance to the ESP, are not likely to achieve significant control of
vapor phase POM (22). Condensation of vapor phase POM compounds will occur in
the wet scrubber, but significant collection of particles remaining in the gas
flow through the scrubber is not likely to occur. Systems equipped with mist
eliminators may exhibit a slight reduction in particulate before the stack.
There were 94 operating utility wet FGD systems as of January 1982 (about 13%
of installed coal-fired capacity) (23).
A more recently applied utility S02 control technique is spray drying.
In this process, the gas is cooled in the spray dryer but remains above the
saturation temperature. A fabric filter or an ESP is located down-
stream of the spray dryer. Thus, this system would provide for significant
control of both particulate and vapor phase POM because the vapor phase com-
pounds are condensed before they reach the high efficiency particulate control
device. One source (22) estimated that over 90 percent of the benzo(a)pyrene
emitted from a representative utility boiler would be controlled by a spray
dryer FGD system. As of early 1983, there were five utility spray drying
systems, all applied to new units in the initial phases of commercial opera-
tion (24).
5-18
-------
RADIAN
CORPORATION
Nitrogen oxide control techniques applied to newer coal-fired boilers
include low-excess air firing and staged combustion. The principle of these
control techniques is to limit the oxygen available for NOX formation in the
combustion zone. However, as discussed in Section 5.3.1.1, lower air to fuel
ratios may lead to Increased POM formation (other conditions remaining con-
stant). Data regarding the effects of "low-NOx" operating conditions on POM
formation is conflicting and very limited (25).
The POM emission factors that have been reported for utility boilers
have either been based on data from uncontrolled boilers or on measurements
made downstream of an emission control device. No emission data were located
that included simultaneous POM measurement both before and after the control
device. Thus, it is difficult to quantify the Impact of the technology on POM
emissions.
Air emission regulations applicable to existing coal-fired utility
boilers include SIP emission limits and the 1971 and 1979 NSPS (Subparts D
and Da, respectively, of 40 CFR 60). The 1971 NSPS applies to boilers con-
structed after August 1971. That standard specifies a 520 ng/J (1.2 Ib/mil-
lion Btu) S02 emission limit and a 43 ng/J (0.1 Ib/m1llion Btu) particulate
matter emission limit. The 1979 NSPS, which applies only to units constructed
after September 1979, is more stringent, requiring 70 to 90 percent S02
removal and a specifying a 13 ng/J (0.03 Ib/million Btu) particulate emission
limit (26). Under this NSPS, new coal-fired utility boilers will generally be
required to be equipped with fabric filters or high efficiency ESPs and FGD
systems. The 1979 NSPS also requires the use of NOX control techniques.
Existing oil-fired utility boilers are generally either uncontrolled or
are equipped with ESPs for particulate emission control. The same consider-
ations discussed above for ESPs applied to coal-fired utility boilers apply to
oil-fired units. S02 control for existing oil-fired boilers is most often
achieved by the use of lower sulfur oils.
5-19
-------
RADIAN
Air emission regulations applicable to oil-fired utilities include the
1971 and 1979 NSPS and SIP requirements.
Gas-fired utility boilers are not equipped with emission controls since
S02» particulate, and NOX emissions from these units are relatively low.
Industrial Combustion Sources
Emission control techniques for industrial coal-fired boilers are similar
to those described for utility coal-fired units. However, application of S02
and particulate control technologies to the current population of industrial
boilers differs substantially with the application of these controls to util-
ity boilers. There are very few industrial boilers equipped with FGD systems
for S02 control. Fewer than 15 1ime/limestone or double alkali systems are
operating on industrial coal-fired boilers. There are about 12 sodium-based
FGD systems operating on coal-fired boilers, primarily at paper mills and
textile plants (27). Five spray drying FGD systems have begun operation in
the last few years (28). However, most existing coal-fired industrial boilers
meet applicable SC>2 regulations by burning low or medium sulfur coal.
The use of particulate control on industrial boilers is more common.
Existing industrial coal-fired units are subject to SIP particulate matter
emission limits. For a 44 MW-^ boiler, these limits vary from 22 ng/J (0.05
Ib/million Btu) in California to 344 ng/J (0.8 Ib/million Btu) in Iowa (24).
Most SIPs fall in the 86 to 172 ng/J (0.2 to 0.4 Ib/million Btu) range, which
generally requires the use of a low-efficiency ESP or a multicyclone collec-
tor. SIP emission limits for larger industrial boilers (>73MW-t) are more
stringent, often specifying a 43 to 86 ng/J (0.1 to 0.2 Ib/million Btu) emis-
sion limit. Medium- and high-efficiency ESPs are required to comply with such
regulations. Finally, unless the SIP is more stringent, large industrial coal-
fired boilers (>73 MW-fc) constructed after August 1971 are subject to an
existing NSPS which specifies a 43 ng/J (0.1 Ib/million Btu) particulate
matter emission limit. Fabric filters or ESPs are commonly used to meet
this standard.
b-20
-------
CORPORATION
As of December 1979, 104 Industrial boilers were using or planning to use
fabric filters (30). No comparable data were located on the number of ESPs,
wet scrubbers* or multicyclones applied to industrial units, although the use
of ESPs on existing boilers is considered to be substantially more widespread
than the use of fabric filters (29). Wet scrubbers are not generally used for
particulate control on coal-fired boilers because they result in a liquid
waste stream that must be disposed of properly (29).
A relatively new particulate control technology, applicable to stoker-
fired industrial boilers, is the side stream separator. This device is a
multicyclone modified to treat part of the gas in a small fabric filter. The
fabric filter enhances the overall collection efficiency of the multicyclone
by removing a portion of the fine particulate (31). To date, side stream
separators have been retrofitted to only a few existing stoker-fired boilers
(31).
An NSPS currently being developed by EPA may require that fabric filters
or ESPs be applied to all new coal-fired industrial boilers above about 29 MW
(100 x 106 Btu/hr). At a minimum, the NSPS will require use of side stream
separators. The NSPS is also likely to require low excess air or staged
combustion for control of NOX emissions. At this time, the standard being
developed contains no provision for S02 control. The NSPS may be proposed as
early as the summer of 1983.
Particulate emissions from oil-fired industrial boilers are generally not
controlled under current regulations. However, oil sulfur content restric-
tions do apply to some existing units. New oil-fired boilers will be subject
to the NSPS being developed by EPA, which may require some degree of particu-
late emission control.
Mul ticyclones followed by venturi or impingment-type wet scrubbers are
the most common type of particulate matter control devices applied to wood-
fired boilers (32). The same considerations discussed above for wet scrubbers
used on coal-fired boilers apply to the use of the technology on wood-fired
5-21
-------
RADIAN
CORPOnOTtOM
boilers. However, the use of wet scrubbers is more common for wood-fired
units than for coal because wood-fired units are often located at paper mills
or other facilities with existing equipment that can be used to dispose of the
liquid waste stream from the scrubber. ESPs are also used to some extent on
wood-fired boilers. The use of fabric filters, however, is restricted to
those boilers firing a salt-laden type of wood because of potential fabric
filter fire hazards associated with combustion of other types of wood (32).
Wood-fired boilers are currently subject only to SIP emission limits.
Forty-three states have SIP limits for particulate emissions from wood-fired
boilers (40). These emission limits are about the same as those in SIPs
applicable to coal-fired units.
Commercial/Institutional Combustion
Commercial and institutional boilers are generally uncontrolled or equip-
ped only with multicyclones (mechanical collectors) for particulate matter
control. As previously discussed, multicycl ones are not efficient collectors
of the fine particles with which POM compounds are generally associated.
Commercial and institutional units may be subject to SIP particulate
emission limits in states that do not specifically define their regulations as
applicable only to industrial boilers. Most SIP particulate emission limits
are a function of boiler size and regulations are significantly less stringent
for small sources. New commercial/institutional boilers with greater than
14.7 MWt (50 million Btu/hr) heat input capacity will be subject to the NSPS
currently being developed for industrial boilers (33).
Residential Combustion
Residential combustion sources are not equipped with add-on particulate
matter control devices. Combustion modifications designed to improve the
efficiency of coal- and wood-fired residential stoves have not been exten-
sively applied to commercially available stoves (25). Most of the combustion
5-22
-------
CORPORATION
modifications applicable to residential stoves are in the form of stove design
changes and include modified combustion air flow control* better thermal
control and heat storage* and the use of catalytic combustors.
The sulfur content of the coal burned in residential units is regulated
in some areas, but in general residential combustion units are not subject to
Federal or state air emission regulations.
5.3.1.3 Source Locations
Utility Combustion
Table 5-5 shows the existing utility capacity by fuel type for each of
the 50 states. Existing capacity for specific fuels is expressed in MW elec-
trical output and as a percentage of the total existing capacity for that
fuel.
These data indicate several well-established trends:
o Coal-fired units are concentrated in the states of Ohio,
Indiana, Pennsylvania, and Illinois. Other states with
substantial coal-fired capacity are Alabama, Georgia,
Kentucky, Michigan, Missouri, North Carolina, Tennessee,
Texas (lignite), and West Virginia.
o Residual oil-fired units are found primarily in Califor-
nia, Florida, and New York. It should be noted that in
contrast to the installed capacity data in Table 5-4, 1980
fuel consumption data for utilities show that in several
states, little or no residual oil was burned by utilities
(15). States which reported no oil consumption by utili-
ties included Alabama, Indiana, Kentucky, Montana, North
Dakota, Oklahoma, Vermont, West Virginia, and Wyoming.
The decrease in residual oil use is likely the result of
units with dual-fuel capabilities being fired with coal or
natural gas as a result of the sharply increasing residual
oil prices in recent years.
o Natural gas-fired units are highly concentrated in Texas
(over half the installed capacity) and Louisiana.
b-23
-------
CO
to
LLJ
CL
LL.
Q
LU
co
CD
Q-
O
t
CO
o
CO
>
3 -P
•P i-
(O O
Z to
a.
• to
CO O
ID in
(O
6? 0
^g
T5*
r— XI
i- >>
0 -P
•r-
co to
• a.
^ to
o
6?
s
1 —
«J -Q
o >,
0 -P
• TJ
t/} flj
• a.
0 "
gE
^T
0}
•p
to
-p
CO
** LO 0* CO CO
VO VO ^ CM vO CO
O O O CM O O O
O •-! CM CM O P- O
CO O\ I — vO i— 4 CO
^•f ^1" CO VO LO CM
1—4
O CM Tf •«*
r-H CO CM CO VO LO LO
CO CO CM CO vO CO CM
.-4
co oo r*^ c^ oo co ^o
*?f p**1* ^i* ^^ vo ^^ G3
CO «-H ^ CO
CM
CM CO
cj\ ^D r*^ r** r^1
ro o 1-1 o o •— i o
LO T LO CM 0 CM 0
co to co i— i •— i
LO r- .-i LO
o ** CM ^r
r-4
•p
10 3
i- U
U) C O i-
tO to to t- "O 4->
E to c in o to o
iO x O C v^ L Q)
jQ W) N to i— O C
to to i- .*: i— f— c
ON i — I p*1 O\ CO O^ CO C7^ i— I CM ^f CO
COO O CN| •-( i— 1 CO i-4 CO i— IOCMCMr--LOOVOLO
OOOOOOOO^OtOOO^-IOCOi-HOOi-lO
CO r-4 LO CO CO O^ CO i—l CM O^ CO CO *^ C7* LO CM CO C3\ CO O^ CO ^3
CO LO rH O^ *^ r^4 CO CO **f CJ^ VO CO LO CO CO VO
co CM 1-4 vo i— i r-i i— i co i— i r-» r-< LO i— i
CO r— 4 CM i—4 i— 4
r-l
vo i— i oo co co co i*~ oo"»otMr~
COLO^J-OOOCOO^tr-ICJvr-CJiO^COOvOCMOLOi— I**
OTj-i-ir-io^'Oi-4OOOocM'«a-cO'-4Oi-ioooo
•-4
vo LO r** co co co f-H vo P*** *^ »~H o^ o^ o\ o^ ^^ CM ^^ ^f r~4 CM o^
CJ\ CO LO LO ^2 CM CM ^f CT^ f^*- ^f CO OO VO ^f CO OS CO LO CO CO
i— 1 i— 1 CM I— 1 VO i— 4 i— 4 I— 4i— l^tl — *3"r-4 i— t
CM
1 — i—4 C7» LO CO P-
CMCOCM 'S' O i— 1 LO CO 'S' i—4 CMO\OOLOCO'— 4COi— 1
OrH^OOvOvOCMi— iLOOOi— IO**i— lO^-Oi— IOO
i--cr»oo^r^--^-oo-*coo^-oo>vococoi7iLncoo^
'icLor^ i— 4 vo •— i co co •— i co CMr—CMcocoLO'3'Lo
r— co »sr coi— ivoo^i— 4 o coo^ocftocM**
^t i— 1 1 — \O LT> f f <— 1 CO i— 1 LO CM CM COCM
i— 4 i— 1 i—l i— 4 i—4i—4
in a. x:
to ID to a. tn
0) U) >tCX>3C-Pi-i- tO CX
c-tota i-to ^10 cjrroouis-rcijj: E
to T3 1-1- oc tnui- foocnifltnacuiioto
^ i— D) "^" O C rO to 3 in 0) r— tO t~ Q) i— O to (O T3 T"
to J^ L. to ^~ »r- «^- to tn 4^ *^~ c >> in *~ c in in ^ t ^
i — oosfOi—"O5cc3'F-t_(nocintncxi>s
CDi — (DtO'Oi— cOiO° ^r "y TF ^r ^r ^r ^* ^^ ^^ ^^
^^ ^l/ ^. r^ ^^ r~I f~~I J^ ^C. «vj ^^ «£^ ^^ ^^ rf^ ^^ ^^ ^£, ^^ ,^. ^^
^
0)
13
C
•r™
-p
c
o
—
5-24
-------
*-N
0
3
C
•r—
^J
C
0
o
^-s
m
v^x
rt3
LU
Q-
t
i
LU
ID
U.
Q
1 1 1
LU)
^
CO
>-
CD
t
H~t
O
QL
^^
o
t
t— i
_l
1— 1
1
t^
o
LU
—1
5;
^—
CO
z
i— i
o
co
CTi
i— 1
•
in
I
in
LU
_i
3
H-
U)
CD
r—
(0
L.
3
.)-)
Z
r—
O
"(0
T3
U)
0
ce
f__
fQ
o
o
"ro
t- >.
3 -P
-P T-
0} (J
Z (O
Q.
CO O
•
ID U)
ro
6? CD
i
i- >>
0 -P
• O
CO i
O -P
• "o
• Q.
^3 — • i — •
O CN r-l O *5t Tt i— 1
r-l VO CN 00
CO
oo inr-- CTVO CT> •-< coo vor-i •— i
OOCNCNvOi— ICOinmOi-lvOCN CNOr-lTtr-IOr-IO
mocoooi-HOOinoi-ioocNoocNOOi-io
r-i
§vo r*^ CN vo r** CN r-~ i— i c^ vo m in ^^ vo o^ r^* in
m co oo LO ^f oo CTi CN vo ^f m CN *sf r-n CD i— i
r-~^-vocT>CNvoflor— ^tcNcoco CN I-HLOCN r~
OOCTi CN r-CN CO CO r-l
rH
rH i— 1 OO LTl i-l CT>
vocor-i'tcNooLncNi— i mi-ir--cj\oooco'a-'i'r-CTi
Or- ii—i's-r-iooi— ior-Oi— lOco'S-oor-ioincNi— i
<~>
in co r^ rH co r— i c^ r— i i— i ro vo r*" r— i CD m CD oo CD o^ co co
vo r** r** CN co co ^" vo cr* CN c^ CN i— i r— i co co co r** co CN
vo^to^cocNoomcN i-H'S'Ococo ^fcor^CNo
i— i co CN i— i co "f ^~ cy» ^f OCOCN co i— i ^f P- in
r-l CN r-l i— 1 r- 1 i-H
fO fO
c c >a
•r— rO fO "O *r— ,OOO C(OOO CT-
!_"xt_OQ E r-MOQW +J-»- O>T- In 0)
000 oc>< in c c c > c c
^"3 J* ^-~ .f^ ^^ <^ Q (fl QJ _£! ^^ Q) (/) O "r™ *i""™ O "*^
•f_) ^J Q fl|J ^Jj (^ ^^ ^J >|^ ^ (0 ^ C ^J ^^ -^J ^ C
SS3Bj-*-'i-r— ®CO33CXrOS-S_l/)(ninO
zzzzzoooa.o:cocoi— HZ3>>ssss
^
1— 1
CO
*
m
p-
^
00
oo
i— i
^
0
r~H
^
s*
CO
CN
r--
•t
rH
p*»
CN
"TO
0
1-
•
co
3
•
-P*
•r"
o
•r™
£_
-P
U
0
"0
0
+J
rO
U
0
C
0
O)
•p
ro r—
J= 0
-P 3
M-
U)
+J -P
C ro
(O .C
r- -P
a.
r
(- O
0 Ir-
o >%
a. -p
r- U
rO ra
i- Q.
1— ro
•P U
in
3 >,
•o -P
C -r-
•r- 1 —
0 -P
a> 3
£_
rO "O
r— 0
f—
0 r—
•0 ro
3 -P
r— U)
U C
•r-
*4—
>, 0
rO
^r b^3
^- OS
fO X)
-------
RADIAN
Table 5-6 shows existing capacity data for coal-, oil-, and gas-fired utili-
ties aggregated by EPA region.
Industrial Boilers
Industrial boilers are located throughout the United States. Boiler
locations tend to follow industry and population location trends. Table 5-7
shows regional industrial energy use for six major industrial sectors. Al-
though these data represent total industrial energy consumption (i.e.* elec-
tricity, raw materials* boilers* and process heat sources, they are indicative
of industrial boiler geographical concentrations. Most of the coal-fired
industrial boilers are in the Great Lakes, Great Plains, Appalachian, and
Southeast regions. Oil-fired boilers are common 1n the New England, South-
east, and Upper Atlantic regions, while the highest concentration of natural-
gas-fired units is found in the Gulf Coast and Pacific Southwest regions (35).
Wood-fired boilers tend to be located almost exclusively at pulp and
paper, lumber products and furniture industry facilities. These industries
are concentrated in the Southeast, Gulf Coast, Appalachian, and Pacific
Northwest regions (35).
Commercial/Institutional Boilers
These sources are also spread throughout the United States and their
concentrations are tied directly to population centers. Fuel use patterns for
commercial/institutional boilers are likely to parallel those described above
for industrial boilers, since fuel choice decisions in both categories are
made on the basis of fuel availability and prices (including transportation).
Residential Combustion
Locations of residential combustion sources will also be directly tied
to population trends, with one exception: wood-fired stoves and fireplaces.
Wood-fired residential combustion sources are concentrated in heavily forested
b-26
-------
TABLE 5-6. 1980 EXISTING UTILITY CAPACITY BY EPA REGION AND FUEL TYPE
EPA
U.S.
Region
1
2
3
4
5
6
7
8
9
10
Total
Exi
Coal -Fired
489
4,642
41,335
69,751
81,147
24,062
24,845
16,524
6,983
1,945
271,723
stlna Capacity
Oil -Fired
13,489
28,409
16,595
28,483
18,174
7,298
4,718
1,584
29,950
1,488
150,188
(MW)
Gas-Fired
9
179
205
5,760
1,751
58,544
5,421
358
2,042
1,045
75,314
Aggregated from data in Table 5-5.
5-27
-------
RADIAN
m
LU
5<
Z
o
3
LU
o:
•-,**
Ct-eoo
m U
3£-O
5-28
-------
areas of the U.S. This pattern again reflects fuel selection based on avail-
ability and price (36).
Table 5-8 shows the percent of total U.S. wood consumption in fireplaces
and stoves by state and by EPA region.
5.3.2 Emission Factors
Table 5-9 presents total POM emission factors for the various combustion
source categories being considered in this study. Appendix A contains more
detail on the bases of the emission factors shown in the table.
As noted earlier, there are several cautions that apply to the use of
these emission factors as representative of POM emissions from specific
boilers. The emission factors used here are intended to represent only the
best available estimates of POM emission factors for the particular combustion
source population as a whole and generally will not apply to specific units
with a high degree of accuracy. The most important cautions regarding the
emission factors presented in this section are summarized below.
Much of the data used in developing the reported emission factors was
collected in the late 60's and early 70's using modified Method 5 sampling
techniques. As discussed in Section 3, a reportedly significant amount of POM
is present in the vapor phase at temperatures typical of utility and indus-
trial boiler stacks. These POM compounds are unlikely to have been fully
captured with the Method 5 sampling train unless relatively advanced modifica-
tions were incorporated. Thus* the POM emission factor data may represent
primarily only the POM associated with particulate matter at the sampling
point. Therefore* the use of these emission factors, with all other assump-
tions being accurate, could result in an underestimate of nationwide POM
emissions from combustion sources. No useful quantitative data were found in
published literature to allow estimation of how much vapor-phase POM typically
escapes.
5-29
-------
TABLE 5-8. 1981 WOOD CONSUMPTION IN FIREPLACES AND STOVES BY STATE AND
EPA REGION (37)
EPA Region
Region 1
Region 2
Region 3
Region 4
Region 5
State Wood Consumption (% of U.S. Total)3
Connecticut
Maine
Massachusetts
New Hampshire
Rhode Island
Vermont
Regional Total
New Jersey
New York
Regional Total
Del aware
Maryland
Pennsylvania
Virginia
West Virginia
District of Columbia
Regional Total
Al abama
Fl ori da
Georgia
Kentucky
Mississippi
North Carolina
South Carol ina
Tennessee
Regional Total
Illinois
Indiana
Michigan
Minnesota
Ohio
Wisconsin
Regional Total
1.6
1.6
2.3
0.89
0.29
0.82
7.4
2.3
5.6
7.9
0.35
2.3
6.3
4.8
2.0
0.03
15.7
1.5
1.7
2.1
3.1
1.3
5.9
1.3
4,3
21.0
3.8
3.4
4.5
3.0
5.3
3.1
23.1
(Continued)
5-30
-------
TABLE 5-8. 1981 WOOD CONSUMPTION IN FIREPLACES AND STOVES BY STATE AND
EPA REGION (37) (Continued)
EPA Region
Region 6
Region 7
Region 8
Region 9
Region 10
State
Arkansas
Louisiana
New Mexico
Oklahoma
Texas
Regional Total
Iowa
Kansas
Missouri
Nebraska
Regional Total
Colorado
Montana
North Dakota
South Dakota
Utah
Wyoming
Regional Total
Arizona
California
Hawaii
Nevada
Regional Total
Al aska
Idaho
Oregon
Washington
Regional Total
Wood Consumption (% of U.S. Total)3
1.2
0.79
0.98
1.8
1.2
6.0
0.36
0.32
3.7
0.19
4.6
1.2
0.83
0.07
0.12
0.49
0.31
3.0
0.57
5.1
<0.01
0.26
6.0
0.39
0.83
2.0
2.3
5.4
of total U.S. wood consumption in fireplaces and stoves.
b-31
-------
CORPORATION
TABLE 5-9. TOTAL POM EMISSION FACTORS FOR COMBUSTION SOURCE CATEGORIES3
Source Category
Utmtv Combustlpn
Coal
on
Gas
Reported
Emission Factor11
19 ug/kg
2 x 10"6 lb/106 Btu
DNLf
Converted to Heat
Input Basis0
pg/J
0.71 (1.65)
0.86 (2.00)1
Emission Control
Level
Represented''
"Controlled" (ESP)
None
Emission
Factor
Basis9
1
3
Industrial Combustion
Coal
Oil
Gas
Wood
Commercia1/Institutional
Combustion
Coal
01lJ
GasJ
41 yg/kg
21 ug/i
11 ug/m3
2059 yg/kg
DNL9
21
11
1.53 (3.57)
0.50 (1.17)
0.30 (0.69)
105 (242)
0.50 (1.17)
0.30 (0.69)
Multl cyclone
None
None
h
None
None
Residential Combustion
Coal
Oil
Gas
Wood Stoves
Wood Fireplaces
67 mg/kg
120 yg/i
65 yg/m3
0.27 g/kgk
0.029 g/kg
2500 (5826)
2.86 (6.67)
1.74 (4.06)
13,472 (31,435)
1447 (3375)
None
None
None
None
None
2
2
2
2
3
aMore detail on each of these emission factors can be found 1n Appendix A. All emission factors are from
Reference 1, Table III-l ("Intermediate" factors were used) except for emissions for oil-fired utilities
(Ref 38), wood-fired Industrial boilers (Ref 39), and wood-fired residential units (Ref 36).
bUn1ts reported 1n reference cited (Mass of POM emitted per unit of fuel burned).
cSee Appendix A for sample calculation. Fuel heating values assumed for conversions: Coal = 11,500 Btu/lb;
Oil = 150,000 Btu/gal; Gas = 35,300 Btu/m } and Wood * 8600 Btu/lb.
dAs reported in reference cited for emission factor.
el = weighted average reflecting boiler population.
2 = arithmetic or geometric average of boiler types.
3 = one fadl 1ty.
fDNL = Data not located. The emission factor for gas-fired utilities 1s expected to be lower than for coal
or oil because of the high efficiency combustion achieved in gas-fired utility boilers. Notice that the
emission factor for Industrial boiler gas combustion is significantly lower than for Industrial o1l-f1red
boilers.
9The emission factor for commercial/Institutional coal combustion is expected to be higher than for
utilities due to the larger fuel feed size and less efficient combustion associated with commercial units.
^The emission factor represents an average of data for 7 wood-fired Industrial boilers, 3 with mechanical
collectors and 4 uncontrolled. On controlled units, emissions were measured downstream of the collection
device.
fThe fact that the estimated emission factor for oil 1s higher than for coal may reflect the fact that ESPs
applied to coal-fired units achieve a certain degree of fine particle (and thus POM) control. The reported
emission factor for the oil-fired utility boiler reflects a "no control" situation.
JThe emission factors reported for industrial oil- and gas-fired boilers were used for commercial/1nstu-
tional oil- and gas-fired boilers, since the Industrial boilers tested were relatively small and represen-
tative of commercial/institutional units.
kRange of POM emission factors reported In literature was 0.05 to .37 g/kg of wood burned. The emission
factor selected represents an average of six wel1-documented tests on representative stoves.
5-32
-------
RADIAN
The methods used In past source tests to sample for and analyze POM
compounds from combustion sources have varied considerably with respect to
sample collection, preservation, and preparation, and with respect to compo-
nent analysis techniques. Thus, it is difficult to develop valid comparisons
between emission factor data for various source categories. One must rely to
a large degree on the theoretical principles that describe POM formation in
combustion sources in assessing the relative potential for various categories
to be significant sources of POM emissions.
Boiler design and operating parameters (including firing method, combus-
tion zone temperatures, fuel/air mixing parameters, fuel/air ratio, and fuel
properties) influence POM formation and emissions. Other factors being con-
stant, boilers that are designed and operated in a manner that maximizes
combustion efficiency may have lower POM emission factors than inefficiently
operated units. However, quantitative data on boiler design and operating
practices for specific boiler categories and the relative effect of these
practices on POM emissions are lacking. The best emission factors for the
purposes of this study were considered to be those which reflected a "weighted
average" of emission factors for the boiler population. These factors were
available only for coal-fired utility and industrial boilers. Other published
emission factors represented arithmetic or geometric averages of emission
factors for various boiler types within a category. And, in some instances,
available emission factors were based only on data from one facility.
Details on the type of emission control equipment used on the sources
tested and on the design and operating parameters of the control equipment
were not typically included with the "controlled" emission factor data
reported in the literature. Clearly, these parameters would be useful in
determining how representative the sources tested are with respect to the
current boiler population. Furthermore, quantitative data on the effect of
particulate matter, S02> and NOX control devices on POM emissions were not
located in the literature reviewed. Therefore, it was not feasible to use
uncontrolled POM emission factors combined with estimated POM removal effi-
ciencies of the control devices currently in use to develop total emissions.
5-33
-------
Despite these caveats* the national emission estimates presented in the
next section should be useful for preliminary evaluations of POM source cate-
gories. They will be most useful in pinpointing combustion source categories
with the greatest potential for POM emissions. They are also appropriate for
providing an idea of the amount of POM that could be controlled by regulating
various categories.
5.3.3 National Emission Estimates
Table 5-10 shows 1980 national estimated POM emissions from the various
combustion source categories considered in this study. The emission factors
and fuel consumption data used to calculate the national emissions are also
shown.
Estimated total POM emissions from wood-fired stoves account for 90
percent of annual POM emissions from all combustion sources. The emission
factor for wood stoves is four to five orders of magnitude higher than those
for virtually all other combustion sources except coal-fired stoves and wood
fireplaces. The higher POM emission factor is attributable to highly ineffi-
cient combustion in wood stoves.
The next most significant combustion source categories of POM are resi-
dential coal stoves and wood-burning fireplaces* which together account for
8.6 percent of the toal estimated POM from combustion sources. Thus, resi-
dential sources are estimated to account for 99 percent of POM emissions from
fuel combustion.
In terms of absolute magnitude of emissions* the other significant com-
bustion source category is wood-fired industrial boilers. However, emissions
from this category may be over-estimated somewhat because they are based on a
POM emission factor that represents only the application of mechanical collec-
tor particulate emission controls. Many of the larger wood-fired boilers are
equipped with ESP or wet scrubbers which could provide significantly more POM
control than mechanical collectors. These same cautions about overstating
b-34
-------
TABLE 5-10. 1980 NATIONAL TOTAL POM EMISSIONS ESTIMATES FOR COMBUSTION SOURCES
Source Category
Estimated 1980
U.S. Fuel Consumption
101SJ (1012
Emission Factor
pg/J CIb/1012 Btu)
National
Emissions
Mg/yr (tpy)
% of Total
from Combustion
Sources
Utility Combustion
Coal
011
Gas
Combustion
Industrial
Coal
011
Gas
Wood
Commercial/Institutional
Combustion
Coal
011
Gas
12,855 (12,150)
2592 (2450)
4031 (3810)
1009 (954)e
959 (906)e
3555 (3360)e
579 (547)f
100 (94.7)9
599 (566)
2825 (2670)
0.71 (1.65)
0.86 (2.00)
0.52 (1.2) est.b
1.53 (3.57)
0.50 (1.17)
0.30 (0.69)
103 (237)
1.7 (4.0) est.c
0.50 (1.17)
0.30 (0.69)
9.1 (10.0)
2.2 (2.4)
2.1 (2.3)
1.5 (1.7)
0.48 (0.53)
1.1 (1.2)
59.6 (65.6)
0.17 (0.19)
0.30 (0.33)
0.84 (0.92)
0.12
0.03
0.03
(0.18)
0.02
0.01
0.01
0.76
(0.80)
rfeg
neg
neg
neg
Residential Combustion
Coal
011
Gas
Wood Stoves
Wood Fireplaces
TOTAL
68.8 (65.0)
1310 (1241)
5152 (4870)
520 (492)"
346 (328)n
2500 (5826)
2.86 (6.67)
1.74 (4.06)
13,500 (31,400)d
1450 (3370)d
172 (189)
3.8 (4.2)
14.7 (16.2)
7022 (7733)
502 (553)
7792 (8581)
2.2
0.05
0.19
90.1
—£u±-
(99.0)
100.0
Reference 15, pp. 5-9, except as noted. More data on fuel consumption estimates can be found 1n Appendix
A.
''Estimated based on the difference between reported emission factors for oil- and gas-fired Industrial
boilers.
GEs1mated to be slightly higher than emission factor for coal-fired Industrial boilers.
dRounded to 3 significant figures from Table 5-9.
eTotal Industrial fuel consumption values from Reference 15 adjusted to reflect only amount used 1n boilers.
(See Reference 17 note.) Adjustment factors (% total consumption used 1n boilers): Coal - 30!5j Oil - 67!5;
and Gas - 40%.
^Reference 40. Total estimated heat Input capacity for wood-fired boilers multiplied by 60/5 load factor to
determine annual Btu consumption.
9A11 reported commercial/Institutional combustion assumed to be in boilers.
Reference 37.
5-35
-------
emissions also apply to those estimates presented in Table 5-10 for industrial
coal-fired boilers. However, as discussed previously, the published POM
emission factor data used may not reflect that portion of POM emitted as vapor
at typical stack temperatures, thus underestimating actual POM emissions
from a single boiler.
The estimates in Table 5-10 indicate that utility boilers account for
only about 0.2 percent of total POM emissions from the combustion source cate-
gories examined. There is, however, substantial uncertainty related to (1)
the POM emission factor data, (2) the (indirect) effect of particulate and S02
controls on POM emissions, and (3) the fuel consumption data used in develop-
ing the national estimates. Thus, it is appropriate to conduct a sensitivity
analysis using the extremes of reported data and making some "worst-case"
assumptions. For example, one recently debated topic is the potential impact
on POM emissions of existing oil- and gas-fired utilities converting to coal.
These concerns are related to the fact that combustion of coal results in
higher POM emissions per unit heat input than combustion of oil or gas in
similar units.
In the "worst case", assume that all 1980 fuel consumption by utilities
was in the form of coal (about 19,500 x 1015J based on Table 5-10). By multi-
plying this consumption by the highest POM emission factor for pulverized coal
combustion in utility units (controlled with an ESP, since coal conversions
would likely be subject to at least SIP particulate emission limits), a "worst
case" estimate can be developed:
A.o Pq POM Vl9,500 x io15J\ (metric ton] , ig>5 tons pQM
V J C0al burnedA Yr J VlO18 pg J Yr
[This emission factor is the "maximum" reported in Reference 1, Table III-l,
converted to pg/J using a coal heating value of 26,800 kJ/kg (11,500 Btu/lb)].
Even with these worst case assumptions, emissions from the utility category
would be only about 0.26 percent of the estimated total for combustion cate-
gories.
5-36
-------
Another category where a sensitivity analysis is appropriate is combus-
tion of wood in residential stoves. Emission factors for POM reported in the
literature ranged from 0.05 to 0.37 g/kg wood burned (References 42 and 36,
respectively). Using these emission factors and the fuel consumption data
shown in Table 5-10, a range of estimates between 1300 and 9600 metric tons
per year can be developed [based on a wood heating value of about 20*041 kJ/kg
(8600 Btu/lb of dry wood.)] The fact that such a range exists is clearly
indicative of the variability and uncertainty of POM emission factors and the
associated impact on the relative significance of various source categories.
Table 5-11 presents a limited comparison of national emission estimates
developed in this study with previously published estimates. It is beyond the
scope of this study to investigate the differences between these estimates in
detail. However, the estimates appear to" be relatively consistent with two
exceptions. Reference 36 reports significantly higher emissions from coal-
fired industrial boilers. The estimate was based on a 56 pg/J emission fac-
tor, versus 1.5 pg/J reported in Reference 1, and was representative only of a
pulverized dry bottom industrial boiler. The 1.5 pg/J factor appears to be
more representative. It is about twice the emission factor reported for
utility pulverized coal units, reflecting primarily the difference in emission
control device application between the utility and industrial sectors. Anoth-
er indication that emissions from coal-fired utility and industrial boilers
are similar is benzo(a)pyrene (BaP) emission data reported in Reference 2.
The BaP factor for utility coal combustion was reported to be about 0.05 pg/J
and for industrial units it was reported to be about 0.03 pg/J.
The other discrepancy observed in Table 5-11 is the difference in esti-
mates for wood stoves. However, this difference can be accounted for by the
differences in fuel consumption data used to develop the estimates. Reference
36 estimated 1976 residential wood combustion at about 16 million metric tons.
However, recent data for 1980 indicate that that figure has increased to about
26 million metric tons (37). More information on trends in residential wood
combustion is presented in Section 5.3.4.
5-37
-------
TABLE 5-11. COMPARISON OF TOTAL POM EMISSION ESTIMATES WITH THOSE
REPORTED IN PREVIOUS STUDIES
Combustion
Source
Category
Utility
Coal
Oil
Gas
Industrial
Coal
Oil
Gas
Wood
Estimate in
Current Study
1980
(metric tons/yr)
9.1
2.2
2.1
1.5
0.48
1.1
59.6
Reference 36
-1976
(metric tons/yr)
12.9
0.3
0.3
69. Oa
1.3
2.1
1.2
Reference 41
-1976
(metric tons/yr)
9.5
0.3
0.3
15.7
1.2
2.0
DNR
Commercial/
Institutional
Coal
Oil
Gas
0.17
0.30
0.84
DNR
3.1
1.8
1.2
Residential
Coal
Oil
Gas
Wood Stoves
Wood Fireplaces
Total
172
7.2
14.7
7022
502
7795
102
7.4
9.8
3759°
78
4042
3664b
4.7
5.7
69.8
3779d
a
Emissions reported in Ref. 36 based solely on pulverized coal dry bottom
boiler, emission factor of 56 pg/J versus 1.5 pg/J as reported in Reference 1
as an average for all boiler types. The 1.5 pg/J emission factor is about
twice that reported for coal-fired utility units, reflecting primarily the
difference in control devices (See Table 5-9).
^No documentation of emission factor or production data used to calculate
emissions
cBased on about 16 million metric tons of wood consumed in 1976, vs. 1980
consumption which was about 26 million metric tons.
documentation of emission factor or production data accompanied report.
5-38
-------
CORPCMMTIOM
One final consideration that should be kept 1n mind when reviewing the
national estimates 1s that emissions from residential combustion sources used
for heating vary considerably with respect to the season. An example of this
variation is shown in Figure 5-1 for wood consumption in fireplaces and
stoves. These data show the difference in wood consumption and estimated POM
emisions between winter and summer months. Similar trends can be expected for
emissions from residential coal, oil, and gas combustion sources used primar-
ily for heating.
5.3.4 Trends Influencing POM Emissions
Utility Combustion - POM emissions from utility combustion will be
influenced primarily by three factors: the type of fuel burned, emission ,
controls used, and growth in new plant construction.
With respect to the type of fuel burned, the brief sensitivity analysis
described above indicates that even if all existing oil- and gas-fired utility
capacity were converted to coal, POM emissions from utilities would still
comprise only a small fraction of the total POM emissions from combustion
source categories. And, although many new utility plants will be designed to
burn coal, those plants will be subject to the relatively stringent particu-
late emission limit [13 ng/J (0.03 lb/106 Btu)] specified in the 1979 NSPS
applicable to coal-fired utilities. Compliance with this standard will re-
quire the use of a high-efficiency ESP or a fabric filter, both of which
provide over 90 percent control of fine particles (43). As discussed above
control of fine particles will result in substantial control of POM emissions.
Projections for new utility construction have generally been revised
downward in recent years because conservation has reduced the need for new
electrical generating capacity.
5-39
-------
*;
,
•
C
C
C
f
0)
o
CO
1— 1
0.
01
•H
Pm
^
D
3
^
0
1-4
-------
RADIAN
CORPORATION
Industrial Combustion
Industrial growth will be accompanied by the need for additional indus-
trial boiler capacity. The same factors influencing POM emisions from utility
combustion are applicable to this category: type of fuel burned* emission
controls used* and projected growth rates.
Despite possible economic advantages of using coal rather than oil or
natural gas, there are several potential impediments in the current regulatory
environment to a widespread or vigorous increase in industrial coal use. One
recent study (44) found that about 70 percent of industrial facilities are
located in non-attainment areas. The authors noted that obtaining permits for
construction of new coal-fired boilers (or conversion of existing units to
coal) in these areas will be difficult because of an "apparent unwillingness
of existing industries" to provide emission offsets to competing firms. In
addition, the concentration of industries-in urban areas makes coal burning
subject to strong local opposition if efficient emission controls are not
applied. The capital investment required for the emission control systems,
combined with the relatively large capital investment for the boiler itself,
may, in many cases, make coal less economically attractive to the industrial
user.
When these economic factors are combined with traditional uncertainties
in securing an uninterrupted supply of coal (e.g. strikes, bad weather), there
appears to be only certain cases where industrial coal use would be favored.
Using specific case studies, the report cited above (44) concluded that coal
use appears most economical for energy-intensive industries, such as petroleum
refining and primary metals, that are located near suitable coal supplies.
The study also concluded that the use of coal in large energy-intensive indus-
tries located along the Gulf Coast (e.g., chemicals and primary aluminum
plants) is less likely to be cost-effective in the near term. For smaller,
less energy-intensive industries, capital requirements may be unjustifiable,
regardless of location.
5-41
-------
RADIAN
CORPORATION
Future fuel choice decisions for industrial boilers will be significantly
affected by oil and gas prices* federal energy policies* coal transportation
costs, technology developments and environmental regulations. It 1s difficult
to assess what emission control requirements will be applicable to new indus-
trial boilers, since the NSPS for that category has yet to be proposed.
However, the NSPS will likely require a higher degree of particulate emission
control than current state regulations. Also, as discussed above, local air
quality considerations, especially in non-attainment areas, are likely to
require the use of efficient emission controls on new boilers.
Figure 5-2 shows the growth in industrial wood consumption over the past
30 years. (The consumption values in Figure 5-2 include wood fired in process
heaters, which was not included in the calculation of national POM emissions
from wood combustion presented in the previous section.) Figure 5-2 shows that
industrial wood consumption has been fairly steady for the last four years.
The pulp and paper and lumber and wood products industries* major users of
wood-fired boilers, are tied directly to the housing industry. Process tech-
nology changes, energy conservation, and a shift away from the most energy
intensive products may result in some reduction in energy consumption in the
pulp and paper and related industries (37). One source predicts that 1950 MW
(6.65 x 10^ Btu/hr) of new wood-fired boiler capacity will be installed
between 1982 and 1990 (45). The added capacity is foreseen as a result of
growth in the pulp and paper and lumber industries and trends in these indus-
tries toward replacement of fossil fuel-fired boilers with wood units.
Replacement of older existing wood-fired boilers will also account for some of
the new capacity installed.
An NSPS for wood-fired industrial boilers has not been developed. The
same local air quality considerations discussed above for coal-fired boilers
would apply to construction and regulation of new wood-fired units.
5-42
-------
1970 '71 '72 '73 '74 '75 '76 '77 '73 '79 '80 '81
30'
1950
1960
Figure 5-2. Industrial Wood Consumption Trends (37)
*Includes consumption in boilers and process heaters.
5-43
-------
Commercial/Institutional Combustion
No quantitative data were located on the expected growth 1n commercial/
institutional combustion. However* some qualitative assessments are possible.
Continued population growth in urban areas will result in construction of new
commercial and institutional boilers. Fuel choice decisions for these boilers
will be driven by the same factors that affect industrial fuel choices:
regional location, oil and gas prices* energy policies* coal transportation
costs* technology developments* and environmental regulations. However*
development of alternate energy sources, such as solar heating, and energy
conservation practices may reduce the growth in this category* however.
New commercial/Institutional boilers with heat input capacities of great-
er than 14.7 MW (50 x 106 Btu/hr) will likely be required to use the same
particulate control systems specified for industrial boilers under the NSPS
being developed (33). However* many commercial boilers are smaller than 14.7
MW and will likely be subject only to SIP emission limits. In most states*
particulate emission limits are significantly less stringent for small boil-
ers.
Residential Combustion
Because of the potential for significant POM emissions* the predicted
growth of wood-fired stoves is of particular concern. Figure 5-3 shows 30-yr
trends for residential wood combustion in stoves. These data support the
analysis that although residential wood consumption dropped sharply in the
50's and 60's increasing oil and gas prices resulted in a significant increase
in the use of wood in the residential sector during the last decade.
It seems likely* that if oil* gas* and electricity prices continue to
increase* residential wood consumption will increase* especially as more
efficient stoves are introduced. However* the use of alternate energy sources
such as solar heating could become competitive with wood.
5-44
-------
35
30
25
20
w
0)
o
(0 >>
•H ^»
=15
§ §
H £10
li
03
c
o
T3
0 =;
5
0 1
1950
1960
1970 '71 '72 '73 '74 '75 '76 '77 '78 '79 '80 '81
Figure 5-3. Trends 1n Residential Wood Consumption (Stoves) (37)
5-45
-------
RADIAN
Because of the uncertainty associated with predicting residential fuel
prices and fuel choice decisions it is difficult to estimate future levels of
residential wood consumption. To provide an idea of the potential increase in
POM from residential wood combustion in stoves* it can be assumed that the
consumption levels will grow at annual rate equal to that demonstrated between
1980 and 1981 (about 1.4 percent). Assuming that no significant changes occur
in control of POM from wood stoves, annual emissions by 1990 could potentially
reach 8160 metr1c/tons/yr (assuming 0.27 gm POM emitted per kg of wood
burned).
5.4 COKE PRODUCTION
This section deals with POM emissions from by-product coke production.
5.4.1 Source Category Description
5.4.1.1 Process Description*
Coke production is an integral part of iron and steel manufacturing.
Coke provides the heat and carbon for the smelting and reduction reactions
that occur in furnaces. About 93 percent of the coke produced is used to
convert iron ore into iron. Iron foundries, nonferrous smelters* and chemi-
cals plants account for the remainder of coke consumption.
By-product coke production is carried out in enclosed slot-type ovens.
There can be 10 to 100 ovens per coke battery and there are usually several
batteries located at each plant. Figure 5-4 depicts a typical plant. The
major components of the by-product coke process (a batch-type process) are:
o charging the ovens with pulverized coal that has been
blended to the desired size and composition*
*The material presented in this subsection is summarized from References
46 and 47.
5-46
-------
RADIAN
CAMVCVM Ml IS
T 0,;; T
| IflMAfir I
L-rmJ
CAM WHM*f
rtUSHMK IKMMM
r~i^—
SIMIACC I
^'"J "1 lofllVONAIlM
-H±l
run riMiHtn
MUttSlWM
on roA ruiL
tM*H AUttOMI*
LlOUOO
Sf I tin
IAHNS
Flow sheet showing the major steps in the by-product coking process.
(copyright 1971 by United States Steel Corporation)
Figure 5-4. By-Product Coking Process (46)
5-47
-------
RADIAN
comxxumoM
o coking reaction (coking time varies between 16 and 20
hours),
o pushing (after coking is complete, the coke is mechani-
cally pushed into quench cars for transport to the next
process step),
o quenching the hot coke with water, and
o recovery of by-products such as tar, ammonium sulfate,
benzene, and naphthalene
With the exception of by-product recovery, existing data indicate that each of
the process operations listed above is a source of POM emissions.
Fugitive emissions are associated with charging and leaks from coke oven
door, lids, and offtakes. (Leaks from lids and offtakes are often grouped
together under the category of topside leaks.) Battery stacks are located on
the ovens to provide a natural draft for the combustion gas that is used to
heat the battery. Oven gases leak through the oven walls and are emitted
through these stacks. Pushing results in fugitive emissions. In the quench
tower, emissions are carried up and out of the tower by the steam produced
during quenching of the hot coke. Finally, although by-product recovery is a
potential source of POM, very little data are available on this portion of the
process.
5.4.1.2 Emission Controls/Regulations
Table 5-12 shows the emission controls applicable to by-product coke pro-
duction. No quantitative date were located on the use of these systems on ex-
isting sources. However, between 80 and 90 percent of the existing capacity
is subject to some type of control requirements, either under SIPs or by spe-
cial consent decrees (49). The SIP and consent decree requirements vary from
being relatively stringent for new and some existing sources to no control
required (50).
5-48
-------
CORPORATION
x-s
oo
^^
•^
co
1-
^^
i
CL
LU
O
O
1—
0
Q
O
or
Q.
i
^-
03
O
1-
LU
CO
^£
o
1— 1
_J
CL
Q.
co
_J
o
"Z,
8
z
o
1— 1
CO
CO
i— i
2
LU
•
IN
i— 1
1
LT>
LU
wJ
<£
1-
c
o >^
0 U
C
0 -r-
-P 0
E £
t- >4-
•P LU
in
LU
in
0
Q-
•^
C
u
0
h-
r—
O
s_
-p
c
o
o
•»
1- t-
CL 0
in -p
<0 -P'
x>
O)
c o
i- -P
O) c
i_ i-
.c in
u c
o
0 In
O) in
-PS
CO 0
-P
c
0
c
•1—
(O
-p
c
o
o
i— 0)
c 0
t- 10 O
4— "O *r—
o c in >
n J_ 0
0) 0-0
o xj
C • XJ i—
(O >> 3 O
C i— i_ I-
0 -P U -P
•p x: in c
CO) O
I- 1- -P O
03 -P 0
E 30
0 -P
-o in 10
CO • T3
(O i— in i-
o t_ -o
0 o c
in o o 10
ZD -P T3 O
0
i_
3
-P -P i-
c a. o
0 IO £-
S 0 -P
c cm
T- t_ O 0
« o o o
-P -v T-
c -a -o >
o c c 0
O (O (O Q
in
(O
0
t-
§
Q
X)
O
LO
O
-P
LO
m
i
0
•p
in
•o
c
3
0
CL
O
O)
C
•r™
r«
IO
0
in
M-
o
c
O
-p
(O
u •
i- in
t— -*
CL (O
CL 0
< i—
•P
C
0
E
c
n—
IO
•P
C
o
o
in
^
fO
0
1
0
-^
i^
in
CL
0
o
•p
o
LO
1
lf_
o
in
= Q.
0 co
•0 LU
•r*
in -p
0 •
0 * in
.* 0
O T3 U
U C i-
s (0 >
0
t_ in -a
0 t-
> 0 i—
0 XI 0
^"^ t.
U) 3 -P
•o s_ c
0 O O
x: in u
in
•p 0
0 -P
•• S *
in xj
0 1-
!_ • -0
3 >, C
in i. to
o 0 u
i— -P
u -P 0
c 10 t-
LU XI (O
"U
c
(0
0 r- U)
t- O 0
3 t. O
•p -p 1-
CL C >
in , 0
XJ i— 4->
C i—
0 O
..- 0
o
C i—
0 O
3 ^
0--P
c
>> o
i. U
•o
•o
•P c
3 (O
<*%
0
•> i.
O) 3
c -p
•i- Q.
O O
c
0 0
3 i.
O" "^^
3
t-ff
TO U
^c
•z.
Q
O)
c
^j
(O
0 £-
J- O
-P
t- in
o co
0.
tn LU
i —
f— ^
10 in
f t-
0
C -Q
0 X)
> 3
O J-
o
c in
•i—
c
in i- •
jif in
u in i_
IO 0 (D
i- in -p
O IO r—
cn t-
C -P
i- in o
XI 3 •*—
O iO J-
-P x: xj
•0 X 10
a. 0 4-
r~>
o
£_
-P -P
C C
0 O
E 0
c in
t- s-
O C (D
0 10 Q
in
.*:
u
•p
co
>^
L.
0
^J
-P
iO
m
^c
•z.
o
M-
O
in
0
m
3
0
H
IO
>
0
-Q
0
•p
>» ^J
*0 0
-* E
i- C
1 »r-
»-- IO
-P
< C
•z. o
o u
^c
o
•p
u
3 >,
O 0
fl Q
^•> 0
CD CC
.
OO
0
u
c.
0
L.
0
0
a:
c
0
N
•^ • •
S- OO LO
0 ^t LO
-P
U 0 0
iO O O
S- C C
IO 0 0
x: L. L.
u 0 0
in 0 0
< a: a:
iO XJ O
•
0
XJ
IO
'Io
IO
-p
o
c
IO
•p
*o
Q
II
Q
5-49
-------
RADIAN
CORPORATION
There 1s currently a program underway at EPA to develop a NESHAPs for
coke ovens. The alternatives being considered for the NESHAPs, which would
apply to both existing and new sources, range from achieving 35 to 80 percent
reduction over current estimated levels of BSO emissions (51) from charging,
door leaks, and topside leaks (51).
The existing and developing regulations for coke ovens are generally in
the form of work practice standards because of the extreme difficulties
associated with measuring mass emissions from these sources. Typical
regulations for charging specify the amount of time that visible emissions can
occur during charging. Other regulations specify the allowable percent of
leaking doors (PLD), of leaking lids (PLL), and of leaking offtakes (PLO)
(50). Regulations for quench towers do, in some cases, specify mass emission
levels, but many just restrict the quality of water that can be used in
quenching (52).
The Occupational Health and Safety Administration (OSHA) has developed
and is enforcing regulations related to worker safety. These regulations
dictate certain work practice procedures (50).
5.4.1.3 Source Locations
By-product coke plants are located in 18 states. Most of the plants
are located near steel plants and coal supply points. Forty-seven of the
estimated 60 plants are owned by or affiliated with iron and steel firms (53).
In 1976, 57 percent of total U.S. coke was produced in Pennsylvania.
Ohio and Indiana were the next largest producers. Other states with signifi-
cant coke production capacity include Alabama, West Virginia, Maryland, New
York, and Michigan. The relative amount of coke produced in various states
has been fairly stable (53).
5-50
-------
RADIAN
5.4.2 Emission Factors
Most of the emission data reported for coke ovens are based on emissions
of BaP or benzene soluble organics (BSD). Very little total POM data are
available. Table 5-13 summarizes emission factor data from references
reviewed in this study. Total POM data were located only for door leaks
(uncontrolled and controlled) and quenching operations. Comparison of BSD
data available for the various process operations shows that charging* door
leaks» and quenching (with contaminated water) are the largest potential
emission sources.
Tables 5-14 through 5-17 show the type of POM compounds included in the
total POM data reported in Table 5-13.
5.4.3 National Emission Estimates
Development of national total POM emission estimates for coke production
is difficult and uncertain due to the lack of total POM data and the wide
variability of the limited data published. However* a very rough estimate of
the lower end of the range of national emissions can be developed by assuming
that POM emissions are at least equal to those from door leaks and quenching.
An emission factor of 4 gm/metric ton of coal charged is representative
of door leaks without an add-on control device. For quenching* an emission
factor of 2 gm/metric ton of coal charged was assumed. Thus* an estimated
total POM emission factor of 6 gm/metr1c ton of coal was used to calculate
national emissions. This factor is based on the use of relatively clean water
for quenching. Most states limit particulate emissions from quenching to 1.24
kg/metric ton of coal charged or less (54). However* compliance with such
limits is based on the use of clean water since emissions from these sources
are very difficult to measure accurately (55).
5-51
-------
ZT
0
1— 1
1
o
r>
Q
o
a:
CL
LiJ
8
ce:
o
1 1
^£
h"~
^£
Q
o:
o
o
i^
^*
o
1— 1
CO
co
i— t
?s*
LU
•
m
r-i
i
in
LU
_j
CD
«dt
h-
0
U
c
!_
0
t.
*^
0
cc
(/)
-p
c
0
E
O
^— v
•^J
0
0
L.
<0
JZ
O
U) r—
S-
LT>
u
p**.
l—<
o
•p
oo
OT
c
•1—
J=
Ul
CL
O^
in
o
o
o
oo
CN
O
-P
i-H
r— 1
in
0
o
o
•
o
0
a\
o
Ol
•
o
CN
O
O
O
O)
c
•r-
JZ
o
c
0
<§
r— f
VO
JZ
o
£Z
0
3
CT
1_
O
t|_
!_
0
•P
rO
*
C
>
(_
0
4J
-P
ro
CO
•
•o
0
1
>
0
t-
U)
0
u
c
0
L.
0
0
s_
c
•r-
•^
0
•P
ra
8
0
{-
0
X
ra
•p
ra
•o
s_
o
•p
O
rO
t^
c
0
In
in
1
0
o
c
-P
ra
JC
-p
0
•p
ro
U
•r—
T3
C
•r-
in
\r
C
rO
r—
CD
•
^~
i-H
1
in
0
r—~
JO
rO
C
T-
c
8
JZ
in
Ul
(0
Ul
c
3
0
Q.
U
^T
CL
m
i— i
in
0
•o
3
I —
U
c
1— 1
n
-p
c
0
E
Ul
Ul
0
Ul
Ul
"*••
•
\*
u
0
CD
•
_l
0
0
-1
T5
C
ra
•
cc
rl
t-
"Q
C
<
^
E
*o
0
s_
1-
• •
Ul
ra
c
0
>
i—
%
E
ra
JZ
I_
3
Q
-P
I_
O
CL
0
cr
^^
CL
LU
^_
ra
c
s_
0
•p
c
1— 1
«^
Ul
0
!„
0
-p
•p
ra
CO
c
0
o
0
o
0
0
a
H
|
+J
O
O)i—
in
•^
ra
•p
rO
"°
0
Ul
0
_c
-p
s_
o
4-
0
U
c
0
L.
0
4-
0
L.
l»
ro
C
•^
CO
J_
M-
m
c
o
Ul
Ul
g
LU
U
C
ro
D
L.
O
Ul
3
O
•a
£_
(O
N
ro
D1X
«r-
t-
0
u
M-
O
•
in
^-i
I
in
0
•
VO
.— I
in
0
JO
ro
in
T3
o
<•
-p
Z
<
CL
LU
T3
0
•r-
l^K
T3
O
E
ro
5
•i—
^
•o
0
-P
U
3
•o
c
o
o
Q)
C
•1—
"o.
E
rO
CO
•
0
O
U
c
0
0
J_
p-
1
0
1—
JO
rO
H
•
Ul
"O
c
3
O
CL
S
u
2jT
2
L.
rO
^_
0
CL.
CO
i— I
•^
C
rO
in
•a
c
3
O
Q.
O
i
CL
PO
m
t_
0
«4-
^*
^^
CL
^
^_
rO
C
O
ro
•P
ra
•o
Ul
-p
c
0
in
0
£_
CL
in
in
0
u
c
0
£_
0
t|_
0
cc:
•
in
-p
c
3
o
E
rO
i.
0
J=
01
•r-
JZ
i— h- roco
(-1
ra
h-
c
••"
c
8
x:
in
in
c
• 3
00 0
r- a.
i-H O
o
VO S!
^£
o ui
£_ 0
ro T3
2 3
» "u
O C
•Z. I-l
-o
c.
c
8
-C
Ul
Ul
•o
c
3
o
CL
U
•s.
£
o
1— 1
Ul
0
T3
3
13
c
1— 1
0 '
1
c
0
c
T3
C
ra
c
0
0
i-
01
t_
o
i^
Ul
•p
in
0
-p
4-
o
0
0)
ra
i-
0
>
<
4-
^T
V^
O
>^
JO
•a
0
N
>,
ro
C
rO
Ul
0
"cu
i
Ul
• •«
Ul
3
-P
ro
£_
ra
Q.
CL
rO
>%
"0
•f—
-P
rO
f—
0
U
c
•^
•^
c
3
O
**"
in
c
3
O
Q.
U
U)
g
JZ
Ul
•
in
-p
Ul
0
•P
0
5
c
L.
O
M-
T3
0
r"—
Q.
-------
TABLE 5-14. POM COMPOUNDS MEASURED IN UNCONTROLLED COKE OVEN DOOR LEAK SAMPLES
(59)
Benzo(a)phenanthrene
Benzo(e)pyrene
Benzofluoranthenes
Benzo(k)f1uoranthene
Chrysene
Dlbenzanthracenes
Dibenzpyrene
D1methy1benz(a)anthracene
FT uoranthene
Indeno(l,2»3-c,d)pyrene
Naphthalene
Pyrene
Benzo(a)pyrene
5-53
-------
CORPORATION
TABLE 5-15. POM COMPOUNDS MEASURED IN CONTROLLED DOOR LEAK SAMPLES (60)
Naphthalene
FTuoranthene
Pyrene
Benz(c)phenanthrene
Chrysene
Benz(a)anth racene
7,12-Dimethylbenz(a)anth racene
Benzof1uoranthenes
Benzo(a)pyrene
Benzo(e)pyrene
Choianthrene
Indeno(l,2,3-c,d)pyrene
Dibenz(a,h)anthracene
Dibenzacridines
Di benz(c,g)carbazole
Dibenzpyrenes
3-Methyl cholanthrene
5-54
-------
TABLE 5-16. POM COMPOUNDS IDENTIFIED IN COKE QUENCH TOWER EMISSIONS
INTERNAL EPA DATA (59)
Anthracene and phenanthrene
Methyl anthracenes
Fluoranthene
Pyrene
Methyl pyrene and fluoranthene
Benzo(c)phenanthrene
Chrysene and benz(a)anthracene
Methyl chyrsenes
Di methylbenz(a)anth racene
Benzo(a)pyrene
5-bo
-------
TABLE 5-17. SELECTED POM COMPOUNDS IN COKE QUENCH TOWER SAMPLES
PUBLISHED DATA (61)
Benzo(a)pyrene
3-Methyl cholanthrene
7,12-Dimethylbenz(a)anthracene
Dibenz(a,h)anthracene
Dibenzo(a,h)pyrene
Dibenzo(a»i)pyrene
Benz(a)anth racenes
Pyri dine
Indeno(l,2,3-c,d)pyrene
Phenanthrene
Phenol
Cresol
Quinoline
6-56
-------
RADIAN
CORPORATION
The emission factor was then multiplied by the 1980 coke production
level of 41.8 million metric tons (56) and a typical yield of coke from coal
of 69 percent (57) to yield a national estimate:
/ 6gm POM \ /ton coal charged | | 41.8 x 10 metric tons cokej
^metric ton coal charged! I 0.69 ton coke J \^ yr J
Jjneiric_iojiJ = 3fi3 m@trfc tons POM/yr (400 tons/yr)
v iou gm j
This estimate is likely to represent the low end of the estimated range
of national POM emissions from by-product coke production because only door
leaks and quenching tower emissions are included and it was assumed that clean
water was used in quenching.
5.4.4 Trends Influencing POM Emissions
The two major factors that will influence POM emissions from by-product
coke production are:
o proposal/promulgation of coke oven NESHAPs and
o coke production levels.
Published alternatives being considered for the coke oven NESHAPs are
expected to require between 35 to 80 percent reduction of BSO emissions over
current levels, depending on the emissions source and the alternative selected
(51). These standards would apply to both new and existing coke ovens.
Production levels are related to the iron and steel industry which, as
discussed in Section 5.5.4 1s expected to experience only moderate growth in
the next few years (1 to 3 percent annually).
5.5 IRON AND STEEL PROCESSES
This section describes, in general, some iron and steel processes that
potentially emit POM. The discussion focuses on iron and steel sintering and
5-57
-------
CORPORATION
ferroalloy production* two categories for which at least some Information on
POM emissions was located. Because there were Insufficient published total
POM emission data* the development of national emission estimates for iron and
steel sintering was not feasible.
5.5.1 Source Category Description
5.5.1.1 Process Description
Several iron and steel processes may potentially emit POM. Specific
information was located for iron and steel sintering and ferroalloy produc-
tion. The POM emissions from non-ferrous alloy production and other metals
manufacturing may be similar to those described for ferroalloy production.
Sintering—Sintering is the process by which pulverized ore is agglo-
merated before it is used in a blast furnace. A traveling grate moves the ore
over a series of wind boxes where air is pulled through the grate and the ore
is ignited with a burner (62). The ore contains coke and oil scrap and POM
compounds are generated from the burning of this material. Also* coke and
scrap particles with absorbed POM can escape at several points in the process
(62). Much of the POM may be emitted as vapor at sintering process tempera-
tures (62).
Ferroalloy Production—In production of ferroalloys, reducing materials
such as coal or coke and ores are charged to a furnace on a continuous or
cyclic basis. Furnace operation is continuous. In the most commonly used
process (electric arc furnaces) reactions occur in the zones surrounding large
electrodes. Burning of the coal or coke generates carbon monoxide, particu-
late matter, vaporous metallic compounds, and POM (63) which are emitted
through the furnace exhaust. Electric arc furnaces which can be open, closed,
or semi-sealed are also used in ferrous foundries, specialty and alloy steel
production, and nonferrous alloy production.
5-58
-------
5.5.1.2 Emission Controls/Regulations
Sintering—In 1976, two-thirds of the existing Iron and steel sintering
process factilities had no emission controls (62). Possible control technol-
ogies include ESPs, fabric filters, and wet scrubber, all of which would
capture significant amounts of POM. Wet scrubbers would likely be the most
efficient in reducing the total POM because vapor phase compounds would be
condensed in the scrubber. No federal NSPS exists for iron and steel sinter-
ing.
Ferroalloy Production—Exhaust gases from the electric arc furnaces are
typically vented to wet scrubbers or fabric filters, both of which should
provide significant POM control. Fabric filters are used at 26 of the 31
ferroalloy plants in the United States (63). A 1974 NSPS limits particulate
emissions, opacity, and carbon monoxide emissions from "new" electric arc
furnaces producing silicon-, manganese-, and chrome-based alloys. Federal
NSPS also exist for electric arc furnaces in iron and steel producion facil-
ities and non-ferrous alloy manufacture and an NSPS is being developed for
electric arc furnaces used in ferrous foundries (64). Most states also have
particulate emissions regulations applicable to existing metal production
facilities that have required the use of emission control equipment.
5.5.1.3 Source Locations
Sintering—Sintering occurs in conjunction with operation of large
blast furnaces. These plants are concentrated in Ohio, Pennsylvania, and
Indiana (62).
Ferroalloy Production—Production of ferroalloys is concentrated in the
same states as sintering (Ohio, Pennsylvania, and Indiana) as well as in West
Virginia, Alabama, Kentucky, and Tennessee. There are also large facilities
located in Colorado and California (65).
b-59
-------
RADIAN
5.5.2 Emission Factors
Sintering—Only BaP data were available for the sintering process (62).
The BaP emission estimates ranged from 0.66 to 1.1 gm per metric ton of
sinter feed.
Ferroalloy Production—POM emission data were collected from a semi-
sealed and a closed electric arc furnace in a recent study supporting review
of the ferroalloy NSPS (66). POM emissions from the semi-sealed furnace
(producing 50 percent FeSi) were 91.0 gm/Mw-h. This facility was controlled
with a 1ow energy wet scrubber followed by a flare. Emissions were measured
before the flare. Tests on the closed furnace (producing SiMn) showed con-
trolled POM emissions of 1.0 gm/Mw-h. The closed furnace was controlled by a
high pressure drop wet scrubber. Emissions from the closed furnace measured
before the scrubber were 156 gm/Mw-h when the furnace was producing FeMn.
Table 5-18 shows the compounds "detected in the furnace exhausts sampled.
No emission factor data were presented for open furnaces* which are used
1n the large majority of ferroalloy plants (65). However* one report reviewed
did present an annual nationwide POM estimate for open furnaces (see Section
5.5.3).
5.5.3 National Emission Estimates
As stated above* no useful total POM emission factors were available for
the iron and steel sintering. Therefore* development of a national POM emis-
sion estimates for this category was not feasible.
Reference 67 presented total uncontrolled POM emission estimates for
electric submerged-arc furnaces used in ferroalloy production:
5-60
-------
TABLE 5-18. POM EMISSIONS IN FERROALLOY ELECTRIC ARC FURNACE EXHAUST (66)
Semi sea led
Test C-2,
Furnace
during
50% FeS1 Production
FT uorene
Carbazole
Anthracene
Phenanthrene
Cycl openta ( def ) phenanth rene
Methyl anth racenes
FT uoranthene
Pyrene
Me thy Ipy rene
Benzo ( gh 1) f 1 uoranthene
Benzo(a) and benzo(b)f1 uorene
Chrysene
Benz(a) anthracene
Methylchrysenes
7,l2-d1methylbenz(a)anthracene
Benzotelpyrene
Benzofl uoranthene
Benzot 1 ) fl uoranthene
Benzo ( k ) f 1 uoranthene
Benzo(e)acephenanthylene
Perylene
Benzo(a)pyrene
Methyl benzopyrenes
3 -methy 1 chol anth rene
Indeno(l»2»3-c,d)pyrene
Benzo(gh1 )perylene
Anthanthrene
01benzo(ath)anthracene
D1benzo(c,g)carbazole
D1benzo(a1+ah)pyrenes
Coronene
TOTAL
mg/Nm3
75.0
b
18.3
18.3
10.7
1.5
27.4
28.5
0.07
8.9
1.24
8.1
10.5
0.81
3.5
0.16
3.5
0.43
1.64
1.10
3.2
0.83
1.0
g/MW-h
45.9
11.2
11.2
6.5
0.92
16.7
17.4
0.04
5.4
0.75
4.9
6.4
0.49
2.1
0.10
2.1
0.26
1.00
0.67
1.9
0.51
0.61
91.0
Cl c^e,^ Furnace
Test D-2, during Test D-l,
during
S1Mn Production FeMn Production3
mg/Nm3 g/MW-h mg/Nm3
1.5 0.36 16.0
9.6
2.1 0.51 220.0
0.070 0.017 24.0
0.24 0.058 220.0
0.22 0.053 2.3
0.005 0.0012 14.0
0.016 0.0039 49.0
5.2
0.58
51.0
3.1
1.2
0.39
6.0
1.4
0.90
0.079
O.S4
0.51
1.0
g/MW-h
4.0
2.4
54.9
6.0
54.9
0.57
3.5
12.0
1.3
0.14
13.0
0.77
0.30
0.10
1.5
0.35
0.22
0.020
0.13
0.13
156.0
aBefore scrubber.
bBlanks Indicate compound not detected.
5-61
-------
RADIAN
Estimated POM Emissions
Furnace Type (metric tons/vr)
Covered furnaces 209 to 1879
Open furnaces 135 to 1211
These calculations were based on actual emission factor data measured
upstream of emission control devices such as scrubbers and fabric filters.
However* it is likely that such control devices would substantially reduce the
amount of POM emitted to the atmosphere (67). Reported "organic" collection
efficiencies of the scrubbers on five furnaces tested ranged from 16 to 97
percent (67). As noted elsewhere in this report* fabric filters are expected
to achieve significant reductions in POM since the fine particulate onto which
POM is typically adsorbed is efficiently collected by fabric filters.
POM emissions were reportedly significantly higher for closed furnaces
than for open furnaces on the basis of POM emitted (uncontrolled) per unit of
furnace capacity.
5.5.4 Trends Influencing POM Emissions
Only low to moderate growth (1 to 3 percent annually) is foreseen for the
iron and steel Industry (68). Increased metals imports* high capital outlays
required for modernization expenditures to meet environmental regulations, and
the slow growth of steel-intensive industries are the major factors contribut-
ing to sluggish growth in this industry.
Any new capacity will be subject to the NSPS which generally requires the
use of fabric filters or high pressure drop wet scrubbers.
5.6 ASPHALT PRODUCTION - PAVING AND ROOFING
This section describes two asphalt-related sources of POM emissions:
manufacture of asphalt hot mix used in paving and air blowing of asphalt and
subsequent saturation of felt for asphalt roofing products.
5-62
-------
5.6.1 Source Category Description
5.6.1.1 Process Descriptions
Asphalt Hot Mix—Production of asphalt hot mix involves mixing asphalt
aggregates with hot liquid asphalt cement in a batch or continous process. A
large percentage of the hot mix is produced in existing plants using the batch
process shown in Figure 5-5a (71). The drum-mix process, in which wet aggre-
gate is dried and mixed with hot liquid asphalt cement as it is simultaneously
dried in a rotary dryer, has been used in most new plant designs (132). This
process is shown in Figure 5-5b. Drum-mix plants are also being used to re-
cycle salvaged asphalt pavement. Most asphalt mix plants are relatively
small: the average production rate is 160 metric tons/hr (176 tons/hr) (69).
POM compounds are emitted from combustion of gas or oil in the direct-feed
rotary dryer and from the mixer. Because the drum-mix process is based on a
parallel flow design (i.e., hot gases and aggregate flow through the dryer in
the same direction), particulate emissions from this process are less than
from a conventional process. However, because the asphalt is heated to a
higher temperature for a longer time, the drum-mix process may result in
greater POM emissions.
Fugitive emissions of POM may result from asphalt loading and handling.
As with most batch processes, emissions are highly variable.
In recent years, virtually all new plants are drum-mix operations in
which mixing and drying of the aggregates occurs within the same vessel (132).
Asphalt Roofing Products—The two major steps in the production of
asphalt-based roofing products are air blowing of the asphalt and saturating a
felt with the air blown material.
Although the processes are not always carried out at the same plant, air
blowing is an integral part of felt production. Air blowing involves accom-
plishing oxidation by bubbling air through liquid asphalt at 220 to 260°C (428
5-63
-------
<£ O .
o <: *
UJ Q* LU
OS O on
O I Q£
o <^<:
< o
o
(3
O
O
•3
O
(-J
PH
•H
S
O
sa
o
+J
ctf
PQ
4-1
o
B
tfl
M
M
rt
O
O
pq
cfl
LO
I
oo
•H
5-64
-------
S
m
td
,c
P.
co
S
3
OJ
a
a)
S-i
cd
M-l
O
00
o
O
O
r-H
M
3
00
5-65
-------
RADIAN
to 500°F) for 1 to 4 hours. Emissions are highly variable due to the cyclic
nature of the process (70).
Saturating the felt involves dipping it in airblown asphalt* spraying it
with the asphalt, or, in some cases, a combination of both techniques. The
final felt product, with controlled thickness coating of asphalt, is made into
rolls or shingles (70).
5.6.1.2 Emission Controls/Regulations
Asphalt Hot Mix—Wet scrubbers and fabric filters are typically used at
asphalt hot mix plants (74). No data were located on the effect of these
devices on POM emissions from the hot mix process. However, it can be
expected that both types of control equipment will provide a significant
degree of control of process emissions of POM because of their relatively high
fine particle collection efficiencies. Fugitive emissions from hot mix
production, on the other hand, are largely uncontrolled.
An NSPS for asphalt hot mix production was proposed in June of 1973. The
NSPS limits particulate matter emissions and the opacity of outlet stack gases
from plants (75). The NSPS applies to facilities constructed after June 1973.
The NSPS is currently being reviewed by EPA (132).
Asphalt Roofing Products—Typically, controls for air blowing of as-
phalt consist of a primary cyclone (to catch larger particulate matter) fol-
lowed by a fume incinerator (70). No data were located regarding the effect
of such a control system on POM emissions.
Saturating operations are generally enclosed by fume hoods which are
vented to emission control devices. The control devices used on various
plants include low voltage ESPs, afterburners, and high energy air filter
(HEAP) systems. Limited data showed HEAP systems reduced total POM by about
70 percent, while direct-fired afterburners had no effect on POM (70).
0-66
-------
RADIAN
eomxMumoi
An NSPS limits participate emissions and opacity from saturators and
blowing operations used in manufacturing roofing products and constructed
after November 1980 (75). ESPs, high velocity air filters and afterburners
are expected to be used to meet this NSPS (75).
5.6.1.3 Source Locations
Asphalt Hot Mix—About 20 percent of the 4300 operating asphalt hot mix
plants are mobile (71). Most permanently installed plants are located in
urban areas. As of 1974, states with 5 percent or more of the national
asphalt hot mix capacity were California, Illinois, New York, Pennsylvania,
and Ohio (72).
Asphalt Roofing Products—The majority of asphalt roofing product
plants are also located in urban areas. Six large plants (out of an estimated
200 total plants) were reported to account for 20 percent of production in
1973 (73). Four states were reported to have 15 or more roofing product
plants: California, Illinois, Texas, and New Jersey (73).
5.6.2. Emission Factors
The limited emission factor data located for these categories are
presented below. In both cases, published data were limited to one or two
emission tests conducted at what were considered to be representative
facilities.
Asphalt Hot Mix—Controlled emissions from the stack exhaust and mixer
at a batch production facility were reported to be 13.0 mg per metric ton of
hot mix produced (76). The plant was equipped with a primary cyclone followed
by a wet scrubber. No information was provided on the emission test method
used. Table 5-19 shows the POM compounds identified in the stack exhaust.
Reported analytical errors ranged from 6 to 144 percent for the different
compounds. No fugitive emission factor data were located, nor were data
b-67
-------
TABLE 5-19. POM COMPOUNDS IDENTIFIED IN STACK EXHAUST EMISSIONS FROM A
CONTROLLED ASPHALT HOT MIX PLANT (76)
Dlbenzothiophene
Anthracene/phenanthrene
Methyl anthracenes and phenanthrenes
9-Methy1anth racene
Fluoranthene
Pyrene
Benz(c)phenanth rene
Chrysene/benz(a)anthracene
7»12-Dimethylbenz(a)anthracene
3,4-Benzofluoranthene
Benzo(a)pyrene
Benzo(e)pyrene
Perylene
3-Methylcholanthrene
Dibenz(a,h)anthracene
Indeno(l,2,3-c,d)pyrene
7H-Dibenzo(c,g)carbazole
Dibenzo(a»h and a»i)pyrene
5-68
-------
RADIAN
available for the newer drum-mix facilities. [In the course of its NSPS
review* EPA plans to conduct emission tests at drum-mix facilities (132).]
Asphalt Roofing Products—Controlled POM emissions from air blowing and
saturator operations were reported as (77):
Air blowing: 2.1 mg POM/metric ton felt produced
Saturators: 4.3 mg POM/metric ton felt produced
These data represent the average emissions from two representative facil-
ities. The air blowing operations tested were equipped with fume incinera-
tors. At one air blowing operation, POM compounds were reduced by about 80
percent across the fume incinerator, but at the other facility POM increased
by a factor of two across the incinerator (77).
One of the saturators tested was equipped with a HEAP system which showed
70 percent reduction of POM. The other saturator tested was equipped with a
fume incinerator which reportedly had no effect on POM emissions (77).
The data were collected using a Method 5 sampling train modified by
placing the filter after a series of impingers. Table 5-20 shows the POM
compounds identified in samples from the air blowing operations tested.
5.6.3 National Emission Estimates
The emission estimates for these categories were calculated by
multiplying estimated 1980 national production levels by the emission factors
presented above.
Asphalt Hot Mix—Annual production of hot mix in 1980 was reported to
be 22.1 x 106 metric tons (24.3 x 106 tons) (78). This figure agrees rela-
tively well with a 1976 production level of 20 x 10^ metric tons used in an
earlier study (79), but is an order of magnitude lower than a 1975 estimated
production level of 300 x 106 metric tons attributed to the National
b-69
-------
TABLE 5-20. POM COMPOUNDS IDENTIFIED IN SAMPLES FROM ASPHALT
AIR BLOWING AND FELT SATURATION (77)a
Benz(c)phenanthrene
7,12-01methyl(a)anthracene
Benzo(a)pyrene
Benzo(e)pyrene
3-Methylcholanthrene
D1benz(a»h)pyrene
Dibenz(a,1)pyrene
aSample extracted with methylene chloride and analyzed by GC/MS.
b-70
-------
Asphalt Pavement Association (80). Since the 1980 estimate was more recent
and since the original reference for the higher estimate was not readily
available* the 1980 estimate was used in calculating national emissions:
|22.1 x 1Q6 metric tons hot mixj /13.0 x
I yr J ^metric
10 "3 am PONM /metric ton!
metric ton hot mixy \ 10° gm
= 0.29 metric ton/yr (0.32 ton/yr)
These estimates are based on the emission factors for batch plants and do not
reflect emissions from newer drum-mix plants.
Asphalt Roofing Products—In 1980, an estimated 5.3 x 10& metric tons
(5.8 x 106 tons) of asphalt roofing products were manufactured (78).
Based on the controlled emission estimates presented above national POM
emissions were calculated as follows:
A. 01 4 /5.3 x 106 metric tons fe1t\ [2.1 x 10"3 am POM) /metric ton]
Air Blowing: ^ ^^j ^etrlc ton felt j ( lO* gm )
= 0.011 metric ton/yr (0.025 ton/yr)
04. 4. /5.3 x 106 metric tons fe1t\ (4.3 x 1Q"3 gm POM ) /metric ton |
Saturators: ^ yp J ^ metric ton feu;^l()6gm )
= 0.023 metric ton/yr (0.012 ton/yr)
Total= 0.011 + 0.023 = 0.034 metric ton/yr (0.037 ton/yr)
5.6.4 Trends Influencing POM Emissions
Asphalt Hot Mix—Asphalt hot mix production levels are related directly
to the paving industry. Production levels are likely to remain relatively
steady (79). New or modified plants will be covered by the existing NSPS
which requires particulate emission control. Also, as mentioned above, newer
facilities are likely to be drum-mix designs.
5-71
-------
CORPORATION
Asphalt Roofing Products—Production levels of roofing materials are
dependent on the housing industry and there is also a large replacement market
for these products. The 1980 U.S. Industrial Outlook (81) predicts a 2 per-
cent growth in new construction through 1984, although the growth in this
industry is highly dependent on the general state of the economy and on mort-
gage loan interest rates. No information was located regarding the capability
of existing roofing materials production facilities to keep pace with the
anticipated housing industry growth. New or modified plants will be subject
to the NSPS for this category.
5.7 CATALYTIC CRACKING IN PETROLEUM PRODUCTION
This section deals with POM emissions from catalytic cracking as used in
petroleum refining. Catalytic cracking is used to upgrade heavy petroleum
fractions to produce high octane gasoline and distillate fuels.
5.7.1 Source Category Description
5.7.1.1 Process Descriptions
There are three basic types of catalytic cracking techniques currently
used in the petroleum industry:
o fluidized catalytic cracking,
o thermofor catalytic cracking, and
o houdriflow catalytic cracking.
Fluidized catalytic cracking accounted for about 94 percent of the total
refinery catalytic cracking in 1977 (82). The basic process involves heating
a mixture of gas and oil to about 480°C over a silica alumina catalyst and
then fractionating the mixture (82). The spent catalyst is regenerated in a
kiln by burning off the coke that has coated the catalyst particles during
cracking. The regeneration process in carried out about 540°C. Venting of
5-72
-------
CORPORATION
the exhaust gases from the regeneration kiln is a potential source of POM
emissions.
5.7.1.2 Emission Controls/Regulations
In many cases, exhaust gases from the regenerator are vented directly to
a carbon monoxide (CO) waste heat boiler to recover the useful energy in the
gases. CO boilers which are fired with auxiliary fuel or contain a catalyst
have been reported to be 99 percent efficient in reducing PAH emissions (82).
Direct-fired afterburners (plume burners) are reportedly not efficient con-
trols for POM.
Catalytic cracking units constructed after June 1973 are subject to an
NSPS that limits CO and particulate matter emission limits (83). The NSPS
generally requires the use of a CO waste heat boiler and an ESP. State
requirements applicable to new and existing catalytic cracking units also
generally require some degree of particulate matter and CO control (84).
5.7.1.3 Source Locations
Refining capacity is centered in Texas, Louisiana, and California, but
most states have at least one refinery (82).
5.7.2 Emission Factors
The most representative POM emission factor would appear to be one that
represents the use of a CO waste heat boiler. A weighted average of the
controlled emission factor data for the three types of catalytic cracking
techniques was reported to be 930 yg POM per cubic meter of oil cracked
(intermediate value in Reference 85). No information was provided regarding
the emission test method or the POM compounds included in the sample.
t>-73
-------
RADIAN
5.7.3 National Emission Estimates
National POM emission estimates for catalytic cracking were derived by
multiplying the emission factor given above by the reported quantity of oil
subject to catalytic cracking in 1980.
The Oil and Gas Journal (86) reports that as of January 1981, U.S.
catalytic cracking levels totaled 1,017,891 m3/stream-day (6,401,833
bbl/stream day). This figure includes fresh and recycle streams. A stream
day is defined as operation at full capacity for short periods. Catalytic
cracking units generally operate at stream-day capacity about 90 percent of
the time (86).
Based on this information, national POM emissions from catalytic cracking
were estimated as follows:
^1,017.891 m3 cracked\ L9 stream-dav\ /365days\ 1930 x 1Q~6 am Wrnetrictgnl
U
"A1
I stream day I \ day I \ yr J \m^ oil cracked J\ 106 gm
= 0.31 metric tons/yr (0.34 tons/yr)
5.7.4 Trends Influencing POM Emissions
Two factors will influence future POM emissions from catalytic cracking:
oil production levels subject to cracking and recent technology developments
in the cracking process itself.
Estimates of oil production levels fluctuate widely with changes in the
U.S. and world economic situations, but many forecasters predict a drop in
motor gasoline and distillate fuel production as a result of conservation
induced by price increases, more fuel efficient cars, and a switch from dis-
tillate oil to natural gas for home heating. The Oil and Gas Journal pre-
dicted a 2 to 3 percent drop in the amount of oil subject to cracking between
1981 and 1982 (87).
5-74
-------
RADIAN
CORPORATION
More complete combustion of CO to C02 in the catalytic cracking process
could lead to reduced POM emissions. Recent developments in this regard
include the UOP hot regeneration and the Amoco Ultracat process. The rela-
tively higher temperature for regeneration used in the UOP process improves
combustion efficiency and thus could reduce POM emissions. However* the high
temperatures require special materials of construction making the UOP process
more suitable for new units than retrofits. The Amoco process, however, is
based on improving the catalytic reactor efficiency, allowing more complete
combustion to occur in the regenerator without having to operate at high
temperatures (88). Thus, the technology similar to the Amoco process may be
suitable for retrofit to existing units.
5.8 COMBUSTION OF MUNICIPAL, INDUSTRIAL, AND COMMERCIAL WASTES
Combustion of waste can be carried out in boilers or incinerators. Waste
fuel-fired boilers recover the heat from combustion to generate steam (or, in
rare cases, electricity). The primary purposes of incinerators, on the other
hand, are to:
o reduce the volume of waste to be disposed of,
o reduce the toxicity of waste, or
o recover valuable resources from the waste.
Although POM emission data were located only for municipal solid waste
(MSW) and commercial incinerators, several other categories are discussed
below because of their similar potential for POM emissions. The categories
discussed here include:
o municipal solid waste combustion (incinerators and
boilers),
o industrial solid and liquid waste combustion (incinerators
and boilers), and
o commercial solid waste combustion.
s-75
-------
RADIAN
Hazardous waste Incinerators were not reviewed 1n this study because of
their extensive coverage under the current RCRA programs (89).
5.8.1 Source Category Description
5.8.1.1 Process Descriptions
POM emissions from waste combustion will be a function of waste composi-
tion and Incinerator or boiler design and operation.
Municipal Waste Combustion—Municipal waste is generally a heterogenous
mixture of wood, paper, metal, glass, and refuse. However, municipal waste
can contain industrial wastes, combustion of which may result 1n relatively
higher POM emissions (90). One source (91) estimates that in 1980 municipal
solid waste contained 5 percent plastics by weight, 20 percent of which was
polyvinyl chloride (PVC). Combustion of plastics may potentially increase POM
emissions. Increasing utilization of plastic products would result in even
higher concentrations. Another factor that may result in relatively higher
POM emissions from municipal waste combustion is the expected higher percent-
ages of organic matter, such as leaves and tree clippings, that is likely to
be present during the Fall.
There are several types of municipal incinerators (90). The waste can be
combusted on a moving belt, in drum-type rolling chambers, or on various types
of grates. Waste can be fed continuously, as is the case for all larger,
modern facilities, or it can be fed in batches, sometimes manually. Batch
operations are typical of older, smaller municipal incinerators many of which
have been shutdown (see following sections). About 150 to 200 percent excess
air is supplied to the incinerator [compared with the 30 to 50 percent excess
air typical of industrial coal-fired boilers (92)]. The large volumes of
combustible exhaust gases generated are sometimes burned in secondary chambers
(multichamber units). The exhaust gases, which contain POM resulting from the
incomplete combustion of organic material in the waste, exit through stacks to
the atmosphere. Large units operating continuously at high temperatures
5-76
-------
achieve more efficient combustion than smaller units and thus produce less POM
per mass of waste material combusted (90).
The average MSW incinerator can process about 16,000 kg/hr [385 metric
tons/day (424 tons/day)] of solid waste (90). At most large facilities, the
waste is burned as received with an effort made only to separate out very
large non-combustible items.
Boilers fired with MSW also burn the waste as received, with only minor
separation. Typical new MSW boilers can process about 181-2720 metric tons/
day (200 to 3000 tons/day) of refuse (133). Heat input capacities of exist-
ing units range from 1.3 to 85 MW (4.5 x 106 to 290 x 106 Btu/hr) (93). [85
MW is approximately equal to 545 metric tons/day of waste burned for these
units.]
Industrial Waste Combustion—Industrial wastes combusted in incinera-
tors consist primarily of processing wastes and plant refuse and contain
paper, plastic, rubber, textiles, and wood. Because of the variety of manu-
facturing operations, waste compositions are highly variable between plants,
but may be fairly consistent within a plant. Also, existing industrial
boilers are used to burn refuse-derived fuel (RDF), either alone or with coal.
In addition, a recenty published paper (89) reports that as much as 18
million metric tons (20 million tons) of hazardous wastes are burned annually
in existing industrial boilers.
Industrial incinerators are basically the same design as municipal
incinerators. Available data (based on a sample of over 300 units) indicate
that 91 percent of the units were multichamber designs, 8 percent were single
chamber designs, and 1 percent were rotary kiln or fluidized bed designs (94).
About 1500 of the estimated 3800 industrial incinerators are used for volume
reduction, 640 units (largely in the petroleum and chemical industries) are
used for toxidty reduction, and the remaining 1700 units are used for re-
source recovery, primarily at copper wire and electric motor plants. Some
5-77
-------
RADIAN
Industrial boilers are specifically designed to fire waste fuels. There are
about 50 of these units, ranging up to about 7.6 MW (26 x 106 Btu/hr) heat
input capacity (93).
Refuse-derived fuel is a solid waste fuel that has been processed or
classified to remove non-combustible materials. RDF can be co-fired with coal
in existing boilers or in boilers designed specifically for RDF. There are an
estimated six facilities either operational or under construction that will
fire only RDF (93).
Apparently some existing coal- and oil-fired industrial boilers are being
used to burn hazardous wastes (89). Characteristics of these boilers were
discussed in Section 5.3.
Commercial Waste Combustion—Incinerators are used to reduce the volume
of wastes from medical"facilities, large office complexes, schools, and com-
mercial facilities. Small multichamber incinerators are typically used and
over 90 percent of the units require firing of an auxiliary fuel. Because of
the inefficient combustion in these units, they can be a potentially signifi-
cant source of POM (95). Operating skills of commercial and industrial
incinerator operators are usually limited.
5.8.1.2 Emission Controls/Regulations
This section presents a summary of the emission controls typically used
on existing waste combustion sources and general information on the air emis-
sion regulations applicable to these sources.
No quantitative data on the effect of emission controls on POM emissions
were located, but the general principles described in Section 5.3 apply here:
o particulate matter control technologies designed to col-
lect fine particles (ESPs, fabric filters) will be effec-
tive in controlling POM compounds which preferentially
condense on smaller particles, and
5-78
-------
RADIAN
CORPORATION
o wet scrubbers can be effective in reducing POM emissions
that are 1n the vapor phase* because those compounds are
condensed as the gas Is cooled In the scrubber.
Highlights of the emission regulations applicable to existing and new
waste combustion sources are discussed below.
Municipal Waste Combustion—MSW incinerators and boilers firing MSW
with a capacity of 45 metric ton/day (50 tons/day) or greater and constructed
after August 1971 are subject to an NSPS (96) that requires about 93 percent
control of particulate emissions (97). Boilers and incinerators burning
municipal wastes are also subject to SIP emission and opacity limits in most
states. These are sometimes regulated together under the designation of "fuel-
burning equipment" (98).
New MSW-fired boilers will also be regulated under the industrial boiler
NSPS currently being developed by EPA. The proposed standard will likely
require the use of ESPs.
Industrial Waste Combustion—Industrial incinerators are subject to SIP
particulate and opacity standards in all 50 states. The typical standard
requires that outlet partlculate emissions be controlled to about 0.45
gr/Nm3 (0.2 gr/dscf) at 12 percent (X>2 (99), or slightly greater than 80
percent control (100). Opacity limits generally range between 10 and 30
percent (99). Most industrial incinerators are currently equipped with after-
burners, but new units may be required to have scrubbers or ESPs (101).
New industrial boilers firing waste or RDF will be subject to the NSPS
being developed by EPA, which will likely require the use of scrubbers or
ESPs. Existing industrial boilers are subject to SIP emission limits
depending on location, size, and fuel type. As discussed in Section 5.3.1.2,
emission controls used on existing sources vary from relatively inefficient
mechanical collectors to high efficiency ESPS and fabric filters. At least 40
5-79
-------
RADIAN
percent of the existing RDF and RDF/coal capacity is controlled with ESPs
(102).
Commercial Waste Combustion—No specific information was located for
these sources* but they may be subject to SIP particulate and/or opacity
levels. However* they are unlikely to be equipped with efficient emission
controls. Reference 95 reports that most "commercial" incinerators are
equipped with afterburners* but includes industrial incinerators in their
definition.
5.8.1.3 Source Locations
Municipal Waste Combustion—MSW incineration without heat recovery has
been declining in recent years due to the inability of older units to comply
with applicable air emission regulations. As of 1979-1980* there were slight-
ly more than 100 operating MSW incinerators spread over 19 states* most
located in the eastern and middle portions of the country (103). The highest
concentrations of these sources were in New York City and Chicago.
Most of the MSW-fired boilers are also located near urban population
centers (104).
Industrial Waste Combustion—No specific location data were available
for industrial incinerators or waste-fired boilers. However, the locations of
these units will obviously parallel those of the industries that rely on them.
The lumber and wood products, primary metal, and printing industries are
expected to be the major users of large incinerators. The lumber and wood
products industry is located primarily in the Southeast and Northwest. Pri-
mary metals are concentrated in the Appalachian and Great Lakes regions.
Commercial Waste Combustion—Concentration of these units will follow
urban population patterns.
5-80
-------
5.8.2 Emission Factors
Emission factors for total POM from waste combustion were available only
for MSW incinerators and commercial incinerators, and even these data were
quite limited. Most of the available data were originally reported in a 1967
study (105). No information was provided on the composition of the waste
fired during the tests, a major factor influencing POM emissions.
Table 5-21 shows available PAH emission data for four MSW incinerators
and a commercial incinerator. The PAH compounds reported are indicated in the
footnotes. Two previously described trends are supported by these data:
o Uncontrolled emission factors were highest for the commer-
cial incinerator, followed by the smaller MSW unit. The
larger MSW unit had the lowest uncontrolled emission
factor. These data, although very limited, tend to sup-
port the trend of less POM per unit of refuse burned'for
larger, more efficient incinerators.
o The scrubber and scrubber/ESP combination both showed
significant reduction of PAH relative to uncontrolled.
Emissions of polychlorinated biphenyl compounds (a subset of POM not
included in PAH) have been reported to be between 0.10 to 0.16 gm/metric ton
of municipal solid waste incinerated (based on two studies) (106).
5.8.3 National Emission Estimates
Development of national POM estimates was possible only for MSW incinera-
tors and commercial and industrial incinerators. Due to a lack of data re-
garding the amounts of waste combusted, emissions were developed by multiply-
ing the average unit capacity by the estimated number of units, an average
annual capacity factor, and an emission factor.
The average MSW incinerator capacity has been reported to be 385 metric
tons/day (424 tons/day) and there are an estimated 104 units nationwide (95).
A 70 percent average annual capacity factor was assumed in lieu of specific
5-81
-------
RADIAN
co
cr
o
l-
o:
LU
z
i— i
o
i— i
_,
^C
o
t—l
a:
LU
^r
^5»*
O
O
Q
"^
^c
^»
CO
2
o:
o
Lu
CO
cr
o
i-
o
^c
LJ_
z
o
I— 1
CO
co
r-l
2
LU
5
0.
•
r-l
CN
1
LT)
LU
I
CQ
*5f
1-
in c «-*
c o ~o
O -P 0
•r- -P
in u ro
in i- t-
•r- t. CD
E -P C
LU 0 T-
E U
$ S "r-
Q- CO
c
O r—
•r- O
in L.
i- C
E 0
LU O
^•x
>^
ro
"u
>^
>> c
-P 0
1- +J
o
rO O
d. *r-
ro L.
0 -P
0
E
<**.
0
O
0
J^
0
H-
0
Q£
**•»»
0
ex
£"
0
u
o
co
fO fl3 iO
O VO CN
r-l r— 1 Kj"
moo
• • •
o o o
>^
ro
t_
CL
0 CO 0
§ I §
Z 0 Z
"(0
•^
LO in
** CN
CN
^— x
vO
O
{- r-l
rj) —
E 0
rO -P
CO _c ro
C O U.
"+> +J ^
rO r— CO
CJ 3 C
O 2 i-
t_ i—
CL in 0
j£ *r~ 3 >
CO O O rO
20 31.
S- C -P
J= 1 i-l
O t- ~ -P t-
•P O vO CO
ro -P 0 O -P
CD (O •— 1 O -- 1_
» 0 * 0
r— C 0 0 C
r— -i— -P CO i—
(O O (O t_ U
ECS- ro C
CO 1— 1 CO _l 1— 1
_Q
^f
r— 1
0
•
o
0.
co
LU
•s^
i.
0
_Q
f^
3
t_
o
in
-p
0
vO
r-l
CN
L
o
-p
rO
£_
0
C ^
•r- VO
O 0
C r-l
HH s-^
in 0
3 -P
o *o
3 U
C O)
*^-
-P S-
C ro
O .0
O
CO
» C
0 f-
co JK:
rO O
_J t-
, E -c o
+J ro -P r-l
I— *r- ^-*
t- O U S
OS -»
-P -P 0 t- CO
(O i— 0 C
J_ M- COJ3 i-
0 O C E S-
C •*— rO *^
•r- C W .C >t-
O ro O
C 0 ro "i— in
f-H E -P ro
0 r- CO
r- U C 3
i- El S-
U -P 1 ra «J
S- 0 r—
0 £ in I- r—
§5 -fJ 0 T-
0 T- -C X
CO C -P 3
O «-* 3 O ^
•> L.
0 0
c a.
0
^> *
a. 0
c
«. 0
0 t-
C j=
0 -P
C- C
>% ro
CL C
^ o
lO f"
+^s C^
o
N •>
C 0
0 C
CQ 0
•p •
" C TJ
0) rO CD
r— L. -P
CL O U
£ 3 0
rO r- -P
in >*- 0
C »
•r- 0 -P
C O
•o 0 c
0 r—
3 t- 3
in 0 .a
ra CL
0 '-- i.
e T- o
•> 4-
in -c
C CO 0
3 — • N
o o >>
CL N r—
JC ro
0 C
-Q ra
ra
•rj
C
(0
CD
0
^^
c
0
CL
•i
in
0
c
0
JC
•p
c
rO
L.
O
p.
4-
O
N
C
0
A
0
c
0
in
t
r~
O
^
0
c
0
o
rO
£_
JZ
•P
C •
ra 0
*-* c
ro 0
N >•
C CL
0 ^
^
in u
0 rn
^3 •*
3 CN
r— •*
U r-l
C *~-
i- O
C
O 0
in T3
r— C
< -r-
•
t.
O
•p
o
rO
4-
c
0
in
U)
•r—
E
0
c
M—
•o
0
•rj
3
r—
U
c
i—
in
•a
c
3
O
CL
O
2E!
£
c
o
c
o
'-p
rO
O
M-
c
•r-
O
z
u
5-82
-------
RADIAN
CORPORATION
data for MSW incinerators. This assumption was based on a reported average
commercial incinerator capacity utilization of 71 percent (95). The PAH
emission factor presented in Section 5.8.2 for large MSW incinerators
controlled with the scrubber/ESP system was selected as representative of the
current population. The size of the unit tested was similar to the average
size MSW incinerator. The emission factor reported for PCB was not included
because of the relatively small size of the incinerator tested (0.11 metric
tons/day). Thus, the PAH emission factor may result in a low estimate of
total POM. National PAH estimates for MSW incinerators were calculated as:
(365 days) [385 metric tons incinerated] /O. 7 capacity f actoA /104 unitsNx
yr day-unit
0.014 om PAH] metric ton
metric ton \106 gm
= 0.14 metric tons/yr (0.16 tons/yr)
A similar method was used to calculate PAH emissions from commercial and
industrial incinerators:
p65 davs| /0.71 capacity factoA 12.5 metric tons] /100,000 units\ |6.8 gm PAHjx
I yr J \ I \~ unit day I \ I I metric ton/
(metric ton] = 440 metric tons/yr (485 tons/yr)
Average capacity, capacity utilization, and the number of units were obtained
from Reference 95. Industrial incinerators were included with commercial
incinerators in the available data on capacity and number of units. In lieu
of other data being available, the emission factor for commercial units was
assumed to apply to industrial incinerators.
It should be emphasized that these emission estimates are highly uncer-
tain. They are based on a very limited data; they may not reflect total POM
emissions (i.e., polychlorinated biphenyls); and they do not reflect the
5-83
-------
RADIAN
actual composition and amounts of waste combusted. Also* no data were avail-
able to develop accurate POM emission estimates for MSW-fired boilers, indus-
trial incinerators or industrial boilers burning solid waste, hazardous waste-
s, or RDF. Of these sources, industrial incinerators may potentially be the
most significant because of their relatively lower combustion efficiencies and
lack of controls.
5.8.4 Trends Influencing POM Emissions
Municipal Waste Combustion—The trend is toward larger, more efficient
MSW incinerators due to air emission regulations (90). These larger units
will generally be equipped with ESPs. One reference reports that the trend in
emission control selection for MSW incinerators has been forced away from
scrubbers because of the failure of some earlier poorly designed impingement
scrubbers (108).
Reported growth projections for MSW incineration are that about 5 new
units will be built by 1985, each at about 2270 metric tons/day (2,400 tons/
day) capacity. Six large MSW-fired boilers are expected to be built by 1990
(93). These new units will likely be controlled with ESPs or wet scrubbers.
Industrial Waste Combustion—Industrial incineration is expected to
decline, especially for use in volume reduction purposes (109). This trend is
reportedly due to land disposal techniques becoming more economical as a
result of more stringent air emission regulations (109). Exceptions to the
decline in industrial incinerators include (109):
o incineration with heat recovery (especially in the lumber,
pulp, and paper, and other industries with combustible
wastes), and
o resource recovery from copper wire and electric motor
incineration.
5-84
-------
One study has suggested that the complications of RCRA rules and the
associated recordkeeping requirements may lead to increased (and unregulated)
combustion of hazardous wastes in existing boilers (89). This could be a
significant source of POM emissions depending on the types of waste
incinerated, existing emission controls, and regulatory development related to
RCRA.
Six new RDF-fired units with a total heat input capacity of about 380 MW
are expected to be build by 1990 (93). The co-firing of RDF in existing coal-
fired boilers may increase due to general incentives to reduce fuel-related
costs.
Commercial Waste Combustion—No significant changes in capacity or
types of wastes incinerated are expected in this category.
5.9 CARBON BLACK PRODUCTION
5.9.1 Source Category Description
5.9.1.1 Process Description
Carbon black is produced by pyrolysis of an atomized liquid hydrocarbon
mixture. One of its main uses is as a reinforcing agent in rubber tires.
About 90 percent of U.S. carbon black is produced by the oil-furnace process.
Temperatures in the refractory-lined steel furnaces vary between 1320 and
1540°C. The heat for the decomposition reaction is supplied by combustion of
natural gas. The pyrolysis reaction is a source of POM emissions (110).
Hot furnace gas, containing carbon black particles, is cooled to about
230°C and then passed through a fabric filter for recovery of the carbon black
(110). Exhaust gas from the fabric filter is vented to the atmosphere or sent
to an emission control device. The main process vent is the major source of
POM emissions in a carbon black plant. Emissions may depend to some degree
5-db
-------
on the type of carbon black produced due to required differences in furnace
operation as a function of product specifications.
5.9.1.2 Emission Controls/Regulations
About two-thirds of U.S. carbon black plants use a combustion device
(flare, CO boiler, or thermal incinerator) on the main process vent (110). No
quantitative data were available on the effect of these systems on POM emis-
sions. Although the use of a combustion device might be expected to reduce
POM emissions, "high temperatures" (in the combustion devices) could report-
edly lead to formation of additional POM (110). No temperatures were speci-
fied in the reference reviewed.
However, in addition to the combustion control device, most carbon black
furnaces are equipped with fabrip filters for product recovery. Thus, POM
condensed on the captured carbon black particles will be removed from the
exhaust gas stream.
Existing carbon black plants are generally subject to SIP particulate
emission limits. New plants would be covered by prevention of significant
deterioration (PSD) requirements, but no NSPS exists for this source category.
5.9.1.3 Source Locations
Carbon black plants are located in Louisiana, California, West Virginia,
Arkansas, Oklahoma, and New Jersey (111). As of 1977, about 60 percent of the
U.S. production capacity was located in Louisiana.
5.9.2 Emission Factors
Total uncontrolled POM emissions from an oil-furnace carbon black plant
were measured in a previous study using a modified Method 5 sampling train
(110). Samples were extracted, separated by liquid chromatography, and
5-86
-------
analyzed by GC/MS. A total POM emission factor of 1.9 gm/metric ton (3.8 x
10~3 ib/ton) of carbon black produced was reported. Acenaphthylene accounted
for about 46 percent by weight of the POM in the samples collected. Table
5-22 lists the POM compounds detected in the collected samples.
5.9.3 National Emission Estimates
Reference 110 puts the U.S. carbon black production capacity at 1.72 x
10^ metric tons (1.9 x 10^ tons/yr). In 1980, the value of carbon black ship-
ments was $540 million (112). At about $0.50/bulk kg of carbon black (113),
this translates into approximately 1.08 x 10^ metric tons/yr (1.19 x 10^
tons/yr), or 57 percent of capacity. Thus, national POM emisisons from carbon
black production can be estimated as:
(1.08 x 10 metric tons carbon black | f 1.9 gm POM j /metric tonj
yr J Uetric ton produced/ ^ 106 gm J
= 2.05 metric tons/yr (2.26 tons/yr)
This estimate may be somewhat high since it is based on an uncontrolled POM
emission factor (110) and, as stated above, about two-thirds of the plants
currently have combustion devices for emission control. However, no data were
available to quantify the effect of such combustion-type controls on POM emis-
sions.
5.9.4 Trends Influencing POM Emissions
A two to three percent annual growth rate has been projected for the car-
bon black industry (114). Some of this growth appears possible without con-
struction of new facilities as indicated by the 57 percent capacity utiliza-
tion figure presented above. New plants will likely be subject to PSD permit-
ting requirements that require some form of emission control.
5-87
-------
RADIAN
TABLE 5-22. POM COMPOUNDS IN SAMPLES FROM OIL-FURNACE
CARBON BLACK PLANT (110)
Acenaphthylene (46% of total)
Anth racene/phenanth rene
Benzof1uoranthenes
Benzo(g,h»1)fluoranthene
Benzo(g»h,1)perylene/anthanthrene
Di benzanth i ocenes
D1benzo(c,g)carbazole
Dibenzopyrenes
Dibenzothlophene
Dimethyl anthracenes/phenanthrenesa
7,12 Dimethylbenz(a)anthracene
Fluoranthene
Indeno(l,2,3-c,d)pyrene
Methylanthracenes/phenanthrenes3
Methy1cholanth rene
Methylfluoranthene/pyrene
Pyrene (26% of total)
aTogether these groups accounted for 12% of total.
5-88
-------
RADIAN
5.10 WOOD CHARCOAL PRODUCTION
5.10.1 Source Category Description
5.10.1.1 Process Description*
Charcoal* primarily used for outdoor cooking* is manufactured by the
pyrolysis of carbonaceous raw materials. The raw materials used are primarily
medium to dense hardwoods. Softwoods* sawmill residue* nutshells* fruit pits
and vegetable wastes can also be used. (This study was limited to examining
charcoal production from wood.) There are two major types of techniques used
in wood charcoal manufacture:
o Missouri-type batch kiln, and
o Continuous Herreshoff furnace.
The batch process accounts for about 45 percent of the national produc-
tion (a small portion of this capacity is accounted for by older beehive
kilns). Continuous furnaces, which generally have a much larger capacity than
batch kilns, account for the remainder for production capacity. The kiln or
furnace products are sold directly or made into briquettes.
A Missouri-type batch kiln normally processes about 45 to 50 cords of
wood in a 10- to 25-day cycle. Kiln temperatures are in the 450 to 510°C
range and pyrolysis of the wood products is a source of POM emissions.
Missouri-type kilns have exhaust stacks along the side walls. The required
burn time and resulting emissions from the batch kiln vary with kiln capacity,
operational practices* wood species, and wood moisture content.
*The information presented in this subsection is summarized from References
115 and 116.
5-89
-------
RADIAN
Continuous charcoal production is accomplished in Herreshoff multiple
hearth furnaces. Furnace temperatures range between 480° and 650"C. The
off-gases exit through the stacks located on top of the furnace or are used as
a heat source (e.g. predrying of feed, drying of briquettes produced at an
adjacent location, or combustion in a waste heat boiler to produce steam.) Use
of continuous Herreshoff furnaces requires a large and steady source of raw
materials. The typical continuous furnace capacity is 2.5 metric tons/hr
(2.75 tons/hr).
5.10.1.2 Emission Controls/Regulations
Emission control of batch kilns is complicated by the cyclic nature of
the process. Direct-fired afterburners are probably the most feasible control
system, but auxiliary fuel use is required with these devices. If fuels other
than natural gas are used as the auxiliary fuel there is a potential for
additional POM emissions. No data were located on the percentage of batch
kilns equipped with afterburners. No data were located on the effect of
afterburners on POM. Natural gas would likely have to be used as the auxil-
iary fuel to avoid additional POM emissions.
Many of the batch kilns are older and relatively small batch kilns are
likely to be uncontrolled (115,116).
Continuous furnaces can also be controlled by direct-fired afterburners.
Auxilliary fuel firing is required only during start-up or process upsets on
continuous units because of the higher heating value of the exhaust gases. An
incinerator is used to control emissions in at least one continuous furnace
plant, but afterburners are reportedly used on most continuous furnaces (117).
There is no NSPS for charcoal manufacturing, although the feasibility of
developing a standard was investigated (115). Charcoal manufacturing facili-
ties are generally subject to SIP particulate emission limits and opacity
regulations. Some plants may be subject to CO standards.
5-90
-------
COKMMUmOM
5.10.1.3 Source Locations
Charcoal manufacturing facilities are located primarily in Missouri,
Arkansas, and other Southeastern states (118). Missouri accounts for a large
share of the national production, although about 24 states have a wood
charcoal industry.
5.10.2 Emission Factors
POM emission factor data were available only for a Missouri-type batch
kiln (119). An average of five sampling runs showed total POM uncontrolled
emissions of 0.35 gm/hr (770 x 10~6 Ib/hr). Based on limited production
da-ta provided in Reference 119, this translates into about 3.5 gm/metric ton
of charcoal produced (0.007 Ib/ton). The authors report that the samples,
collected with a modified Method 5 apparatus and analyzed by gas chromato-
graph, contained benz(c)phenanthrene, benzo(a)pyrene, and "POM-like" material.
Four POM compounds were analyzed for, but not detected: dibenz(a,h)anthra-
cene, 3-dimethylcholanthrene, 7,12 dimethylbenz(a)anthracene, and 3,4,5,6-
dibenzocarbazole.
It should be noted that the emission tests were considered of question-
able value due to the difficulty of sampling the kiln and "the improvisational
sampling techniques" used. No estimate of the accuracy was provided. Also,
no emission data were reported for the initial ignition process during which
gas can escape from the kiln.
5.10.3 National Emission Estimates
The estimated 1978 charcoal production capacity was 376,340 metric tons
(414,000 tons) distributed as follows (120).
5-91
-------
Process Production % Total
Missouri-type 198,180 52.7
batch kiln
Beehive kiln 2,730 0.73
Continuing Herreshoff 175,460 46.6
furnace
376,370
Charcoal production in 1980 was assumed to be about the same since there has
been no indication of growth in the industry (121).
Based on these data, national POM emissions from Missouri-type kilns were
calculated as:
ii\ (Lj
J \rnei
r!98,180 metric tons charcoal] [3.5 gm POM] [metric ton^
I yr / Imetric tony V 106 gmJ
= 0.69 metric ton/yr (0.76 ton/yr)
Although an emission factor was not available for continuous furnace
production, is is expected that POM emissions will be substantially less (on a
mass per mass of charcoal produced basis) than for batch kilns. This is
because of the capability of the continuous furnace to use non-direct fired
afterburners or incinerators or to recover the heating value of the exhaust
gases through other combustion devices. These techniques should reduce POM
emissions. Thus, at a maximum, total POM emissions from this category should
be:
198,180 + 175,460 metric tons charcoal] [3.5 am POM] /metric ton]
yr \metr1c ton 106 gm
1\ [3.5 am POM] /m
J ynetric tony V
=1.3 metric tons/yr (1.4 tons/yr)
5-92
-------
RADIAN
This is a "worst-case" estimate since it is based on the higher emission
factor of the Missouri-kiln.
5.10.4 Trends Influencing POM Emissions
There is a general trend in the industry toward fewer but larger plants
that are more amenable to emission control than many of the existing small
plants. In addition, small plants have been forced to close in some states
(Florida* Illinois, Ohio, and Oklahoma) because of problems in complying with
state emission regulations (121). Also, the level of charcoal production has
remained relatively stable in recent years. There are no indications of
future growth in the industry, as briquettes are increasingly being manufac-
tured from other materials such as lignite. Thus, POM emissions from this
category are not expected to increase in the future.
5.11 VEHICLE DISPOSAL
5.11.1 Source Category Description
5.11.1.1 Process Description
Three types of techniques are used to remove the organic material from
auto bodies before they are used as scrap by the steel industry (usually in
electric arc furnaces). The organic material can be removed by (122):
o open burning of whole auto bodies
o incineration of whole auto bodies, or
o shredding of auto bodies and incineration of the shredded
steel.
Incineration of shredded steel is generally accomplished in rotary kilns.
The combustion efficiency of the kilns is relatively high, reducing the
potential for POM emissions (122). Open burning has a considerably higher
5-93
-------
RADIAN
CORPORATION
potential for POM emissions than incineration because of the low combustion
efficiency typical of open burning.
5.11.1.2 Emission Controls/Regulations
Open burning is generally prohibited especially in most metropolitan
areas, but may still occur either illegally or outside restricted areas.
Incineration of whole auto bodies would likely be subject to SIP particu-
late matter emission limits applicable to incinerators. Mechanical col-
lectors, wet scrubbers, and ESPs are candidate control technologies for incin-
erators. No specific information was located with regard to the application
of controls to incineration of auto bodies.
5.11.1.3 Source Locations
No source location data were readily available. However, incineration of
auto bodies would be expected to coincide with steel industry locations.
5.11.2 Emission Factor Data
No POM emission factors for vehicle disposal were located.
5.11.3 National Emission Estimates
No emission factor or consumption data were located.
5.11.4 Trends Influencing POM Emissions
The demand for auto scrap by the steel industry is increasing due to an
increase in the use of electric arc furnaces. These furnaces can accomodate a
high proportion of scrap in the furnace feed (122). Concurrently, the in-
crease in demand has made shredding and subsequent incineration of auto body
steel in rotary kilns more economically attractive (122). The trend toward
5-94
-------
RADIAN
this higher combustion efficiency technique* combined with continued enforce-
ment of open burning prohibitions, may result in steady or decreased POM
emissions even, as the demand for auto body scrap is increased.
5.12 MOBILE SOURCES
This section discusses the following mobile source categories with
potential for POM emissions:
o gasoline-fueled autos and trucks,
o diesel autos and trucks, and
o rubber tire wear.
5.12.1 Source Category Description
5.12.1.1 Process Descriptions
Gasoline Autos and Trucks—Emissions of POM from combustion of gasoline
in autos and light-duty trucks are dependent on several factors, including
(123):
o inefficient combustion because of less than stoichiometric
air to fuel ratios,
o driver operation techniques that lower fuel efficiency,
o engine deterioration and combustion chamber deposits,
o aromatic content of the gasoline and presence of lead
additives.
The pyrolysis of motor oil deposits built up on the engine may also be a
potential source of POM emissions (123).
Diesel Autos and Trucks—The primary causes of POM emissions from
diesel combustion in autos and trucks are overloading and poor engine main-
tenance (124). However, even under normal operating conditions, diesel autos
b-95
-------
RADIAN
CORPORATION
tend to emit more POM than their gasoline-fueled counterparts. The higher
emissions of POM may be due to the relatively lower combustion-chamber temper-
atures typical of diesel engines (124). Diesel autos emit 30 to 100 times
more particulate matter per mile than gasoline autos equipped with catalytic
converters. Diesel particulates are usually less than 1 pro in diameter and
consist of a carbonaceous core. Many different soluble organic compounds are
absorbed on these particles, some of which are POM (125).
Rubber Tire Wear—Carbon black and other tire materials possibly con-
taining POM are released to the atmosphere through oxidation and normal wear
of rubber tires (124).
5.12.1.2 Emission Controls/Regulations
Gasoline Autos and Trucks—The control devices used on recent model
autos have the capability to reduce POM emissions. Exhaust gas recirculation,
which was introduced in 1968, has been reported to reduce PAH emissions by 65
to 80 percent over uncontrolled (123). Catalytic converters, which generally
are used on post-1975 models, have been reported to reduce PAH emissions by up
to 99 percent (123).
Federal and state mobile source emission regulations generally require
the use of unleaded gasoline and catalytic converters for recent year models.
The regulations also limit emissions of NOX, CO, and hydrocarbons (HO and
evaporative losses. Some states have implemented regular inspection and
maintenance (I/M) programs for autos.
Diesel Autos and Trucks—Exhaust controls are not commonly used on
diesel-fueled vehicles. However, proper loading, fueling, and maintenance of
diesel engines can reduce their POM emissions (124).
Federal and California state regulations limit HC, NOX, CO, and evapora-
tive losses from diesel autos and light duty trucks. States have the option
5-96
-------
of implementing either the Federal or California standards, but are restricted
from implementing more stringent standards for diesel-fired vehicles (126).
Rubber Tire Wear—Emission controls or regulations are not applicable
to this source.
5.12.1.3 Source Locations
Source locations data is not applicable to mobile sources, although thei.r
concentrations are obviously higher in urban areas.
5.12.2 Emission Factors
Reference 131 provided emission factors for several PAH compounds emitted
from gasoline and diesel vehicles (Table 5-23). These values were based on
measured emission data and "derived emission factors." The emission factors
were then multiplied by estimated 1979 fuel consumption data for the various
types of mobile sources to develop the total estimates presented in the next
section. The PAH compounds included are listed in Table 5-24.
The authors note that the derived emission factors are quite uncertain
and may vary by a factor of two or more. The two major assumptions used in
developing the emission factors in Table 5-23 were (131):
o that the "PAH distributions" for both gasoline and diesel
vehicles are the same as the distribution for average
light-duty gasoline vehicles not equipped with catalytic
converters, and
o that all PAH emissions are reduced by catalytic converters
as much as BaP emissions are reduced by such devices.
5.12.3 National Emission Estimates
Table 5-25 shows the 1979 estimated PAH emissions from several mobile
source categories including gasoline and diesel autos and trucks, buses,
5-97
-------
^-%
r-l
m
^1
CO
LU
_l
O
I— I
1 1 1
LLl
>•
_J
LU
CO
LU
1— i
Q
•z.
^C
LU
•z.
HH
_l
O
CO
<
o
cc
0
LU
CO
o:
o
l-
0
<
LU
O
i— i
CO
CO
1— 1
LU
5
Q_
•
CM1
1
IT)
LU
_l
3
H-
ra La
L. i-
O E
-P -^
O C7)
LL° ^
LU
C
o
•w-
in
U)
^ "rO
LU O)
^.
X 0)
^ 3.
<£|
0
Q.
>,
1-
f__
0
3
Lu
0
a.
>,
&
0
"u
•i—
jr
0
>
un
8
,
-P
3
•a
-P
jC
^
0
-P
(_
0
>
C
o
U
U
•r—
4->
>>
ra
O)-P
•r~
_J
ra
U
CO
.— I
^t
r-
m
"S-
r-
0
c
•f—
f—
O
in
ra
CD
•o
0
•a
ra
0
c
Z)
c
o
tf_
•P
,
-P
3 +>
T3 0)
^^
JI ro
CD-P
•^ ro
—1 U
in
f
r-l
CO
co
0
c
•r-
r^*
O
in
ra
CD
T3
0
TJ
ra
0
c
r?
>,
ra
i
0
L.
.c
-P
«
>,
-P
3 -P
•o in
I >-,
•P r—
x: ro
CO-P
i- ro
_l O
CM
LO
O
CJ»
CM
TJ-
CM
LO
•^~
0
C
•r—
^_
O
in
ro
CD
T3
0
•o
ro
0
_l
O
C
*
>,
•P
3
T3
>,
>
ra
0
X
L.
CD
4->
L.
CD
>
C
o
o
U
•^
-P
>>
ra
•P
ra
o
LO
CM
CTv
1— 1
CTi
tO
CM
CO
•^~
^_
0
3
LU
^_
0
in
0
•^
Q
X,
-P
3
T3
i
.c
U)
•^
_i
CO
t
c*
to
CM
CO
•«t
^_
0
3
LU
^_
0
in
0
•r-
Q
>,
-P
3
•a
>^
>
ro
0
X
•
-o
CD
•o
3
r—
U
C.
•1—
in
•o
c
3
O
CL
U
z
2
CD
.C
-P
in
•P
in
•r—
r—
"d-
CM
1
IT)
CD
_Q
ra
h-
ro
x-x
lo
0)
*•».
•r™
E
•1 •
r—
(O
O)
-•»
O)
3.
^^
11
•r-
E
^
O)
3.
-Q
5-98
-------
TABLE 5-24. POM COMPOUNDS INCLUDED IN DERIVED MOBILE SOURCE
EMISSION FACTORS (131)
Anthracene
Phenanthrene
Methylphenanthrene
Dimethylfl uorene
Dimethylphenanthrene
Fluoranthene
Pyrene
Benzof1uorene
Benzoanthracene
Trlphenylene
Cyclopentapyrene
Chrysene
Indenofluoranthene
Idenopyrene
Methylchrysene
1-Nitropyrene
Benzof1uoranthene
Benzo(e)pyrene
Benzo(a)pyrene
Perylene
Cyclopentabenzopyrene
Benzochrysene
Anthanthrene
Di benzanthracene
Benzoperylene
Coronene
Cyclopentabenzoperylene
5-99
-------
TABLE 5-25. NATIONAL 1979 PAH EMISSIONS FROM MOBILE SOURCES3'b (131)
(metric tons)
Anthracene
Phenanthrene
Methylphenanthrene
Dimethylfluorene
Dimethylphenanthrene
Fluoranthene
Pyrene
Benzof1uorene
Benzoanthracene
Triphenylene
Cyclopentapyrene
Chrysene
Indenofluoranthene
Indenopyrene
Methylchrysene
1-Nitropyrene
Benzofl uoranthene
Benzo(e)pyrene
Benzo(a)pyrene
Perylene
Cyclopentabenzopyrene
Benzochrysene
Anthanthrene
Dibenzanthracene
Benzoperylene
Coronene
Cyclopentabenzoperylene
350
1,400
900
470
320
750
950
120
37
30
390
150
19
30
6
17
110
52
43
4
26
1
22
13
110
80
18
6,400
aTotal for all PAHs: about (3)(6,400) = about 19,000 metric tons.
blncludes autos, trucks, motorcycles, railroads, aircraft, ships, and farm
and military mobile sources.
5-100
-------
RADIAN
CORPORATION
motorcycles, railroads, ships, aircraft, and farm and military mobile sources.
The emission factors for heavy-duty gasoline and diesel vehicles were used to
estimate emissions for the categories other than gasoline and diesel autos and
trucks.
The largest portion of PAH emissions from the various mobile source
categories is attributable to older gasoline autos not equipped with catalytic
converters (131). However, emissions from this category are expected to
decrease as these older vehicles are taken out of service.
5.12.4 Trends Influencing POM Emissions
Gasoline Autos and Trucks—Consumption of gasoline in autos and trucks
is expected to decrease in future years due primarily to the use of more
fuel-efficient vehicles, conservation of fuel induced by increasing prices,
and the increased use of diesel autos. Between 1978 and 1982, gasoline demand
dropped by slightly less than 12 percent (128). Combined with continued
emission control requirements and I/M programs, the trend of decreasing
gasoline consumption may result in decreased POM emissions from this category.
Diesel Autos and Trucks—A wide range of estimates exist regarding the
projected increase of diesel fuel in autos. However, it is likely that at
least 20 percent of light duty vehicles will be diesel-fueled by 1995 (129).
Total distillate fuel demand is projected to increase at least 2 percent
between 1982 and 1983, with the majority of the increase attributed to diesel
fuel consumption in autos and trucks (128).
This trend could result in increased POM emissions from diesel-fueled
vehicles if emission control technologies continue not to be used. Potential
control techniques for diesel engines include turbocharging and modification
of the combustion chamber and fuel injection system (130). These modifica-
tions should improve combustion efficiencies and in turn reduce the POM
emitted per unit of fuel consumed.
5-101
-------
Rubber Tire Wear—No significant changes are foreseen in POM emissions
from rubber tire wear other than increases associated with total miles
traveled by the vehicle-operating population.
5-102
-------
RADIAN
5.13 REFERENCES
1. Energy and Environmental Analysis, Inc. Preliminary Assessment of the
Sources, Control and Population Exposure to Airborne Polvcvcllc Organic
Matter (POM) as Indicated bv Benzo (A) Pvrene (BaP). Final Report,
Prepared for U.S. Environmental Protection Agency, Research Triangle
Park, North Carolina, p. 95, November 10, 1978.
2. Chalekode, P.K. and T.R. Blackwood. Source Assessment; Coal Refuse
Piles, Abandoned Mines, and Outcrops - State of the Art. EPA-600/
2-78-004v, U.S. Environmental Protection Agency, Cincinnati, OH, July
1978.
3. Reference 2, p. 2-11.
4. Formica, Peter N. (TRC). Controlled and Uncontrolled Emission Rates and
Applicable Limitations for 80 Processes. EPA-340/1-78-004, U.S. Envi-
ronmental Protection Agency, Research Triangle Park, NC, pp. II-l to
II-2, April 1978.
5. Reference 1, p. 99-100.
6. Reference 2, Section 5.
7. Evans, John S. and Douglas W. Cooper. "An Inventory of Particulate Emis-
sions from Open Sources," Journal of the Air Pollution Control Associa-
tion, 30(12), pp. 1298-1303, December 1980.
8. Reference 1, p. 98-99.
9. Shih, Chris, et al. (TRW, Inc.). POM Emissions from Stationary Conven-
tional Combustion Processes. CCEA Issue Paper, U.S. Environmental
Protection Agency, Research Triangle Park, NC, p. 15, January 1980.
5-103
-------
10. Reference 9, p. 3.
11. Reference 1, pp. 43-44.
12. Baig, S., et al (TRW) and Hurley, T. et al. (Radian Corporation) Con-
ventional Combustion Environmental Assessment Final Report. EPA Con-
tract No. 68-02-3138, U.S. Environmental Protection Agency, Research
Triangle Park, NC, p. 4-8, July 1981.
13. Reference 1, p. 22.
14. U.S. Department of Energy. Electric Power Annual. DOE/EIA-0348(81),
Washington D.C., p. 13, November 1982.
15. U.S. Department of Energy. State Energy Data Report; 1960 through
1980. DOE/EIA-0214(80), Washington, DC, p. 9, July 1982.
16. Emission Standards and Engineering Division. (U.S. Environmental Pro-
tection Agency.) Fossil Fuel-Fired Industrial Boilers-Background Infor-
mation Volume 1; Chapters 1-9. EPA-450/3-82-006a, p. 3-2, March 1982.
17. Reference 15, p. 7. Total adjusted to reflect only the fuel consumed by
boilers. Adjustment factors (1974 data) were obtained from: Energy and
Environmental Analysis, Inc. Fossil Fuel-Fired Industrial Boilers -
Background Information for Proposed Standards, Appendices F and G. Pre-
pared for U.S. Environmental Protection Agency, Research Triangle Park,
NC, p.F-12, March 9, 1981. Adjustment factors (% of total energy con-
sumption used in boilers): Coal-30%; Oil-67%; and Gas-'
18. Reference 15, p 6. All consumption reported assumed to be in boilers and
furnaces.
19. Reference 1, p. 55.
5-104
-------
20. Reference 1, p. 51.
21. Reference 5, p. 5.
22. Reference 12, p. 5-57, 5-52.
23. Laseke, B.A., M.T. Mella, and N.G. Bruck. (Red Co. Environmental Inc.)
"Trends in Commercial Application of FGD Technology." Presented at the
EPA/EPRI Seventh Symposium on Flue Gas Desulfurization, Hollywood,
Florida, May 17-20, 1982, p. 2-1. The % capacity figures shown in the
table are based on 275,000 MW installed coal-fired capacity as reported
in: U.S. Department of Energy. Inventory of Power Plants in the United
States - 1981 Annual. DOE/EIA-0095(81), Washington DC, p. 19, September
1982. (Assumes all FGD capacity reported applied to coal units since no
other data were readily available).
24. Kelly, M.E. and M.A. Palazollo. Status of Dry SCb Control Systems—Fall
1982. Final Report, EPA Contract No. 68-02-3171, Task 58, U.S. Envi-
ronmental Protection Agency, Research Triangle Park, NC, p. 19, January
1982.
25. Reference 2, p. 5-58.
26. Code of Federal Regulations. 40 CFR 60, Part 30, Subparts D and Da, pp.
234-261, Revised July 1, 1982.
27. Reference 16, pp. 4-66 to 4-81.
28. Reference 24, p. 23.
29. Reference 16, Chapter 3.
30. Reference 16, p. 4-16.
-------
RADIAN
CORPORATION
31. Reference 16, p. 4-39.
32. Foster, Kathye L., et al. "Permitting a Wood-Burning Boiler in a Major
Metropolitan Area," Journal of the Air Pollution Control Assoc. 32(8),
pp. 872-877, August, 1982.
33. Jones, Larry (EPA-OAQPS). Personal communication with M.E. Kelly, Radian
Corporation, regarding regulatory status of commercial/institutional
boilers. April 21, 1983.
34. U.S. Department of Energy. Inventory of Power Plants in the United
States-1981 Annual. DOE/EIA-0095(81), Washington, D.C., September 1982.
35. Bergman, Michael K. and Robert M. Dykes. (U.S. Environmental Protection
Agency). Prospects for Increasing the Direct Use of Coal in Industrial
Boilers. Draft Report, Research Triangle Park, NC, p. 68-80.
36. DeAngelis, D.G. et al. (Monsanto Research Corporation). Source Assess-
ment: Residential Combustion of Wood. EPA-600/2-80-024b, U.S. Envi-
ronmental Protection Agency, Research Triangle Park, NC, p. 21, March
1980.
37. Energy Information Administration (U.S. Department of Energy). Esti-
mates of U.S. Wood Energy Consumption from 1949 to 1981, DOE/EIA-0341,
August 1982.
38. Piper, B.F., S. Hersh, and W. Nazimowitz (KVB, Inc). Combustion Demon-
stration of SRC II Fuel Oil in a Tangentially Fired Boiler. EPRI Final
Report, FP-1029, Palo Alto, California, pp. 7-22 to 7-23, May 1979.
39. Wainwright, Phyliss B. et al. (NC Dept. of Natural Resources). A POM
Emissions Study for Industrial Wood-Fired Boilers. Department of
Natural Resources, Raleigh, NC, April 1982.
5-lOb
-------
RADIAN
CORPORATION
40. Radian Corporation. Background Information Document for Non-Fossil Fuel-
F1red Boilers. Draft Report (EPA-450/3-82-007), EPA Contract No. 68-02-
3058, U.S. Environmental Protection Agency, Research Triangle Park, NC,
Chapter 3 and p. 6-18, March 1982.
41. Eimutis, E.G. and R.P. Quill (Monsanto Research Corporation). Source
Assessment; Non-Criteria Pollutant Emissions. EPA-600/2-77-107e,
p. 78, July 1977.
42. Peters, J.A. and D.G. DeAngells (Monsanto Research Corporation). High
Altitude Testing of Residential Wood-Fired Combustion Equipment. Pre-
pared for U.S. Environmental Protection Agency, Research Triangle Park,
NC, September, 1981.
43. Reference 16, pp. 4-58 to 4-60.
44. Reference 35, Chapter 1.
45. Reference 40, p. 3-6.
46. Emission Standards and Engineering Division (U.S. Environmental Protec-
tion Agency). Coke Oven Emissions from Bv-Product Coke Oven Charging,
Door Leaks, and Topside Leaks on Wet Coal-Charged Batteries—Background
Information for Proposed Standards. Draft Report, U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina, March 1981.
47. Suta, Benjamin E. (SRI International). Human Population Exposure to
Coke Oven Atmospheric Emissions. EPA Contract No. 68-02-2835, Final
Report, U.S. Environmental Protection Agency, Research Triangle Park, NC,
Revised May 1979.
48. Reference 1, p. 68.
5-107
-------
RADIAN
49. Reference 46, pp. 3-60, 3-61.
50. Reference 46, pp. 3-54, 3-62.
51. Reference 46, p.7-7.
52. Emission Standards and Engineering Division (U.S. Environmental Protec-
tion Agency). Coke Wet Quenching—Background Information for Proposed
Standards. Draft Report, U.S. Environmental Protection Agency, Research
Triangle Park, NC, March 25, 1982.
53. Reference 46, pp. 3-1, 3-2.
54. Reference 52, p. 3-38.
55. Laube, A.M. and B.A. Drummond (York Research Corporation). Coke Quench
Tower Emission Testing Program. EPA-600/2-79-082, U.S. Environmental
Protection Agency, Research Triangle Park, NC, p. 7, April 1979.
56. Berry, Kent (U.S. Environmental Protection Agency). Personal communica-
tion with Mary Kelly, Radian Corporation, March 3, 1983 (through Ray
Morrison, U.S. Environmental Protection Agency).
57. Reference 46, p. 3-2.
58. Reference 46, p. 3-18.
59. Reference 1, p. 26.
»
60. Reference 46, pp. 3-41, 3-42.
61. Reference 55, p. 3-4.
62. Reference 1, p. 78-79.
5-108
-------
RADIAN
63. Emission Standards and Engineering Division (U.S. Environmental Protec-
tion Agency). A Review of Standards of Performance for New Stationary
Sources - Ferroalloy Production Facilities. EPA-450/3-80-041, U.S.
Environmental Protection Agency, Research Triangle Park, NC, December
1980.
64. Pan!, Dale (EPA-OAQPS). Personal communications with Mary E. Kelly,
Radian Corporation, regarding regulatory status of various iron and steel
source categories, April 13 and April 21, 1983.
65. Reference 63, p. 25-29.
66. Reference 63, pp. 47-50.
67. Industrial Environmental Research Laboratory (U.S. EPA). Level 1 Envi-
ronmental Assessment of Electric Submerged-Arc Furnaces Producing Ferro-
alloys. EPA-600/2-81-038, pp. 6-7, 15, March 1981.
68. Industry and Trade Administration (U.S. Department of Commerce). 1980
U.S. Industrial Outlook for 200 Industries with Projections for 1984.
U.S. Department of Commerce, Washington, DC, Chapter 16, January 1980.
69. Kahn, Z.S. and T.W. Hughes. (Monsanto Research Corporation) Source
Assessment! Asphalt Hot Mix. EPA-600/2-77-107n, U.S. Environmental
Protection Agency, Cincinnati, OH, p. 93, December 1977.
70. Gerstle, R.W. (PedCo Environmental, Inc.) Atmoshperic Emissions from
Asphalt Roofing Processes. EPA-600/2-74-101, U.S. Environmental Pro-
tection Agency, Washington, DC, Chapter 4, October 1974.
71. Reference 69, p 8.
72. Reference 69, pp 34-36.
5-109
-------
73. Reference 70, pp 19-20.
74. Reference 1, p 73.
75. NSPS for asphalt hot mix. 40CFR60, Subpart I. NSPS for asphalt
roofing products: 40CFR60, Subpart UU.
76. Reference 69, pp. 50-52.
77. Reference 70, p. 2 and pp. 43-45.
78. Energy Information Administration (U.S. Department of Energy). DOE
Energy Data Reports: "Sales of Asphalt in 1980". DOE/EIA-0112(80),
U.S. Department of Energy, Washington, DC, June 8, 1981.
79. Reference 1, pp. 72-78.
80. Reference 69, p. 7.
81. Reference 68, p. 7.
82. Reference 1, pp. 60-62.
83. Reference 26, Subpart J.
84. Pacific Environmental Services, Inc. Inspection Manual for Enforcement
of New Source Performance Standards; Catalytic Cracking Regenerators.
EPA-340/1-77-006, U.S. Environmental Protection Agency, Washington, DC,
pp. 2-1, 2-2, April 1977.
85. Reference 1, pp 25-26.
86. Oil and Gas Journal. Refinery Forecast Issue. March 30, 1981.
5-no
-------
RADIAN
87. Oil and Gas Journal. Forecast Review Issue, pp 132-133. January 2,
1982.
88. Reference 84, pp. 3-3.
89. Coyle, J.J. and E. J. Potenta (Fred C. Hart Associates, Inc.). "Popula-
tion Exposure to Hazardous Air Pollutants from Waste Combustion In Indus-
trial Boilers and RCRA Regulated Incinerators". Presented at the APCA
Specialty Conference on Measurement and Monitoring of Non-Criteria
(Toxic) Contaminants in Air, Chicago, IL, March 1983.
90. Reference 1, pp. 85-87.
91. Mcllvaine Company. U.S. Incinerator Air Pollution Control Equipment
Market. Mcllvaine Company, Northbrook, IL, p. 3, October 1981.
92. Reference 16, p. 6-17.
93. Reference 40, pp. 3-6 to 3-9.
94. Reference 91, p. 8.
95. Reference 1, pp. 87-89.
96. Reference 26, Subpart E.
97. Reference 91, p. 21.
98. Reference 40, p. 3-59.
99. Reference 91, pp. 18-19.
100. Estimated from MSW incinerator control level and corresponding efficiency
(Reference 97).
S5-111
-------
RADIAN
101. Reference 91, pp. 39-41.
102. Reference 40, p. 4-2.
103. Reference 1» Appendix F and p. 86.
104. Reference 40, p. 3-6.
105. Hangebrauck, P.P., D.J. vonLehmden, and J.E. Meeker. Sources of Poly-
nuclear Hydrocarbons in the Atmosphere. AP-33, NTIS PB 174-706. U.S.
Public Health Service, Washington, DC, 1967.
106. Sh1h, C. et al. (TRW, Inc.) POM Emissions from Stationary Combustion
Sources with Emphasis on Polvchlorinated Compounds of Dlbenzo-p-Dloxin,
Blphenvls, and Dibenzofuran. CCEA Issue Paper, EPA Contract No. 68-02-
3138, U.S. Environmental Protection Agency, Research Triangle Park, NC,
pp 17-21, January 1980.
107. Reference 1, p. 28.
108. Reference 91, p. 37.
109. Reference 91, pp. 52-53.
110. Serth, R.W. and T.W. Hughes. "Polycycllc Organic Matter (POM) and Trace
Elements Contents of Carbon Black Vent Gas", Environmental Science and
Technology, 14(3), pp. 298-300, March 1980.
111. Reference 1, Appendix E.
112. Reference 68, p. 112.
113. Chemical Marketing Reporter. December 1, 1980.
5-112
-------
114. Reference 1» p. 81.
115. Hulman» P.B. et al (Radian Corporation). Screening Study on Feasibility
of Standards of Performance for Wood Charcoal Manufacturing* Final
Report, EPA Contract No. 68-02-2608, Task 32. Prepared for U.S. Environ-
mental Protection Agency, Research Triangle Park, NC, August 1978.
116. Moscowitz, C.W. (Monsanto Research Corporation). Source Assessment;
Charcoal Manufacturing - State-of-the-Art. EPA-600/2-78-004z, U.S.
Environmental Protection Agency, Cincinnati, OH, December, 1978.
117. Reference 115, p. 42.
118. Reference 116, pp. 19-20.
119. Maxwell, W.H. (Midwest Research Institute). Stationary Source Testing
of a Missouri-Type Charcoal Kiln. NTIS PB-258695, U.S. Environmental
Protection Agency, Kansas City, MO, pp 10-17, August 3, 1976. (Average
of runs 2, 4, 6, 8, and 10).
120. Reference 115, p. 20.
121. Reference 115, pp. 21-22.
122. Reference 1, p. 96-97.
123. Reference 1, p. 103-107.
124. Reference 1, p. 108-113.
125. Klausmeier, Rob (Radian Corporation). A Study of the Impact of Light
Duty Vehicle Growth on Colorado's Air Quality, Draft Report, Prepared
for Colorado Department of Health, Radian Corporation, Austin, Texas,
pp. 2-5, November 3, 1981.
5-113
-------
RADIAN
126. Reference 125, p. 6-17.
127. Reference 1, p. 30.
128. Buck, Robert, "Demand, Imports to Rise in '83; Production to Slip." 011
and Gas Journal 81(4), pp. 71-103, January 31, 1983.
129. Reference 125, p. vi.
130. Reference 125, p. 111.
131. National Research Council. Polycycllc Aromatic Hydrocarbons: Evaluation
of Sources and Effects. National Academy Press. Washington, D.C.,
Chapter 1, 1983.
132. U.S. Environmental Protection Agency, Supplement 12 for Compilation of
A1r Pollutant Emission Factors, AP-42. NTIS PB 82-101213. Research
Triangle Park, NC, pp. 8.1-3 to 8.1-5, April 1981.
133. Meyer, Ron (U.S. Environmental Protection Agency). Personal communica-
tion with Ray Morrison, EPA Project Officer, August 29, 1983.
5-114
-------
APPENDIX A
ADDITIONAL CALCULATIONS FOR ESTIMATING POM EMISSIONS
FROM COMBUSTION SOURCES
-------
APPENDIX A
Additional Calculations for Estimating
POM Emissions
from
Combustion Sources
Section
A.I Conversion of Reported Emission Factors
to consistent units
/
A.2 Wood consumption in industrial boilers
A.3 Wood consumption in residential stoves
A.I Conversion of Reported Emission Factors
(Common unit basis [=*] Ib POM/1012 Btu heat input)
A.1.1 Coal Combustion
Emission factors were provided in terms of Ug or mg per kg of coal
burned (1). Assume a coal heating value of 11,500 Btu/lb, which is represen-
tative of several bituminous and subbituminous coals (2,3).
T«vM«.a. A9xlO"6gm POM\/ Ib coal W0.454 kgw Ib
boilers: ( ^-^j K11)500 Btu)l Ib
12
= 1.65 lb/10 Btu
A-l
-------
T j *. • -, u -i /41x10 °gm POM\/ lb coal W0.454 kg\/ lb \
Industrial boilers: I ; °-^ )(., cnn „—)( ,. 6)(~rn )
\ kg coal AH,500 Btu/\ lb /\454 gm'
- 3.57 lb/1012Btu
_ ,, „. , /67xlO"3gm POM\/ lb coal W.454
Residential: ( kg ^1 )(11>5QO Btu )(~lb
coal W.454 kg\/ lb
~
5826 lb/1012Btu
A. 1.2 Oil Combustion
Emission factors were provided in terms of Ug POM per liter of oil
burned (1). Assume an oil heating value (#6 Resid) of about 150,000 Btu/gal
(4).
Utility: [already in terms of Ib/Btu— (5)] - 2.0 lb/1012Btu
T A i /21xlO~6gm POMw liter \/ gal \/ lb \
industrial: ^ Uter Q±[ ^0>2642 gal/U50,000 Btu/V454 gm/
gal/U50,000 Btu/V454 gm
1.17 lb/1012Btu
-.-6.
o jj tJ i /120xlO "gm POM\/ liter \/ gal \/ lb \
Residential: ^ llter Q±1 A0.2642 gal A150,000 Btu>/U54 gm/
=6.67 lb/1012Btu
A.1.3 Natural Gas Combustion
Emission factors were provided in terms of yg or mg per m3 of gas
burned (1). Assume a natural gas heating value of 35,300 Btu/m3[1000 Btu/ft3]
(6).
A-2
-------
Industrial:
/11x10
-s.
W «'
A35,30
V m3 gas A35,300 Btu/\454 gm
0.69 lb/1012Btu
D -j «~, i /65xlO"6gm POM\/ m3 \/ lb \
Reszdential: { - 5 — ° - )( ,- ... )(TT/ - )
V m3gas A35,300 Btu/\454 gm/
= 4.06 lb/1012Btu
A.1.4 Wood Combustion
Emission factors provided in terms of yg or mg per kg of dry wood
burned (7). Assume a (dry) wood heating value of 8600 Btu/lb (8).
_ , _ . , /2.1xlO~3gm POMV 0.454 kg\/ lb
Industrial: ( kg WOQ^ - A lb §)(8600
lb
kg
237 lb/1012Btu
Residential /0.27 gm POMW0.454 kg\/ lb W lb \
Stoves: \ kg wood A lb A 8600 Btu/Us4 gm/
- 31,400 lb/1012Btu
Residential /0.029 gm POM\/.454 kgw lb W lb \
Fireplaces: V kg wood A lb A8600 Btu/\454 gm/
= 3370 lb/1012Btu
A-3
-------
A.2 Wood Consumption in Industrial Boilers
Reference 9 reports 1600 wood-fired industrial boilers with a total
capacity of 1.04x10** Btu/hr in operation in 1978. At an average capac-
ity factor of 60% (10), this translates into an annual wood consumption of
(^
This was the most recent consumption data for wood-fired boilers, and no
trend data (specific to boilers) were available to allow extrapolation to
1980. Therefore, 1978 consumption figures were used to estimate national
POM emissions from this category.
A.3 Wood Consumption in Residential Stoves and Fireplaces
Reference 8 puts annual wood use in stoves at 26 million metric
tons (dry basis). Using the estimated 8600 Btu/lb dry wood heating value
(8) yields:
Reference 8 reports 1980 wood consumption in fireplaces as 17.3
million metric tons:
A-4
-------
A.4 References
1. Energy and Environmental Analysis, Inc. Preliminary Assess-
ment of the Sources, Control, and Population Exposure Co Airborne Polycyclic
Organic Matter (POM) as Indicated by Benzo(a)pyrene. Final Report, prepared
for U.S. Environmental Protection Agency, Research Triangle Park, North
Carolina, Table III-l (intermediate factors for POM), November 10, 1978.
2. Emission Standards and Engineering Division (U.S. Environ-
mental Protection Agency). Fossil Fuel-Fired Industrial Boilers-Background
Information Volume 1: Chapters 1-9. EPA-450/3-82-006a, Chapter 3, March 1982.
3. Babcock and Wilcox. Steam; Its Generation and Use. 39th
edition, Babcock and Wilcox, New York, Chapter 5, 1978.
4. Reference 2, Table 6-5. Divide heat input rate to boiler by
fuel rate to calculate Btu/gal oil.
5. Piper, B.F., S. Hersh, and W. Nazimowitz (KVB, Inc.).
Combustion Demonstration of SRC II Fuel Oil in a Tangentially Fired Boiler.
EPRI Final Report, FP-1029, Palo Alto, California, pp. 7-22 to 7-23, May
1979.
6. Reference 2, Table 6-5. Divide heat input rate by fuel rate
to calculate Btu/m3 of gas.
7. Industrial: Wainwright, Phyliss B. et al. (NC Dept. of
Natural Resources). A POM Emission Study for Industrial Wood-Fired Boilers.
Department of Natural Resources, Raleigh, NC, April 1982.
Residential: DeAngelis, D.G. et al. (Monsanto Research
Corporation). Source Assessment; Residential Combustion of Wood. EPA-600/
2-80-024b, U.S. Environmental Protection Agency, Research Triangle Park, NC,
p. 21, March 1980.
A-5
-------
8. Energy Information Administration (U.S. Department of Energy).
Estimates of U.S. Wood Energy Consumption from 1949 to 1981. DOE/EIA-0341,
August 1982.
9. Radian Corporation. Background Information Document for Non-
Fossil Fuel-Fired Boilers. Draft Report (EPA-450/3-82-007), EPA Contract No.
68-02-3058, U.S. Environmental Protection Agency, Research Triangle Park, NC,
p. 3-3, March 1982.
10. Reference 9, model wood-fired boiler in Chapter 6.
A-6
-------
U.S. Cnv»«mm*ntal Protection Agency
Region V, Library
230 South Dearborn Street ^./
Chicago, Illinois 60604 _,^
-------