-------
1.0
0.8
O)
IB
0.6
0.4
X 0.2
A
5.2
Before ambient correction
B
7.3
C
7.6
D
7.8
E
9.5
oo
i*
N
CO
I
After ambient
correction
Engine number
13.3 NOX emission,
g/hp-hr § 75 grains/
Ib. dry air, 85°F,
29.92 inches Hg.
For Federal 13 mode composite cycle for HD diesel engines.
DAmbient humidity varied from 35 to 125 grains/1b dry; temperature from 70 to 115°F.
Figure 4-5. Effect of humidity and temperature on emissions scatter for six
HD diesel engines** (Reference 46).
-------
on whether the correction factor can be applied to data from (1) spark
Ignition (SI) or (2) compression ignition (CI) engines. Factors that have
been selected to correct both uncontrolled and controlled emission data from
the nine large-bore engine manfacturers are summarized following this
discussion.
Factors Applicable to SI Engines
Three ambient humidity correction factors are potentially applicable
to large bore, natural gas fueled engines, particularly four-stroke,
carbureted versions. These factors are summarized in Table 4-1(47,48,49)
Figure 4-6 is a comparison of the three factors over a typical range of ambient
humidities. Note that only one of the three factors is at a constant load
factor (Equation (2b)); the others are based on composite test cycles for
vehicles.
As Figure 4-6 indicates, there is a considerable difference in
correction factor depending on the study. All of the studies show,
nevertheless, that ambient humidity has a significant effect on NOx level-
The result for Equation (3), based on light-duty automotive gasoline vehicle,
shows the greatest sensitivity to variations in ambient humidity. These
varied responses of NOX level to changes in ambient humidity are not
unexpected since engines react differently to changes in inlet conditions.
Their response generally depends on their A/F ratio, fuel metering and
distribution system, and ignition characteristics. Since large-bore 1C
engines typically operate at a constant rated load, the constant load
correction factor (Equation (2b)) has been selected for application to the
reported data.
4-20
-------
TABLE 4-1. AMBIENT HUMIDITY CORRECTION FACTORS FOR SI ENGINES
no
Equation No.
(1)
(2a)
(2b)
(3)
Correction Factor
K = 0.634 + 0.00654(H) - 0.0000222(H)2 Compo-
site Factor (9 Mode Federal HD Gasoline Test
Cycle)
K = 0.796 + 0.175(H/100) + 0.129(H/1002)
Composite Factor (Federal Test Cycle, LD
Gasoline Vehicles)
K = 0.844 + 0.151(H/100) + 0.075(H/100)2
50 mph, Constant Load
K = 1/(1 - 0.0047(H-75)) Composite Factor
(Federal Test Cycle, LD Gasoline Vehicles)
H = specific humidity in grains H20/lb dry air.
Reference
47
48
48
49
-------
I
l\3
ro
-s
CD
cn
o
•a
cu
-s
O
3
CU
3
cr
o>
CD
•
CD
O
cn
CO
o
CO
en
K =
o
o
VO
tn
O
ui
tn
ro
o
ro
tn
ro
o
to
-S
a. ro
-.. o
o
O
CU
O
rt-
-"• -5
O -
3
CU
O
O
-S
Cf>
o
CO
o
o
o
ro
o
-Q -O -Q
to ro
CU
-------
Previous investigators have been unable to establish an ambient
temperature correction factor for spark ignition 1C engines because various
automotive engines respond quite differently to inlet air temperature
variations (see Appendix C.2). A limited amount of data exists for large 1C
engines that show the variation in NOX emissions with ambient temperature, or
manifold air temperature for turbocharged units. Figure 4-7(50) illustrates
the change in NOX level with a change in manifold air temperature for a large-
bore, four stroke per cycle, turbocharged (4-TC) gas engine. This response
indicates approximately a 1-percent change in NOX level per °F change in
manifold air temperature. (A change in manifold air temperature is nearly
equivalent to the same change in ambient temperature.)
Figure 4-8(51) indicates the response of NOx emissions with changes in
ambient air temperature for large-bore, blower-scavenged gas engines. These
results show approximately a 2-percent change in NOx Per 0|r Cnan9e-
Both.Figures 4-7 and 4-8 indicate that NOX emissions from large-bore engines
are very sensitive to ambient temperature variations. Therefore the results
of Figure 4-7 will be applied to all turbocharged gas engine data, and the
results of Figure 4-8 will be applied to all nonturbocharged gas engine data.
Factors Applicable to 1C Engines
A survey of the literature established two sources that have reported
ambient correction factors for truck-size diesel engines. In the study by
Krause, et. al., a factor was developed that included the effects of
temperature and humidity(52). The results of this study were subsequently
adopted by the EPA for mobile heavy duty diesel engines. The other study was
conducted by the Coordinating Research Council (CRC) and only investigated
4-23
-------
25 r-
20
s_
-C
Q. 1C
j- I 3
cr>
c
o
•£ 10
80 100 120 140
Air manifold temperature, °F
0»
««
I
<
160
Figure 4-7. Effect of manifold air temperature on a large-bore
4-TC engine (Reference 50).
4-24
-------
7 r
5 -
CO
O
—
_a
O
to
X
O
40 50 60 70
Air temperature, °F
80
90
100
Figure 4-8. Test results of NOx emissions versus intake air
temperature for two blower-scavenged gas engines
(Reference 51).
4-25
-------
the effects of ambient humidity(53). (A more detailed discussion of both
studies can be found in Appendix C.2.)
Figure 4-9(54»55) shows the ambient humi dity correction factors
developed from these two studies. The ambient humidity factors for SI
engines are also shown 1n Figure 4-9. In general, SI engines appear to be
more sensitive to ambient humidity variations than CI engines. Note that the
results from the two CI engine studies are shown for specific CI engine types
(e.g., four-stroke turbocharged, aftercooled engines). In general, the data
show that NOX emissions from different engine types, particularly at low
humidity levels, respond differently to changes in ambient humidity.
The Krause study also investigated the effect of anbient temperature
on NOX emissions. Figure 4-lC)(56) presents the correction factors that were
derived for these smaller bore engines. It shows that NOX emissions from
naturally aspirated and blower-scavenged engines are more sensitive to inlet
air temperature changes than are the emissions from aftercooled units.
Since the Krause study systematically examined the effect of both
temperature and humidity for a number of CI engine types, his correction
factors have been selected for application to similar large-bore engine
types.
4.2.1.3 Summary of Ambient Correction Factors for Application to Large-Bore
Engine Data
Table 4-2(57»58»59) summarizes the anbient correction factors that
have been selected for application to the data reported by the nine large-
bore engine manufacturers. Note that, with the exception of SI temperature
factors, all the corrections are based on studies of smaller bore automotive
engine types. The corrections for CI engines are given for specific engine
4-26
-------
K - (NO><>corrected/
O O
o o
-5 -O
CD fa
O -s
O O
3 3
-h O
Oi -h
O
c-h O
O «-<
-s
Q.
co
I—I
Cu
cr
n>
3
ri-
ro
o
ro
en
ro
CO
CO
ro
i
co
co
0>
en
en
-PI.
i
r>
i
3>
73
(D
cn
A-16249
-------
1.10
i
ro
c»
1.05
I/I
10
O>
X
o
i.o
S-
o
X
o
0.95
0.90
120
Ambient air temperature, °F
Fiaure 4-10. Correction factors for temperature for CI engines (Reference 56).
-------
TABLE 4-2. EXISTING 1C ENGINE AMBIENT CORRECTION FACTORS FOR APPLICATION TO LARGE-BORE ENGINES
Fuel
Diesel &
Dual Fuel (CI)
Humidity and
Temperature
Correction
Natural Gas (SI)
Humidity
Temperature
Correction Factor3
K = l/(l+A(H-75) + B (T-85))
H = observed humidity, grains H^O/lb dry air
T = observed inlet, air temperature, °F
Type A
2-BS -0.00242
4-NA -0.00231
4-TC, AC -0.00231
B
0.00235
0.00255
0.0017
/H \ / H \2
v n PAA 4. n i Ki / l-i. o n7£ i i
K - 0.844 i 0.151 ^1()0J« 0.075 \^QQJ
BS K = 1 - (T - S5)(0.017)
TC K = 1 - (T - 35)(0.010)
Comments
Rated load correction for
humidity and temperature
(from Ref. 57).
Rated load correction for
humidity only (from Ref. 58)
Ambient temperature correc-
tion only (from Ref. 59)
r\3
10
NOV corrected = (K) NO observed
A A
-------
types. In addition all the factors are based on rated load conditions, since
large bore engines typically operate at, or near, rated load,
Although ambient humidity and temperature variations can significantly
affect the NOX emissions that are measured from a particular engine, these
variations, in general, are not responsible for the large variations in
uncontrolled emissions that were reported for similar engine types by
different engine manufacturers. The other sources of data variability
(largely measurement practices and design differences among models) are
discussed in Sections 4.2.2 and 4.3.
4.2.2 Effect of Measurement Practices
Previous studies have shown that sampling instrumentation and
procedures have a large effect on emission levels. For example, a series of
studies conducted by the Coordinating Research Council (CRC) indicated that
uncertainties in the measurement of NOX levels can range as high as 40
percent^ '. This conclusion was based on the standard deviation of
measurements reported by different laboratories for the same emission source,
expressed as a percentage of the mean emission level. CRC concluded that this
uncertainty could be attributed largely to poor calibration and measurement
procedures. The EPA, then in cooperation with CRC, showed that these un-
certainties could be reduced to less than 5 percent using a specific set of
procedures(61). Since then, EPA has proposed that these procedures be used
to certify mobile, heavy-duty diesel and gasoline engines starting with the
1979 model year. The following paragraphs will briefly discuss the
measurement practices of each of the nine large-bore engine manufacturers who
reported emissions data. (Additional details regarding these practices can
be found in Appendix C.3.) Then uncertainties for each manufacturer's
4-30
-------
practice will be estimated relative to the proposed EPA procedure. These
uncertainties will be used in Section 4.3 to establish upper and lower bounds
for estimated average emission levels from large-bore 1C engines.
Table 4-3(62»63»64»65) indicates which of the four measurement
practices was used by each of the manufacturers. Note that two manu-
facturers, Alco and Ingersoll-Rand, used what is essentially the EPA
procedure. For the purpose of this discussion the EPA procedure will serve
as a reference for comparison with the other three procedures, since it is
believed to be the most accurate.
Figure 4-11 illustrates each of the four procedures schematically,
and Table 4-4 summarizes the sources of error for the DEMA, SAE, and EMD
practices. (A more detailed discussion of these procedures can be found in
Appendix C.3.) The primary shortcoming of these practices is their failure to
adequately define instrument performance and sample transfer procedures.
Unheated sample lines, inappropriate water removal devices, system leaks, and
failures of the converter in the chemiluminescent instrument all lead to
errors in the measurement of NOX in the sample gas. The use of NDIR
instruments (SAE/EMD practices), can lead to overstated values of NOX
emissions due to interferences resulting from the presence of water vapor in
the detector cell of the NDIR instrument. Considered together, these sources
of error can cause a large uncertainty in reported NOX levels. Figure 4-12
illustrates the overall uncertainty for data reported by each of the nine
engine manufacturers. (Again, a more detailed discussion of these
uncertainties can be found in Appendix C.3.) Note that manufacturers using
the SAE or EMD procedure could experience uncertainties of +20 percent.
Manufacturers using the DEMA practice, in contrast, are more likely to
experience understated (5 to 15 percent) NOX levels due to a loss of NOX
4-31
-------
TABLE 4-3. LARGE-BORE ENGINE MANUFACTURERS MEASUREMENT PRACTICES
Measurement Practice
Manufacturer EPAa DEMAb SAEC EMDd
Alco X
Caterpillar X
Colt X
Cooper Energy X X
DeLaval X
ElectroMotive (GMC) X
Ingersoll-Rand X
Waukesha X
White Superior X
(Div. Cooper)
aEPA's proposed practice for 1979 Heavy Duty Diesel and Gasoline
Enginesv62).
Diesel Engine Manufacturers Association (DEMA) Exhaust Emission
Measurement Procedure for Low and Medium Speed Engines(63).
°Society of Automotive Engineers (SAE) Recommended Practice J177a^63V
ElectroMotive Division of General Motors Corporation Practice^65).
4-32
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
TABLE 4-4. SOURCES OF ERROR FOR DEMA, SAE, EMD EMISSION PRACTICES
DEMA (CL)
CO
en
Unheated sampling lines permitted
No specification of sample residence
time in sampling line, or system re-
sponse time
Leak checks are not specified
Water removal device can be located
at analyzer, but before NO?-»-NO
absorption in water trap
Chemical driers permitted
No instrument specifications
Converter efficiency checks not
specified
No calibration procedures specifed
Calibration and span gas specifica-
tions not defined
EMD (NDIR/NDUV)/SAE (NDIR)
• Unheated sample lines
• No sample residence or system
response time specified
• Leak checks not specified
• Allow chemical drier
0 Calibration procedures not specific
(e.g., what constitutes out-of-
calioration, how calibration points
are curve fit, etc.)
0 Calibration and span gas blends and
dilutents not specified by SAE.
EMD has "own" specifications.
-------
during the transfer of the sample gas to the relatively interference-free
chemiluminescent analyzer. These estimates of uncertainty will be used to
place upper and lower bounds on the average uncontrolled emission levels
computed in Section 4.3.
4.3 UNCONTROLLED EMISSION LEVELS
This section presents data on uncontrolled emissions from large-bore
engines. Average uncontrolled NOX emissions from these engines, weighted
according to sales, were derived from data supplied by manufacturers. By
applying a specified degree of NOX control to these average uncontrolled
emission levels, potential controlled (regulated) emission levels can be
established. The degrees of NOX control that can be applied are identified
in Chapter 6, which summarizes demonstrated alternative controls.
This approach to setting the standards requires an adequate sample of
emission data for each manufacturer's engine. Section 4.3.1 discusses the
current data base, and shows that emissions data have been reported for about
80 percent of all the large-bore engine models manufactured to burn diesel,
dual fuel, and natural gas. This large existing data base is representative
of all the engines to be affected by standards of performance.
Section 4.3.2 presents the uncontrolled emissions data for diesel,
dual fuel, and natural gas engines, and examines the sources of variations in
these data. Differences in the ambient conditions (temperature and humidity)
and procedures for measuring the emissions account for only small variations
in the data. The largest source of data variations is differences in engine
design. These differences and their effect on NOx emission levels are
discussed in Section 4.3.3.
4-36
-------
Because of these differences, a method is needed to characterize
uncontrolled emissions from each of the three fuels for which standards of
performance will be proposed. In Section 4.3.4, representative uncontrolled
NOX levels for each fuel are determined by weighting each manufacturer's data
(corrected for ambients where possible). Weighting is based on the percen-
tage of total horsepower sold by each mnaufacturer during the past 5 years,
and the weighted levels are bounded by estimates of measurement uncertainty,
based on each manufacturer's procedures.
4.3.1 Existing Data Base
The extent of the current emissions data base is illustrated in Table
4-5, which shows the number of large-bore manufacturers who produce engine
models within the diesel, dual fuel, and natural gas categories. The second
row of the table shows the number of models produced for each engine type
(e.g., 2-BS, 2-TC, 4-NA, and 4-TC). The lower two rows show the number of
models within each fuel and engine type category that contain (1) uncontrolled
and (2) controlled emissions data. Uncontrolled emissions data are available
from every manufacturer of large-bore engines, although a few manufacturers
have not conducted tests to reduce NOX emissions from their engines.
In general, as shown in the last column of Table 4-5, there are
uncontrolled emissions data for about 80 percent of the models produced for
each fuel category. The current data base contains more data than for those
models listed in Table 4-5, since there are data for several engines of the
same model for some manufacturers. This additional data is useful in
determining differences in emissions data from engines of the same, and
different manufacturers, as discussed in Section 4.3.3. Thus, a substantial
data base exists for characterizing uncontrolled emissions from diesel, dual
fuel, and natural gas engines.
4-37
-------
TABLE 4-5. EXTENT OF EXISTING DATA BASE
Fuel
Strokes
Air Charging
No. of manufacturers
No. of models9
No. of models with
uncontrolled data
No. of models with
controlled data
Diesel
2
BS
2
2
2
2
TC
2
2
2
2
4
NA
1
3
1
1
TC
5
8
6
5
Dual Fuel
2
TC
1
1
1
1
4
TC
4
5
4
3
Gas
2
BS
2
2
2
2
TC
2
5
6
6
4
NA
4
5
4
3
TC
6
14
10
7
Diesel
6
15
11
10
Overal 1
Dual Fuel
4
6
5
4
Gas
7
28
23
18
CO
00
A model is a group of engines that share the same fuel, air charging, strokes/cycle, manufacturer,
bore, and stroke.
-------
4.3.2 Uncontrolled Emission Levels
Uncontrolled emissions of NOX, CO, and HC (and nonmethane HC where
measured) are shown on Figures 4-13 to 4-15. On each figure the data is
plotted separately for each fuel (diesel, dual fuel, and gas), and is
differentiated by engine type (i.e., 2-BS, 2-TC, 4-NA, and 4-TC). Since 1n
general, the CO and nonmethane HC levels from these engines are considerably
lower than the limits that apply to mobile vehicles and engines,4/ this
section is concerned primarily with NOX emissions. The effects of NOX
reduction techniques on CO, HC, and smoke emissions are discussed in Section
4.4.12. Figures 4-13 through 4-15 show that uncontrolled emission levels
vary considerably within each category of fuel and engine type. In Figure
4-13, fuel consumption, particularly for diesel and dual-fuel engines,
remains relatively constant despite wide variations In NOX levels among all
engine types. Since both NOX emissions and thermal efficiency increase as
cylinder temperature increases, efficient engines (low fuel consumption)
would be expected to show high NOX emission rates. As shown in Figure 4-16,
this is not the case among all engines of one fuel. NOX levels and four- and
two-stroke diesel engines shown on Figure 4-16(a), however, appear to
increase as fuel consumption decreases, but other trends are not apparent for
other fuels and engine types. Although other design features (e.g., manifold
air temperature, air-to-fuel ratio, speed, torque, etc.) are probably more
4/For example, the proposed Federal Government standards beginning in 1979
for heavy-duty gasoline and diesel engines are 1.5 g/hp-hr hydrocarbon
(HC), 25 g/hp-hr carbon monoxide (CO), and 10 g/hp-hr hydrocarbon plus ox-
ides of nitrogen (HC + NOX). California regulations for 1977-1978 heavy
duty vehicles (greater than 6,000 Ibs) are 1.0 g/hp-hr HC, 25 g/hp-hr CO,
and 7.5 g/hp-hr NOX.
4-39
-------
Otrfr
03
-p.
I
CO
o
o
rt
3
2.
X
(D
^.
(/>
l/i
«j*
O
(/I
O
D.
CD'
ro
ro
CD
III 11
a 2» * >
S « l» 1 ft
h
S
S =
s
&
iu
M
s
7 2
a
s
f
. s
* s
Ul
00
»
tn
S!
~
K
5
in
S S
s
s
3
S
s
s
IM
S
~;— a
NO, level, j/hp-hr
w»wo>or«»o.o.8~
O
o
•
•
•
0
•
*
0
o
0
0
*.
0
o
•
o
•
•
•
o
o
o
o
o
o
o
o
o
o
o
0
Brake specific
fuel consumption,
Btu/hp-hr
8888
o
D
Q
Q
O
Q
Q
Q
D
Q
Q
Q
O
Q
O
D
D
D
a
D
Q
O
O
D
D
Q
O
Q
O
if
!
Diesel:
•Itrolled
t
8
"
1!
|i
?l
i"
3
S
-------
Baseline s
Dual fuel:
Uncontrolled data
NOTE: All data corrected for ambients
o| 9000
2S- 8000
(A O I
u CL
l-sf 70°°
1- 3 +J
CQ H- CO
6000
Q Q
Q
Q Q Q Q Q Q
22
20
18
16
14
12
10
8
6
4
2
O
O
Engine nuafaer
Stroke configuration
13 14 78
2-TC
6 7 16 50 61 76
4-TC
Figure 4-13(b). Uncontrolled NOX emissions from dual fuel engines.
-------
217-1?
to
c
fO
I
CO
o
o
m
ex
x°
en
(/>
.^.
O
ca
Ol
n
rt n c/i am
ill If
tO -1 ft
C
fXJ
5 M
N,
to
A
r S
S g
01
s
s
-
s
k -J*
*s
s
*
s
ro
ro
8
«
U)
* 8
rt §
Ul
s
CO
*
01
w
CJ1
09
RS
NOX level, g/hp-hr
ro > 01 oo o isi A 01 co o fo
•
•
o
0
0
O
0
o
o
o
0
^
•
•
•
0
•
•
•
•
•
•
•
0
P 0
o
o
^
Brake specific
fuel consumption,
Btu/hp-hr
S S S S
1 1 1 1
o
Q
O
Q
&
B
(D
Q
D
ID
Q
Q
Q
-S Q
Q
D
O
Q
D
Q
Q
D
Q
Q
D
O
O
D
Q
VI 09
C
O
Is?
O VI
a.
a.
CU
s
1
m
S1
3-
a.
1
i/i
c
5
i
S1
1
I'
cf
A- I
-------
•^
co
tMtlfltt
swnury
«
S
4
k.
. 3
S
S
2
1
CMjlM
nwbcr
Strati
tlon
Oteiel:
Uncontrolled (tati
O
o
o
.
0 0
W 11 17 W 43
l-K
O
o ©
o ®
3 1? 1» 1» 41 7? 73 74 77
l-Tt
O
?4 35
4-M
0 00
0
o
° 0
0 0
o
0
o
5 a 15 » ?7 36 55 SC 57 U 5* M f* 75
4-Tt
Figure 4-14(a). Uncontrolled CO emissions from diesel engines.
-------
Baseline siMmary
Dual Fuel
Uncontrolled Data
Engine Number
6 7 16 61 76
Stroke Configuration
Figure 4-14(b). Uncontrolled CO emissions from dual fuel engines.
-------
J^
tn
Incline tewy
3.D
Z.O
f
s
1.0
CngliwIMwr
Cimrifirttlan
O
0
1 71
2-K
0
&
o
2 69 70
Z-TC
CM:
u
0
o
o
o
21 22 n 44 74 m
4-W
n i r r
O 0 O O O
o
o
e
o o °
-
*
Figure 4-14(c). Uncontrolled CO emissions from gas engines.
-------
Incite
1.2
1.0
0.8
I
m^
|
0.4
0.2
bite
Stroke
mffojn*-
tlm
•
0
0
0
0 ©
10 11 17 42 43
»
OKI
O
0
0 0
0
o o o
0
3 12 18 11 41 72 73 74 77
2-IC
Mewl:
0
0
24 35
4MM
O
0 0
0
0 0
o o o 0 o
o
0 ° 0
0
0
i » 9 IS » 27 36 47 51 52 S5 56 57 58 59 60 62 75
4-TC
*HC Invl KMKII to to ill
Figure 4-15(a). Uncontrolled HC emissions from diesel engines.
-------
toettae!
l F«el:
lcri Bit*
o
o
J3L
Total HC
O
O
Q
13 14 711
f 7 1C SO Cl 7C
Stnto mflgnvtln
t-K
«-TC
Figure 4-15(b). Uncontrolled HC engines from dual fuel engines
-------
-pi
oo
Baseline
8
7
S
5
. 4
* 3
2
1
tnglne rater
Stroke
conflagration
Gas:
Uncontrolled data
-
-
.
•
-
0
• o
o
1 71
2-BS
o
o
0
ft
2 69 70
2-TC
0
0 °
0
22 79 39 » 44
4-M
Total HC 0
italMCttMtlt* HC U
o
0
o o
0
e
0
o
0 0
0 0 0 0 0
D 0 °
42023 282*373140454648495354810?
4-TC
Figure 4-15(c). Uncontrolled HC emissions from gas engines.
-------
-E*
vo
9000
8000
7000
6000
I Q
^
2-BS 4-NA
Diesel: O Q & V
2-TC 4-TC
NOTE: Darkened symbols
uncorrected for
ambients
o
8 10 12 14
NOX level, g/hp-hr
16
18
20
22
Figure 4-16(a). Uncontrolled NOx levels versus brake specific fuel consumption (BSFC)
for diesel engines.
-------
9UUU
8000
i_
-C
1
0.
-c:
•^
3
•4J
00
1 «->
S fc
O CQ
7000
6000
2-TC 4-TC
Dual fuel ID ^
.
Q
9*
Q ^
• i^
•
•if
V
Q ^
8 10
14
16 18 20 22
N0x level, g/hp-hr
Figure 4-16(b). Uncontrolled NOX levels versus brake specific fuel consumption (BSFC)
for dual fuel engines.
-------
01
9000
8000
7000
6000
m o
2-BS 4-NA
Gas: O O & V
2-TC 4-TC
NOTE: Darkened syMbols
uncorrected for
ambients
10
12
14
16
18 20
22
HOX level, g/hp-hr
Figure 4-16(c). Uncontrolled NOX level versus brake specific fuel consumption (BSFC)
for gas engines. '
-------
Important 1n affecting uncontrolled NOX emissions and fuel consumption,
changes 1n operating conditions intended to reduce NOX emissions, generally
causing fuel consumption to increase, as discussed 1n Section 4.4.
As noted in Figure 4-13(a), not all the NOX data can be corrected for
the effect of ambient temperature and humidity. (Ambient correction factors
for large-bore engines are summarized in Section 4.2.1.) The effects of
ambient variations, differences in measurement practices, and Inherent
differences 1n engine design on the variability of uncontrolled NOX emissions
are summarized 1n Table 4-6(66), The data samples include only those data
from Figure 4-13(a) that could be corrected for ambient variation. As this
table indicates, the largest source of variations 1n data 1s Inherent
differences In engine design. This conclusion is similar to that of an
investigation of sources of emission data variability in gasoline
vehicles^67). In this study, researchers demonstrated that although
measurement and ambient effects were significant, variations among vehicles
caused most of the variations In NOX emissions for a series of tests on similar
vehicles. Sources of emissions variability due to engine design are discussed
in the following section.
4.3.3 Effect of Engine Variability on NO* Emissions
The emission data supplied by the manufacturers vary considerably. As
discussed above, small variation can be attributed to using different
measurements techniques or not correcting for ambient conditions. However,
most of the differences in emissions from uncontrolled engines result from:
(1) variations in the production of a particular model, (2) variations among
different models of the same type (i.e., same strokes/cycle, air charging and
fuel), or (3) variations in the number of cylinders for a given model. In
4-52
-------An error occurred while trying to OCR this image.
-------
this section, the uncontrolled data base is evaluated for these sources of
variability, and furthermore, the data are examined for trends related to
differences in engine design, such as speed, torque (bmep), and manifold air
temperature. All of these analyses use data corrected for ambient variations
by the methods described in Section 4.2.1.
4.3.3.1 Production Variations
It is difficult to quantify variations in emissions among production
units of the same model. Up to now, manufacturers have concentrated on
obtaining emissions data for different engine models. Since no emission
regulations (with the exception of smoke limits) have been 1n effect for
stationary engines, there has been little Impetus for manufacturers of large-
bore stationary engines to make exhaust measurements of engines leaving the
production line. (In general, fewer than 100 units are produced each year
for stationary applications by any one manufacturer.) However, Colt and
GMC/EMD, as well as numerous manufacturers of smaller bore, heavy duty
engines for trucks have reported variations in emissions from production
models.
One large volume manufacturer of medium-bore engines has shown that
their laboratory units must emit at levels at least 25 percent lower than a
performance standard, to insure that their production models will comply with
the standard^**). jhjs margin accounts for production variables that effect
emission levels in mass produced engines. For 75 percent of this
manufacturer's current engines meeting the Federal automotive emission
standard of 16 g/hp-hr (NO + HC), current variation 1n 1.34 g/hp-hr (N0¥ +
A A
HC). Moreover, this manufacturer believes that the magnitude of production
variation is independent of the emission level, and this belief is shared by
4-54
-------
several other manufacturers of medium-bore engines(69). For large-bore
engines, produced individually to higher tolerances, it is anticipated that
this variation should be smaller.
Colt reported less than a 3-percent difference between production
models in two NOX measurements of one in 1972, the other in 1975, from a
two-stroke, blower-scavenged diesel eng1ne(70). $uch a m&}-\ difference was
unexpected. They suggest that the variation would more likely be of
approximately 10 percent for production units, but they have no data to
verify this estimate. Colt has measured NOX levels of production spark
ignited engines (2-TC-6) within 3 percent of each other under similar ambient
conditions^1).
GMC/EMD has reported average NOX levels and standard deviations for
samples of their 2-TC and 2-BS diesel models(7^). These results are
summarized in Table 4-7(73). /\s this table suggests, these variations in NOX
levels of production engines may have resulted from ambient variations.
Inlet air temperatures varied over a wide range for both the turbocharged and
blower-scavenged units, and humidity was not recorded. An attempt was made
to determine whether these observed variations in emissions could be due to
changing ambient conditions. First, a correction was computed for each
extreme of the reported temperature range, using the methodology presented in
Section 4.2.1. These two maximum variations were then compared to the
reported data, to determine if temperature variations alone could account for
the scatter. Next, a correction was computed for both the reported
temperature variations and an assumed humidity variation ranging from 35 to
115 grains O/lb dry air. These corrections were then compared with the
production variability; the results are listed 1n Table 4-8.
4-55
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
Based on these corrections, temperature variations alone could account
for all of the variability in emiss-ions from blower-scavenged engines, but
not for turbocharged engines. If humidity were to vary over the range used
for the calculations presented In Table 4-8, then differences 1n ambient'
temperatures and humidity could account for all the variability reported for
these engines.
4.3.3.2 Model Variations
Variations in levels may be attributed to differences in models for a
given manufacturer's engines. For example, NOX levels for a manufacturer's
4-TC-6 models may vary due to differences in bore, stroke, turbocharger,
configuration (inline cylinders vs. vee), compression ratio, aftercooler, and
other engine design parameters. In an effort to identify the magnitude of
model-to-model variations, average NOX levels and standard deviations were
evaluated for different models of the same fuel type from each manufacturer.
Table 4-9 presents the results of this study. GMC/EMD, White Alco,
and Colt are not included in this table since they each manufacture only one
engine model (with different numbers of cylinders) per air charging method.
Seme of these models are configured for different fuels, for example, Colt
markets their 38D8-1/8 opposed piston engine model as a gas, diesel, or dual
fuel engine, either blower scavenged or turbocharged. GMC/EMD and White Alco
manufacture one basic dlesel-fueled, turbocharged design which differs
primarily in number of cylinders and speed ratings. GMC/EMD also markets
blower-scavenged units.
The other five manufacturers listed 1n Table 4-9 produce different
engine models within a given engine type. NOV levels reported by Cooper for
r\
four 2-TC models varied by an average of 13 percent. These engines differed
4-58
-------An error occurred while trying to OCR this image.
-------
1n bore, speed, number of cylinders, and torque (bmep), but were all operated
at the same Inlet and manifold air temperatures. Del aval's data Indicated
only a 4- to 5-percent variation between models for both gas- and diesel-fueled
engines, The percent variations shown for Waukesha, Ingersoll-Rand, and White
Superior, which were uncorrected for ambient conditions, should not be compared
to the Cooper and Del aval results because differences due to ambient conditions
could not be factored out. To the extent that conclusions can be drawn from
such a small sample size, 1t appears that NOX emissions for any type of engine
(given strokes/cycle, fuel, and air charging) vary more from manufacturer
to manufacturer than among models 1n a manufacturer's line. Emission variations
due to differences among manufacturers could be related to differences 1n
speed, bmep, or manifold air temperature. This possibility 1s addressed 1n
Section 4.3.3.4.
4.3.3.3 Variations With Number of Cylinders
Several manufacturers have suggested that NOX levels will vary for a
basic engine design depending on the number of cylinders, since the manifold
interacts with the turbocharger. Figure 4-17, which is a plot of NOX level
(corrected for ambients) vs. number of cylinders, shows that NOX levels for
4-TC gas engines decrease significantly as number of cylinders increase, but
NOX emissions from 4-TC diesel and dual-fuel engines do not indicate a trend
with number of cylinders.
Figure 4-18 presents a different interpretation. The NOX levels
(corrected for ambients) have been plotted vs. the number of cylinders for
each manufacturer's engines, which tends to reduce other sources of emissions
variation (such as design differences among manufacturers) that may have been
4-60
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
reflected in Figure 4-17. Figure 4-18 indicates that there is no clear trend
between NOX emissions and number of cylinders for either diesel or gas units.
The effect of changing the number of cylinders causes uncontrolled NOX levels
to vary from 3 to 9 percent. Because EMD data could not be corrected for
ambients, this data may not represent the effect of differences in the number
of cylinders.
4.3.3.4 Variations in NOX Level Due to Other Engine Variables
The results described above suggest that the variations in NOX levels
reported for engines of a given type are probably due to design parameters
that differentiate one manufacturer's engines from those of the others.
Consequently, uncontrolled NOX data (corrected for ambients) were plotted vs.
speed (rpm), manifold air temperature, and torque (bmep) to reveal any
emission trends with these design parameters.
Figure 4-19 illustrates NOX level variation with speed for two engine
types. The data for gas engines indicate increased NOX emission with in
creased speed. This is in contradiction to what one would expect from the
reasoning that decreased residence time (increased speed) should result in
lowered NOX emissions. Apparently other factors (e.g., increased cylinder
temperature or inherent design differences among different engines) are
responsible for this trend. The 4-TC dual fuel and diesel NOX levels appear
to decrease somewhat with increasing speed, as would be expected for lower
exhaust gas chamber residence times.
Variations in NOX level with manifold air temperature are shown in
Figure 4-20. The 4-TC-6 NOX levels appear to be very sensitive to the design
air manifold temperature, but the diesel and dual fuels NOY levels do not.
/\
4-63
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
These results confirm expectations that NOX levels from premixed, vaporized
fuel combustion in SI engines may be strongly influenced by the degree of
aftercooling. On the other hand, CI engines are characterized by droplet
combustion, and NOX production under these conditions would depend more on
local A/F ratio than on overall air temperature.
Finally, Figure 4-21 illustrates the wide variation of NOX level with
torque (bmep) for 4-TC-G and 4-TC-D, DF units. If the cluster of CI data
around 150 psi is ignored, a trend of decreasing brake-specific NOX emissions
with increasing bmep for these units is apparent. No trend is apparent for
SI units, except that they are generally not manufactured with bmep's
exceeding 200 psi.
Based on these preliminary studies, it appears that certain engine
design parameters may explain more of the variation NOX levels for engines of
a given type than variations in ambient humidity or temperature. That is,
NOX emissions for any type engine (given strokes/cycle, fuel, and air
charging) vary more from manufacturer to manufacturer than they do among
models within a manufacturer's line. Differences among manufacturers are
related to differences in speed, torque (BMEP), manifold air tanperature, and
combustion chamber design. For example, limited data show that NOX emissions
from 4-TC diesel and dual fuel engines decrease as speed increases. NOX
emissions from 4-TC, natural gas (SI) engines increase directly as design
manifold air temperature increases. However, no clear trends can be
established for the effects of number of cylinders and torque (bmep). These
observations suggest that some form of weighted average is required to
characterize uncontrolled NOX emissions from each of the three fuels. This
approach is discussed in the following sections.
4-66
-------An error occurred while trying to OCR this image.
-------
4.3.4 Sales-Weighted Uncontrolled Emissions
Since the sources of variability due to engine design cannot be
specifically identified, a procedure is required to characterize uncontrolled
emission levels of engines which are sold for similar applications.
The procedure adopted here is to compute a weighted, average uncon-
trolled emission level for engines in the diesel, dual fuel, or natural
gas categories. The three weighted levels are based on sales of engine
horsepower during the past 5 years for domestic applications. Sales of
horsepower to standby services were excluded from this computation, since
engines sold for standby applications will be exempted from standards of
performance (see Chapter 9). Therefore, these engines should not influence
the selection of regulated emission levels.
The sales-weighted averages for diesel, dual fuel, and natural gas
engines are presented in Figure 4-22, which also show each manufacturer's
uncontrolled NOX data. The weighted averages are based on data corrected for
ambient conditions where possible. The weighted average uncontrolled NOX
level for diesel engines is 11.0 g/hp-hr, for dual fuel units, 8.1 g/hphr, and
for natural gas engines, 15.0 g/hp-hr. The emission reductions discussed in
Section 4.4 and summarized in Chapter 6 can then be applied to these levels
to determine potential regulated levels of NOX.
Measurement uncertainties are associated with each of these weighted
levels and are shown in Table 4-10. These uncertainties should be applied to
the controlled NOX levels that are determined by applying the NOX reductions
demonstrated by the alternative control systems.
4-68
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
IZ-fr
^
ro
ro
00
o>
re
VI
CO
c
n
§
ut
(fl
•rft
o
ua
»
en
ro
in
rf o i/i am
-" 2 <•* E 3
§|l It
o>
1
IV
09
8 2
M
LJ
2
INJ cn
* R
01
»-J
2
3
fS)
*S
s
*
a
B
a
3
y
r
rt s
*
*
s
s
a
2
NOX level , s/hp-hr
f
• 1
-
0
-
0
o
0
0
"
o
•
"
r •
-
-
t
•
•
•
•
00 1
" r
B ' • ' • • i i i*
0
o
5
•<
v
]•
!_ ^
if
| 0
tl
\
*
•
•
0
o
o
©
•
Brake specific
fuel consumption,
Btu/hp-hr
i I 1 i
1 "^^"^^^^^ww
Q
Q
D
D
D
a
o
B
D
D
O
Q
Q
-3 Q
Q
D
D
D
D
Q
Q
D
Q
D
D
O
0
O
O
No fuel
No fuel
fi
qi
• MMM^
f
3£
i
&
£
2
3
|
1
I
]"
=
i
$
I
s'
-------
TABLE 4-10. WEIGHTED MEASUREMENT
UNCERTAINTIES FOR SALES
WEIGHTED NOX LEVELS
Fuel
Diesel
Dual Fuel
Natural Gas
Upper,
g/hp-hr
1.0
1.0
1.3
Lower,
g/hp-hr
0.9
0.4
0.5
4-72
-------
4.4 NOX EMISSION REDUCTION TECHNIQUES
This section describes techniques that have been used, or are being
evaluated for use, to control NOX emissions from 1C engines. Sections 4.5
and 4.6 discuss control techniques that are designed primarily to reduce
other pollutants (HC, CO, smoke). Since SOX emissions are directly related to
the fuel sulfur content, these emissions are discussed in 4.4.13, combustion
of nonstandard fuels.
The data presented here come from tests on engines whose operating
conditions were altered or which were equipped with emission reduction
devices. These tests were conducted in manufacturers' laboratories rather
than in field installations. The discussion of each potential control
technique centers on how the technique works, its effectiveness, resulting
fuel penalties, effects on other pollutant emissions, technical limitations
to its applications, and cost implications (i.e., additional fuel,
maintenance, or hardware expense incurred by the application of the control).
Most techniques for controlling emissions from 1C engines involve
engine modifications rather than add-on tail gas treatment facilities.
Engines are designed for optimum operation within one or more of the
following constraints: application, initial cost, fuel consumption,
maintenance requirements, reliability, and commitment of a company's
engineering staff to a design.-' Each engine design satisfies the constraint
— Stationary reciprocating 1C engines, and particularly the large ones, may
be required to deliver thousands of hours of continuous operation at rated
load under varying ambient conditions without significant maintenance, or
to start without failure by remote control and deliver full power within
10 seconds. Since these are severe demands, manufacturers feel committed
to a proven design and are, therefore, reluctant to make significant design
changes (e.g., changed piston or cylinder shape or strokes per cycle).
4-73
-------
differently. For example, one engine may operate at 4e BTDC while another at
5° BTDC to meet the same NOX emission level. Therefore, the data are grouped
by engine type and fuel in the tables and graphs that follow. In addition,
whenever there 1s a specific, known reason why one type of engine responds
differently to the application of controls than does another, these
differences are explained in the accompanying discussion of the control
technique.
The reductions 1n NOX shown here were achieved by investigators for current
production engines. In general, no attempt was made to optimize the engine
for the controlled settings (I.e., decrease fuel consumption, reduce maintenance,
etc.). Thus, these results must be viewed as those achievable 1f no attempt
1s made to reoptlmize an engine's controlled setting.
As discussed in Section 4.2.2, the manufacturers' data were measured
using one of four measurement practices (EPA, DEMA, SAE, EMD). Although
differences in three of these practices relative to EPA's may cause
uncertainties in the reported levels, the data are considered adequate for
the purpose of setting standards of performance since these are small 1n
comparison to those in emissions due to inherent differences in engine
design. (Measurement uncertainties for uncontrolled emissions are discussed
in Sections 4.3.2 and 4.3.4.) Furthermore, the reported emissions data have
been corrected to standard conditions of humidity and temperature (when
ambient data were recorded) using the ambient correction factors presented 1n
Section 4.2.1 Dashed lines on the figures 1n this section Indicate NO reduc- ,
A
tions after ambient correction.
The control systems discussed in this section are listed below 1n
their order of presentation.
1. Derating (D)
2. Retard (R)
4-74
-------
3. Changed air-to-fuel ratio (A/F)
4. Turbocharging with aftercooler (TC)
5. Reduced manifold air temperature (MAT or M)
6. Exhaust gas recirculation (EGR) — internal (IE) and external (EE)
7. Water induction (HgO)
8. Combustion chamber redesign (CCR)
9. Catalytic converters
10. Combinations of the above
Several abbreviations will be used on the charts and tables in this chapter;
they are listed in Table 4-11. Fuel consumption data on the charts and
tables are based on a lower heating value (LHV) of 18320 Btu/lb (10160
kcal/kg) for No. 2 diesel oil.
A qualitative summary, by pollutant, of the effect of each control
technique on each engine type as shown by the available data is presented in
Table 4-12. Sections 4.4.1 through 4.4.10 give quantitative results for each
of these techniques. Graphs are presented that show (1) the NOX reduction for
the largest degree of control applied for each manufacturer and (2) the effect
on emissions and fuel consumption as the amount of control is varied. The
information presented in the second set of graphs has been normalized by the
baseline or uncontrolled level. This condition is denoted on the graphs with
a subscript "U" for uncontrolled. The controlled condition is denoted with
a subscript "C" for the controlled level. Section 4.4.11 summarizes the data
presented for each of the above control approaches.
Then in Section 4.4.12, the effect of NOX control on the emission of
other pollutants is examined. This review will help to illustrate whether
standards of performance may be required for other pollutants in addition to
4-75
-------
TABLE 4-11. ABBREVIATIONS FOR ENGINE TYPE AND EMISSION CONTROL
TECHNOLOGY
Abbreviation
Explanation
Fuel
D
DF
G
Strokes/Cycle
2
4
Air Charging
BS
NA
TC
Control Technology6
D
R
TC
A or A/F
M or MAT
EGR(I)
EGR(E)
Sb
INJ
H20
W/F or w/f
PC
VT
CCR
Cat
Diesel
Dual Fuel
Gas (i.e., natural gas)
2-stroke/cycle engine
4-stroke/cycle engine
Blower scavenged
Naturally aspirated
Turbocharged (and intercooled)
Derating
Retard
Turbocharged (and intercooled)
Increased air-to-fuel ratio
Decreased inlet manifold air temperature
Exhaust gas recirculation - internal
Exhaust gas recirculation - external
Increased speed
Modified injectors
Water induction
Water-to-fuel mass induction ratio
Precombustion chamber
Variable throat precombustion chamber
Combustion chamber redesign
Catalytic converter
aRM and RMA are used to denote the combined use of retard,
decreased air temperature, and increased air-to-fuel ratio
blncreased speed included because data are available, but it is
not considered to be a viable control technique (Subsection 5.3.12)
4-76
-------
TABLE 4-12(a). EFFECTS OF CONTROLS ON ENGINES LARGER THAN 350 IN3/CYL: NOX EMISSIONS
Fuel
Strokes/Cycl e
* — ^ir Charging
Control "*" — -^^^
Derate
Retard
A/F
TC
MAT
INJ
EGR(I)
EGR(E)
H20
CR
Diesel
Two
BS
4
4
4
4
4
4
TC
4
—
t
+4
4
Four
NA
4
TC
4
4
4
—
4
t
4
4-
Dual Fuel
Two
BS
TC
4-
4-
—
4-
t
Four
NA
TC
4-
4-
4-
—
4-
t
4-
Natural Gas
Two
BS
4-
4-
4-
4-
4-
4-
TC
4-
4-
4-
—
4-
Four
NA
4
4-
4
t
4-
4-
TC
4
4-
4
—
4
4
t Denotes emission increase with application of control
4 Denotes emission decrease with application of control
•f-4 Denotes conflicting data with application of control
- Denotes no change in emissions with application of control
Blank indicates no data available on effect
-------
TABLE 4-12(b). EFFECTS OF CONTROLS ON ENGINES LARGER THAN 350 IN3/CYL: CO EMISSIONS
i
^j
oo
Fuel
Strokes/Cycle
"•"--^Ajr Cha rgi ng
Control •"••^-^^^
Derate
Retard
A/F
TC
MAT
INJ
EGR(I)
EGR(E)
H20
CR
Diesel
Two
BS
4-
4-
4-4
4
4
4
TC
4-
4-
44
4-
Four
NA
TC
4
t
4-4-
44
4-
t4-
Dual Fuel
Two
BS
TC
t
f
4-
4-
Four
NA
TC
f4-
4-
4-
4-4-
4-
4-
Natural Gas
Two
BS
4-
i
4
4-
4-
TC
4-
4-
4-
4-
Four
NA
4-
—
4-
—
4-
TC
4-
4
4-
4-
4- Denotes emission increase with application of control
4 Denotes emission decrease with application of control
4-4 Denotes conflicting data with application of control
— Denotes no change in emissions with application of control
Blank indicates no data available on effect
-------
TABLE 4-12(c). EFFECTS OF CONTROLS ON ENGINES LARGER THAN 350 IN3/CYL: HC EMISSIONS
-p»
ID
Fuel
Strokes/Cycle
"""^---^^Tr Charging
Control -"-^..^^
Derate
Retard
A/F
TC
MAT
INJ
EGR(I)
EGR(E)
H20
CR
Diesel
Two
BS
—
4-
t
4-
f
TC
t4-
*4-
4-
—
Four
MA
4-
TC
+
t4-
+4-
H
4-
t4-
Dual Fuel
Two
BS
TC
t
t
f
4-
Four
NA
TC
t
4-
t
++
t
t
Natural Gas
Two
BS
t
t
4-
+
t
TC
t
t
t
t
Four
NA
t
t
t
TC
t
t
t
t
t
t Denotes emission increase with application of control
4- Denotes emission decrease with application of control
+4- Denotes conflicting data with application of control
— Denotes no change in emissions with application of control
Blank indicates no data available on effect
-------
NOX. Finally, Section 4.4.13 discusses the effect of burning nonstandard
fuels on emissions from stationary 1C engines.
The data that appear in this chapter can also be found in Appendix
C.I, where they are tabulated by engine. This appendix presents all the
available data on emissions and fuel consumption for large-bore engines.
4.4.1 Derating
When a manufacturer advertises or sells an engine, he guarantees that
1t will produce a given power at a stated speed. These conditions are called
"rated conditions" and can be specified either for maximum, intermittent, or
continuous operating conditions. The maximum rating usually refers to the
peak power that can be achieved by the engine, but manufacturers generally
recommend that the engine not be operated at this level. Intermittent
ratings typically indicate the power output that the engine can produce for
a 1-hour period with at least a 1-hour period of operation at, or below, the
continuous rating before the next surge to the intermittent level(74).
Continuous rating, of course, applies to uninterrupted operation (e.g., 24
hours per day, 365 days per year with shutdowns for maintenance only).
An engine can be derated by restricting its operation to a lower level
of power production than normal for the given application. The effect of
derating is to reduce cylinder pressures and temperatures and thus to lower
NOX formation rates. Although NOX exhaust concentrations (i.e., moles of NOX
per mole of exhaust) are reduced, it is quite possible for this reduction to
be no greater than the power decrease. In such a case brake specific
emissions (i.e., grams NOX per horsepower-hour) are not reduced. This is
especially true for four-stroke turbocharged engines as shown in Figure
4-23(75). In addition, air to fuel ratios change less with derating for
4-80
-------
-p.
CO
I
Q.
c
o
0
40
60 80 100
Percent rated torque
120
Note:
All engines
from same
manufacturer
Figire 4-23. NOX emissions versus torque at constant speed (Reference 75).
-------
turbocharged engines than for naturally aspirated or blower-scavenged units.
Thus NOX emissions are less responsive to derating for turbocharged engines.
Derating also reduces the engine's operating temperature, which then results
in higher CO and HC emissions. This happens because the temperature
dependent reactions that reduce these pollutants are less active(7£>).
Demonstrated NOX emission reduction levels due to derating are shown
in Figure 4-24 for a number of different engine types and fuel. Based on
these data, emission reductions ranged from 1.2 to 23.0 g/hp-hr for naturally
aspirated or blower-scavenged engines and from 0.2 to 10.8 g/hp-hr for
turbocharged units. Since these results were obtained with varying amounts
of derating, it is more informative to compare the effectiveness of this
emission control technique on a normalized basis -- i.e., percent NOX
reduction per percent derate. On this basis, results for naturally aspirated
or blower-scavenged engines varied from 0.25 to 6.2, whereas those for
turbocharged units varied from 0.01 to 2.6. No relationship was found
between normalized effectiveness and uncontrolled emission level, number of
strokes per cycle, or fuel.
Figure 4-25 illustrates the effect of different amounts of derating on
NOX emissions and fuel consumption for diesel, dual fuel, and gas engines,
respectively. Figure 4-25(a) shows that derating decreases brake specific
NOX emission from some diesel engines, but increases them from others. In
general, NOX reductions range from 2 to 25 percent for 25 percent derating
and brake specific fuel consumption increases range from 2 to 5 percent.
With 50-percent derating, NOX reductions range from 15 to 45 percent and fuel
penalties from 4 to 16 percent. Engines No. 10 and 11, both two-stroke
blower-scavenged, achieved the largest NOX reduction (and highest fuel penalties),
4-82
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
Q2-TC
o3
1
0°
g
1.40
1.30
1.20
1.10
i.no
•
#61
#14
#76
#78
1 7
60 70 80
% rated torque
90
Too
Figure 4-25(b). Effect of different amounts of derate on
NOx emissions and fuel consumption on dual
fuel engines.
4-85
-------
l/l
CO
o
LL.
1.30
1.20
1.10
1.00
1.10
i.on
0.90
0.80
0.70
3
5 0.60
^y 0.50
*"><
§. 0.40
0.30
0.20
0.10
#23
#71
O 2-BS
Q 2-TC
& 4-NA
t?7 4-TC
#70
50 60 70 80
% rated torque
90
100
Figure 4-25(c). Effect of different amounts of derate on
NOX emissions and fuel consumption for
gas engines.
4-86
-------
Figure 4-25(b) also shows mixed results for derating dual-fuel engines
but with substantially less variation than among the dlesels. Thus derating
caused an Increase in NOX in only one engine. In general, 25 percent derating
reduces NOX emissions 20 to 35 percent and increases fuel consumption 2 to
8 percent. Derating by 50 percent produces NOX reductions of 30 to 65 percent,
but at the same time fuelusage goes up 10 to 30 percent. In general, small
amounts (25 percent or less) of derating appear effective in reducing NOX
emissions from dual fuel units, and such reductions are accompanied by fuel
penalties of less than 8 percent.
Figure 4-25(c) shows that the derating of gas engines produces a wide
range of NOX reductions. In general, the nonturbocharged engines achieve
the largest reductions, since derating has a greater effect on their air-to-
fuel ratio. For example, blowers on blower-scavenged units operate at
constant speed, independent of load; therefore, as the fuel flow is reduced
to decrease output, the air-to-fuel ratio increases, causing a NOX reduction.
The turbocharged engines, in contrast, maintain a more nearly constant air-
to-fuel ratio, and consequently, experience less of a NOX reduction.
Derating does not require additional engine equipment, and the only
operational adjustment is to the throttle or governor setting in order to
restrict the engine power output. In most cases, this adjustment can be made
in the field, although one could presumably equip a new engine with a fuel
pump or carburetor whose maximum fuel delivery capacity corresponds to a
derated condition. When derated, the engine's efficiency is reduced, and
hence, the fuel consumption is increased. Moreover, when an engine is
derated, a bigger, more expensive unit must be purchased to satisfy a given
power requirement.
4-87
-------
4.4.2 Retarded Ignition Timing
As mentioned in Chapter 3, combustion is initiated by the injection
of fuel oil in a diesel or dual-fuel engine, or by a spark in a natural gas
unit. The effect of variations in the timing of this injection or spark
discharge is the same for both kinds of engines; that is the event can be
described as combustion ignition in both cases. Therefore, this control
technique is termed retarded ignition timing, or retard.
Ignition in a normally adjusted engine is set to occur shortly before
the piston reaches its uppermost position (top dead center, or TDC). At TDC
the air or air-fuel mixture is compressed to the maximum. The timing of the
start of injection or of the spark is given in terms of the number of degrees
that the crankshaft must still rotate between this event and the arrival of
the piston at TDC. The extent of retard is then expressed in degrees
relative to normal ignition. Typical retard values are 2° to 6°, depending
on the engine. Beyond these levels fuel consumption increases rapidly, power
drops, misfiring (erratic ignition) occurs, and smoke from diesel engines
becomes excessive(77).
After ignition, the burning combustion gases expand, driving the piston
downward. This is called the power stroke. When ignition is retarded, the
duration of the combustion process does not change significantly, but rather
is initiated closer to TDC and is extended longer into the power stroke.
Consequently, the combustion process occurs later during higher exhaust
temperatures.
In theory the fuel delivery system in diesel engines could be altered
to reduce the duration of injection and thereby decrease the quantity of fuel
that is combusted late in the power stroke. Such changes would require
increases in the injection pressures above current levels, which are already
4-£
-------
high. One manufacturer of medium-bore engines, Cummins Engine Co., developed
their own high pressure fuel pump for this application because there was no
suitable commercial component on the market. Low-volume manufacturers of
medium-bore engines, however, depend on outside sources for their fuel pumps
and have stated that the inavailability of this critical component restricts
their use of retard (e.g., to 2° to 6°) at this time(78). One manufacturer
of large^bore engines reported an unsuccessful attempt at increasing
injection pressures^79). He found that this higher pressure compressed the
puel and expanded the fuel lines and consequently, the fuel injection time
was not decreased. Presumably, the test could be conducted with fuel
handling equipment so that expansion of the fuel lines, at least, would not
prevent a manufacturer from reducing his injection period.
Retarding Ignition decreases NOX formation at the expense of reduced
efficiency, thus Increasing fuel consumption. Emissions of HC and CO are
generally insensitive to retard except in the extreme case where misfiring
can occur. That is, the higher exhaust temperatures, which tend to improve
the oxidation of any remaining unburned fuel or carbon monoxide, offset the
effect of shorter residence times in the cylinder. Smoke in dlesel engines,
however, increases rapidly after moderate degrees of retard (2° to 6°).
Figure 4-26 presents the level of reduction demonstrated for a range
of engine types. Based on these data, the percent of NOX reduction per
degree of retard ranged from 1.2 to 6.9 for naturally aspirated or blower-
scavenged engines and from 0.6 to 8.5 for turbocharged engines. Actual
reductions due to retardation between 3 to 10 degrees ranged from 0.4 to 7.3
g/hp-hr for all engines. The effect of the control is to consistently reduce
the level of NOX produced, although the magnitude of the reduction can vary
considerably between engine types or within an engine category.
4-89
-------
I
IO
o
retard
9000
Brake specific 8000
fuel consumption
(Btu/hp-hr) 7000
6000
22
20
18
16
14
NOX level 1?
(g/hp-hr)
10
8
6
4
2
finalized reduction3
grees retard from
andard setting
controlled fuel consumption
rcent increase
r charging
rokes/cycl e
jine nunber
F
Gas
:...-' •••
A •
•
29.0
: i
27.9
^ , , *
I I \ t
• * ( 11
2.6 3.1 2.1 16.5 0.4 0.3 4.1 1.1 5.0 3.9
6 4 4 3 9 10 10 6 10 7
7127 6775 6760 7265 78406374 7300 8139 7500 6500
5.2 4.3 4.7 3.415.3 1.4 4.0 0.2 3.5 5.5
BS TC TC TC HA TC TC TC TC TC
1 2 64 65 44 4 20 28 45 46
jel
Dual
*
t *
%
i
i
8.8 1.7 6.3
843
7040 6340 6464
10.1 1.1 2.9
TC TC TC
2* M
*» *t
13 6 16
Diesel
- * I
A
Ji- I * J,
1! 1 ( *
1
7.0 4.4 5.7 7.1 3.8 4.4 2.7 4.9 0.73.83.9
4 10 4 344482 16 4
— 6679 6734 6741 6677 6643 6366 7147 6962 6870 6643
— 7.4 1.9 2.5 0.8 2.6 1.5 6.4 1.618.73.6
BS TC TC TC TC TC TC TC TC TC TC
17 12 18 74 5 9 15 26 55 60 8
NOx reduction/degrees retard
Figure 4-26. Effect of retard on NOX emissions and fuel consumption.
-------
Several manufacturers have investigated the effect of different
amounts of retard on NOX emissions and fuel consumption for diesel, dual
fuel, and gas engines. Their results are shown in Figures 4-27 and 4-28.
Figure 4-26 shows that the effect of retarding fuel injection on NOX levels
and fuel consumption is similar for different diesel engine types. That is,
4 degrees of retard reduces NOX from 22 to 30 percent (i.e., 26 +4 percent),
and 8 degrees reduces NOX 39 to 44 percent. Note that the NOX reduction per
degree of retard decreases for increasing levels of retard. In contrast,
fuel penalties increase at a greater rate with increasing retard. Thus,
4 degrees of retard causes a 2-percent fuel penalty, 8 degrees a 6 percent
penalty, and 12 degrees a 12-percent penalty. Therefore, maintenance and
durability considerations aside, there are diminishing benefits to
retarding diesels beyond a certain point, because increases in fuel
consumption exceed decreases in NOX levels.
Figure 4-28 shows similar results for gas and dual-fuel engines,
although the data are more scattered. In general, ignition retard for gas
engines is not as effective in reducing NOX levels as it is for diesel and
dual fuel engines. For example, 4 degrees of ignition retard gives about a
15 percent NOX reduction in gas engines as compared to around 25 percent for
diesel and dual fuel units. Note that the amount of NOX reduction remains
constant after a certain point for the two naturally aspirated engines, but
fuel consumption continues to increase rapidly. In addition, there are
practical limits of ignition retard for all gas engines. Spark-ignited
engines are more sensitive to ignition timing and, therefore, misfire and
exhibit poor transient performance when the ignition timing is not very close
to the design point.
4-91
-------
1.20
1.10
1.00
#77
8 10
Retard, •
12 14 16
1.0
0.8
* 0.6
0.4
0.2
'#55
6 8 10
Retard, °
| Same engine model
16
Figure 4-27. Effect of different amounts of retard on NOX
emissions and fuel consumption for diesel enqines
4-92
-------
1.10
1.05
1.00
1.
0.2
G Df_
2-BS O
2-TC D (1
4-NA &
4-TC ^ V
G - Gas
DF - Dual Fuel
6 8 10
"Retard
7#28
12 14 16
44
16
Figure 4-28.
Effect of different amounts of retard on
NOX emissions and fuel consumption for gas
and dual fuel engines.
4-93
-------
Special equipment is not required for injection ignition retard as
it involves only an adjustment of the engine spark or injection pump timing.
Typically, the nominal setting of the ignition time is fixed by means of hardware
items, such as crankshafts. Means are then provided for adjustments, or fine
tuning, about this nominal value to compensate for variations in altitude,
fuel, engine wear, etc. Manufacturers usually perform this fine tuning service
during the production run-in of the engine, but the adjustments also can be
made by the operator. This typically occurs every 10,000 hours in the course
of normal maintenance, but the setting is actually verified or corrected weekly.
As stated earlier, peak cylinder temperatures and pressures are
lowered by retard, and, hence, the thermal and structural loadings are
lowered. However, the delayed combustion causes higher exhaust temperatures,
which may lead to rapid deterioration of the exhaust valves if the exhaust
temperatures exceed the design limits of the valve material. According to a
manufacturer of gas-fueled engines, the values in their current production
engines can withstand temperatures up to 1300°F, and the turbochargers are
limited to 1200°F (these two temperature limits are not inconsistent because
the exhaust gas cools between the cylinder exhaust and the turbocharger
inlet)(80). Current cylinder exhaust temperatures range from 900°F to
1250°F. Nevertheless, one manufacturer determined that 4° retard of ignition
in a dual-fuel engine caused a 25-percent reduction in the maintenance life
of his current valve material(81). Another manufacturer reported that his
naturally aspirated SI engines are presently operating near their exhaust
material limits (1300°F) at rated load conditions. Data from one engine
showed that 10 degrees of ignition retard caused the exhaust temperature to
increase from 1263° to 1370°F. (NOX emissions were reduced 17 percent.)
Therefore, the application of retard to meet standards of performance may
4-94
-------
require more frequent engine maintenance or greater Initial cost for higher
temperature exhaust material.
4.4.3 A1r-to-Fuel Ratio Changes
The air-to-fuel ratio is defined as the mass flowrate of air Ingested
by the engine divided by the mass flowrate of fuel consumed. This ratio 1s
termed sto1ch1onetric 1f precisely enough oxygen 1s present in the mixture to
completely oxidize the fuel. When the ratio is greater than stoichlometric,
excess air (oxygen) 1s present, and the mixture is referred to as lean.
Conversely, a lower than sto1ch1ometr1c ratio 1s commonly called fuel rich,
or simply rich, because more fuel 1s present 1n the mixture than can be
completely burned.
The maximum NOX and minimum HC and CO emissions will generally occur
at an air-to-fuel ratio slightly leaner than stoichlometric. Although
maximum flame temperatures occur at less than stoichlometric ratios, maximum
NOX levels do not occur until lean A/F ratios when oxygen availability is
Increased. Perfect mixing of the air and the fuel never occurs in existing
engines; therefore, some excess air 1s necessary for complete combustion and
minimum HC and CO emissions. These relationships are shown 1n Figure
4-29(82) for a gasoline-fueled automobile engine. Similar curves apply to
dlesel- and gas-fired units, with different peak levels for the various curves
and shifts in the air-to-fuel ratio that correspond to peak NOX generation.
When the engine 1s operated rich, HC emissions rise sharply because the
available oxygen is no longer sufficient for complete combustion of the fuel.
The lack of oxygen for combustion also means 1t is not present for NOX
formation and so, despite the high cylinder teiiperatures, NOX formation will
drop sharply at increasingly rich mixtures.
4-95
-------
Relative emission levels
o>
(/>
3
n> o
n>
CD -S
Q>
n> o
ro* o
3 n>
o 3
n> -••
w
OO in
rv) -••
— o
. 3
A-10217
-------
If the ratio is varied in the lean direction, the oxygen availability
will increase but so will the capability of the air and products of
combustion mixture to absorb heat. Consequently, the peak temperature will
fall, resulting in lower NOX formation rates. HC emissions will rise with
leaner mixtures due to the lower temperatures which result in increased
quenching along the cylinder walls. However, since this effect will be
counterbalanced, in part, by the leaner mixtures which occupy the quench
layer, the HC rise rate will not be as sharp as it is with increasingly rich
mixtures.
CO emissions, which are primarily a function of oxygen availability
and only secondarily of temperature, show a pronounced rise as the mixture
becomes richer than stoichiometric, but little variation as it becomes
leaner.
To understand the potential effect of air-to-fuel ratio adjustments on
emissions from a particular engine type, one must first examine the ratios at
which the engines normally operate. The most important engine characteristic
that determines these ratios is the air and fuel charging system.
In injection type engines, which include all diesel and many dual fuel
and gas varieties, the air-to-fuel ratio for each cylinder can be adjusted by
controlling the quantity of fuel that enters each cylinder. Therefore, these
engines can be operated in the lean region where combustion is most efficient
and fuel consumption is optimum. On the other hand, carbureted engines are
beset by large variations in cylinder to cylinder air-to-fuel ratios(83). Of
necessity they must operate near the stoichiometric ratio to insure that no
individual cylinder receives a charge which is too lean to ignite (i.e.,
exceeds the lean misfire limit). Furthermore, A/F ratios cannot be increased
4-97
-------
too far 1n turbocharged, spark-Ignited engines 1f severe misfiring and
detonation are to be avoided.
The most practical use of air-to-fuel ratio adjustment as a control
technique 1s to change the setting toward leaner operation, since
Increasingly rich mixture operation causes substantial penalties 1n both HC
emissions and fuel consumption. This technique can be Implemented best on
Injection type engines. Carbureted engines will require better control of
the air-to-fuel ratio between cylinders before they can operate at leaner A/F
ratios. In fact, some current carbureted engines, when adjusted to leaner
than normal air-to-fuel ratios (but still rich), tend to Increase their NOX
emissions because they are moving towards the peak of the NOX versus air-to-
fuel ratio curve. They are not able to go beyond that point to the lean
region without mlsf1r1ng(84).
The emission levels that can be achieved by air-to-fuel ratio
variation are shown 1n Figure 4-30. Based on these data, the percent
reduction 1n NOX emissions per percent of Increase 1n air-to-fuel ratio was
0.12 for the one blower-scavenged gas engine, 1.4 to 6.2 for two naturally
aspirated units, and 1.4 to 7.1 for the turbocharged gas and dual-fuel units.
Actual reductions in gas engines were about 4 to 6 g/hp-hr for all but the
blower-scavenged units and 2 to 3 g/hp-hr for the dual fuel units. The two
dlesel engines displayed smaller decreases in NOX emissions, 0.7 to 0.9, with
the leaner combustion.
Figure 4-31 shows the effect of changes 1n air-to-fuel ratio (A/F) on
NOX emissions and fuel consumption for dlesel, gas, and dual-fuel engines. In
general, small changes 1n A/F (approximately 10 percent) cause a large NOX
reduction (approximately 30 percent) with less than a 5-percent fuel penalty.
This 1s particularly true for natural gas and dual fuel engines which operate
4-98
-------
I
to
10
A/F
9000
Brake specific 8000
fuel consumption
(Btu/hp-hr) 7000
6000
22
20
18
16
NOX level 11
(g/hp-hr)
12
10
8
6
4
2
Normalized reduction3
A/F ratio, old/new
Uncontrolled fuel consumption
Percent increase
Air charging
Strokes/cycle
Engine number
Fuel
Gas
.
'
A A
A A
-
-
-
-
:' II Ji
1 1 1 1
fy *b -
V
•
^
A
1
+
f
^
f
0.12 4.46 — — 5.08 6.2 1.40 3.28 7.1
62.5/68.2 54.7/59.0 — — 32.0/33.4 17.6/16.5 21.5/17.0 32.5/34.7 16/14.1
7127 6775 6780 6660 6942 8180 7840 6374 7590
1-3 1.4 0.6 2.1 5.2 5.6 4.1 2.3 13.9
US TC TC TC TC M M TC TC
2 2 2 2 244 4 4
1 2 66 67* 69 21 44 4 29
aPercent NOX reduction/percent change In A/F.
bDerate 7X
Dual
A *
Mi
v
1.38 3.9 — 2.40
31/36 27.6/41.8 — 34.3/35.6
7040 6480 6340 6464
2.S 2.1 0.8 2.8
TC Tt TC TC
224 4
13 78 6 16
Diesel
t
A
J.
1
0.94 0.73
43.7/40.0 28.6/21 .1
6632 6870
1.1 9.3
TC TC .
2 4
77 60
Figure 4-30. Effect of A/F on NOX emissions and fuel consumption.
-------
1.20
1.10
£ 1.00
in
CD
u
LL.
l/)
• 0.90
1.00
0.80
ox
0.60
0.40
0.20
#60
i»69
G DF D
EJ
2-BS
2-TC
4-NA
G - Gas
DF - Dual Fuel
D - Diesel
I
\
70 80 90 1-0 1.10 1.20
(A/F)C/(A/F)U
Figure 4-31. Effect of A/F changes on NOx emissions
and fuel consumption.
4-100
-------
near sto1ch1ometric (ie., peak NOX formation). Note that engine numbers 29
and 60 are carbureted and, therefore, are designed to run on the rich side of
the peak on the NOX vs. A/F ratio curve. Hence, they must decrease their A/F
ratio (i.e., move to a richer setting) to achieve NOX reductions. As Figure
4-31 shows, decreased A/F causes rapid increases in fuel consumption.
Increasing the A/F, on the other hand, achieves significant NOX reductions
with more moderate increases in fuel consumption. Note that all of the
engines that increased A/F were turbocharged units which already operate at
lean (greater than stoichiometric) air-to-fuel ratios. Of course, the
maximum increase in A/F is limited by the onset of misfiring, as well as
marginal transient performance and poor fuel consumption.
In practice, leaner air-to-fuel ratios are obtained by either reducing
the fuel input (essentially derating) or by increasing the air input.
Increased airflow is accomplished by installing a turbocharger or replacing
an existing turbocharger with one that is designed to deliver more air. For
blower-scavenged units larger blowers can be installed, or the existing
blower can be operated at higher speeds. These changes, however, tend to be
offset by increased parasitic horsepower as well as higher discharge air
temperatures.
The mass flowrate of air into a cylinder (whose maximum volume is
fixed) can also be increased by raising the density of the incoming air
through cooling. Intercoolers are frequently placed between the turbocharger
exit and the inlet manifold to partially offset the temperature rise of the
air which occurs as it passes through the compressor. (Intercoolers are
required on large, turbocharged SI*.engines to prevent preignition.) Both
turbocharging and intercooling have other effects on emissions and will be
discussed in Sections 4.4.4 and 4.4.5 respectively.
4-101
-------
In the case of carbureted engines, increased air flowrates can be
obtained by a change in the shape of the venturi and fuel nozzles in the
carburetor(85). However, since current carburetor and intake manifold
systems do not distribute a uniform fuel-air mixture to all cylinders,
significant changes toward lean combustion may require the adoption of fuel
injection in place of carburetW86). In addition, increases in A/F ratio
beyond a certain amount reduce the engine's maximum load capability and some
derating is required.
4.4.4 Turbocharging
The exhaust gas stream from an 1C engine contains energy that is
normally lost when this stream is discharged into the atmosphere. Some of
this energy can be regained by passing the exhaust through a turbine, which
is mechanically coupled to a compressor and uses the energy extracted from
the exhaust to increase the pressure of the incoming air. This arrangement,
called turbocharging, is pictured schematically in Figure 4-32(87).
Turbocharging was originally introduced to overcome problems associated with
engine operation at high altitudes where the atmosphere is rarefied. Since
it increases the pressure of the Incoming air, a larger mass of air can be
Injected into the cylinder of a turbocharged engine than of a blower-
scavenged or naturally aspirated one. Hence, more fuel can be Injected into
the cylinder (the air-to-fuel ratio for optimum combustion is relatively
intensitive to pressure), thereby allowing an engine of given size and weight
to produce more power. Although turbochargers are normally designed to
increase an engine's output to approximately 1.5 times its original power,
they can be used to raise the engine's capacity to 2 to 3 times its naturally
aspirated value, provided the basic structure is sufficiently strong.
4-102
-------An error occurred while trying to OCR this image.
-------
Manufacturers frequently use turbochargers to make engines available that are
intermediate in power between existing nonturbocharged units(88). This
procedure is more common among manufacturers of medium-bore engines than
large-bore units, since most of the larger engines are already turbocharged.
Two manufacturers of the medium-bore engines have expressed concern that
emission standards which implicitly require the addition of turbochargers to
all engines will reduce the number of engine sizes available for purchase
because users will no longer have the option of buying engines whose power
rating falls between the rating of adjacent turbocharged models(89). Such a
change could indirectly increase the cost to a user by forcing him to buy a
higher horsepower engine than he would have purchased otherwise.
In addition to increasing the air flowrate, turbocharging also
increases the inlet air temperature due to the temperature rise associated
with compression. This increase, however, is generally offset by the
addition of an intercooler, or heat exchanger placed between the compressor
and the inlet manifold (see next subsection for more details). As seen in
Figure 4-33, three of the four sets of engines for which comparative data on
turbocharged and nonturbocharged are available did not show a rise in the
inlet manifold air temperature as a result of the turbocharger. From the
standpoint of emissions, and in particular NOX formation, two combustion
parameters may change when a turbocharger is added: temperature and air-to-
fuel ratio. If the inlet air temperature rises, the peak cylinder
temperature will be correspondingly higher; hence NOX formation will
increase. Conversely, if the temperature is reduced, NOX formation will be
lower. Changes in the air-to-fuel ratio will also change NOX emissions,
depending on the location of the air-to-fuel ratio of the nonturbocharged
unit relative to the peak in Figure 4-29 and the direction of the change. In
4-104
-------An error occurred while trying to OCR this image.
-------
addition, as the power output is increased, the brake specific emissions will
decrease. The net result, as can be seen in Figure 4-33, is a NOX reduction
ranging from 3.3 to 34.5 percent 1n the two-stroke engines for which data are
available. The largest reduction 1s obtained on the unit that runs at a
lower Inlet air temperature with the turbocharger than 1t does when blower
scavenged. The four-stroke set, engines 22/23, operates very near to
sto1ch1ometr1c, where leaning the air-to-fuel ratio can Increase NOX
formation. More Importantly, 1t was tested at a higher Inlet air temperature
with the turbocharger than without.
Since the combustion reactions for both CO and HC go to completion
more readily at higher cylinder temperatures and 1n the presence of leaner
air-to-fuel ratios, their emissions are reduced by turbocharging(9°).
Similarly, smoke emissions will also be reduced under these lean mixture
conditions^1).
The addition of turbocharging requires the installation of the
turbocoropressor discussed earlier and, depending on the strength of the
originally installed parts, may require the replacement of a number of other
engine components (piston rings, connecting rods, wrist pins and cylinder
heads) with higher strength parts. This may be necessary because
turbocharged engines operate with higher cylinder temperatures and pressures
and thus experience higher thermal and structural loads.
turbocharging is generally offered as an option to most current
naturally aspirated or blower-scavenged engines and, as such, 1s well
demonstrated and readily available. As an emission control technique it 1s
most effective if used in conjunction with some means of cooling and
compressed air, such as the heat exchanger (intercooler) discussed in the
next section.
4-106
-------
The addition of the turbocharger (either with or without an
intercooler) Introduces an operational problem only 1f the engine must
respond to rapidly varying loads and speeds. Stationary engines used 1n
portable air compressors could face such demands, but 1t 1s unlikely that the
large engines for which standards are now being developed would experience a
problem. The problem 1s one of smoke generation due to rich combustion and
arises during acceleration because the fuel flowrate increases much more
rapidly than the air flowrate. The air flowrate depends upon the power
delivered to the turbocompressor, which depends, in turn, on the power that
the turbine can extract from the exhaust gases. The energy carried by these
gases increases only when the engine output increases. Hence there is a
built-in lag in the system which results 1n insufficient air for a short time
after the fuel flowrate 1s increased. Some manufacturers have solved this
problem by careful control of the fuel rate or by using some of the
compressed air from the storage tank (if the engine is used to drive an air
compressor)(92). Another attempt at a solution 1s being tried by one
manufacturer of turbochargers, who is currently developing a turbine with two
kinds of blades. Some of these extract energy efficiently at low pressures
and flowrates while others funtion best at the higher flow conditions
corresponding to rated engine operation^).
Therefore, turbochargers should be considered applicable to all large
engines, either because the application does not involve rapid load
fluctuations or because of compensatory designs.
4.4.5 Reduced Manifold Air Temperature
The installation of a heat exchanger (commonly referred to as an
intercooler or after cooler) between the turbocharger and the intake manifold
4-107
-------
(see Figure 4-32) normally accompanies the use of turbocharging in engines
over 500 hp. The function of the intercooler is to lower the temperature of
the intake air after it has been heated by turbocharging (see Section 5.3.5).
Decreasing the temperature increases the density, thereby allowing more air
to enter the cylinder. This in turn permits higher fuel flowrates and
consequently higher power output. Decreasing the inlet temperatures also has
the secondary effect of reducing the peak combustion temperature and, hence,
the NOX emissions.
Decreasing manifold air temperature will result in an increase in HC
and CO emissions because their respective reduction reactions now proceed at
a slower rate. However, some diesel engines display a contradictory trend
for CO. Presumably the temperature is still high enough to allow the
reaction to proceed to completion, and the increase in power, then, results
in a decrease in brake specific emissions.
An intercooler is either provided with, or available as an option on,
most current turbocharged engine models. In fact, an intercooler must be
used when a carbureted, natural gas engine is turbocharged to prevent the hot
air from detonating the air-fuel mixture prior to its entry into the
cylinder(9*). Additional hardware requirements include the heat exchanger,
a circulation device for the cooling medium (either a fan for cooling air, an
enlarged water pump for engine coolant water, or a new source of cooling),
and a control mechanism to regulate the degree of cooling.
It is also possible to use increased cooling as an emission control in
engines that are already turbocharged and intercooled. Depending on the
level of cooling necessary and the availability of a suitable low temperature
heat sink, this can amount to installation of a larger heat exchanger or the
addition of an entire heat exchanger system using air, well or municipal
4-108
-------
water, or a cooling tower. Self-contained radiators, similar to those found
on automobiles, are the most commonly used cooling system for large engines.
Large radiators could deliver air to the Inlet manifold that 1s no more than
15°F hotter than the ambient air. Thus, 1n a hot area where daytime air
temperatures frequently reach 100°F, the Inlet manifold air temperature could
not be reduced below about 115°F. Cooling towers would have to be used 1n
such a location to reach lower Inlet air temperatures unless a large supply
of cold water were available for once-through cooling. Under typical
conditions for a hot and humid location, the turbocharged Inlet air
temperature of 130°F could be cooled to nearly 100°F with a cooling tower.&
The data 1n Figure 4-34 are all for an Intercooler exit temperature of
100°F. The normalized NOX reduction (percent NOX reduction per degree
Fahrenheit temperature reduction) was used to Interpolate or extrapolate the
available data, where necessary, to the 100°F temperature for ease of
comparison. This normalized NOX reduction value ranged from 0.9 to 1.2 for
gas engines, 0.6 to 2,2 for dual-fuel units, and from 0.1 to 0.3 for dlesel
engines. In general, manifold cooling was most effective on gas engines,
particularly two-stroke models. This 1s probably due to the large quantities
of scavenging air used by these units. When this air 1s cooled, 1t not only
reduces peak combustion temperatures directly but also Indirectly by cooling
the cylinder walls.
Figure 4-35 shows the effect of lowering the manifold air temperature
on NOX emissions and fuel consumption. The results vary and depend on the
-'Based on 100°F ambient air temperature, 90-percent humidity with a
resultant minimum Intercooler water temperature of 97°F, and typical heat
exchanger effectiveness.
4-109
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
engine type. In general, reduction of 10 to 40 percent can be achieved with
a 25-percent decrease in the baseline manifold air temperature. If one
assumes a baseline temperature of 130°F (a common aftercooler design exhaust
temperature), then a 25-percent reduction brings the manifold temperature to
100°F. As noted above, many engines operate in hot climates (e.g., gas
pipeline compressor units and electric generating systems in midwestern
utilities). Such engines would require large quantities of cold water or
refrigeration (which consumes so much energy that derating may be equally
effective) to reduce the inlet manifold air temperature further.
4.4.6 Exhaust Gas Recirculation
Exhaust gas recirculation (EGR) systems function by replacing some of
the incoming excess air with exhaust gases. The purpose of this is to reduce
peak temperature by providing a mass which can absorb some of the heat
released during combustion. The value of using exhaust gas is that it serves
as a heat absorber without making available as much additional oxygen to the
nitrogen atoms as turbocharging or operating with a high air-to-fuel ratio
would. EGR can be accomplished by either reintroducing exhaust gases into
the intake manifold (external EGR) or by restricting the exit of gases that
would normally be exhausted from the cylinder (internal EGR). In both cases
the effects on pollutant emission are quantitatively the same. Externally
recirculated gases can also be cooled before they are reintreduced into the
cylinder. Since these cooled gases are able to absorb more heat than
unrecirculated exhaust gas would, the peak combustion temperature is lower and
less NOX is generated.
A previous study showed that cooled, external EGR was effective in
reducing NOX from a large-bore, two-stroke blower-scavenged test engine^).
4-112
-------
Although this engine was a two cylinder, older version of a locomotive
diesel, trends observed on it should be similar to those that would be
observed on modern production engines of this model. The application of 20
percent cooler EGR at rated conditions resulted in a 55 percent NOX reduction
and an increase in smoke to 17 percent opacity.- HC emissions were unchanged
and CO emissions were increased 72 percent. By comparison, 20 percent hot
EGR resulted in NOX reductions of 51 percent at rated conditions and smoke
increased to 27.5 percent opacity. HC emissions were reduced 17 percent and
CO emissions were increased 167 percent.
These same trends have been reported in tests of cooled EGR on truck
sized engines which achieved NOX reductions of 60 percent using 15 percent
cooled EGR(96). Smoke increased 300 percent and HC remained relatively
independent of EGR rate. CO also increased at EGR rates greater than 30
percent.
Since the oxidation of CO and HC depend upon the availability of
excess air and elevated temperatures, one might expect the reduction of both
oxygen and temperature by EGR to lead to increased emissions of these two
pollutants. However, EGR traps or recirculates some of the unburned, quench
layer hydrocarbons in the cylinder. Since these recirculated remnants of the
fuel are combusted during the next cylinder firing, HC emissions frequently
decrease when EGR is used. Smoke levels increase with EGR due to the
reduction in excess air.
-'Twenty percent EGR means that 20 percent of the exhaust gas is returned
to the engine intake manifold.
4-113
-------
As mentioned earlier, EGR systems can be grouped into two basic categories,
internal and external. The application and associated operational problems
of each will be discussed in the following sections.
Internal EGR
With Internal EGR, the exhaust products are retained in the cylinder by
increasing the exhaust back pressure, retarding the valve timing or by
reducing the scavenging airflow in two-stroke engines. The exhaust back
pressure can be increased by inserting a restriction in the exhaust gas flow
to raise the pressure in the exhaust manifold. This prevents the latter
portion of the exhaust charge, which normally exits at the lowest pressure,
from leaving the cylinder before the beginning of the next cycle.
Retarded valve timing achieves the same effect by prematurely closing
s
the exhaust valves and opening the intake valves. The early closing of the
exhaust valve prevents a complete purge of the combustion gases, while the
quickened opening of the inlet valve allows the trapped gases to enter the
intake manifold and mix with the air or air-fuel charge before reentering the
cylinder. The valve timing in all engine types can be increased by changing
the camshaft timing (position of the camshaft, which controls the movement of
the valves relative to the crankshaft and, hence, to the pistonjv9?).
Reducing the scavenging air pressure in two-stroke engines results in
an incomplete removal of the exhaust gases by the fresh incoming charge. The
scavenging air comes fron the air box located below the piston and is driven
through the combustion volume by the blower. The scavenging air pressure can
be reduced by bleeding off sane of the air through a valve located in the
side of the air box. This results in incomplete scavenging of the exhaust
4-114
-------
products which are then trapped 1n the cylinder and dilute the charge during
the next cycle.
Internal EGR 1s available for both two- and four-stroke engines,
either naturally aspirated or turbocharged. It 1s usable to a greater extent
with turbocharged models due to their leaner operation.
Increased valve overlap has not been adopted for a variety of reasons.
First, severe fuel penalties accompany the use of this technique. In
addition the Inlet valves can become fouled due to the flow of hot dirty
exhaust products over them, and the presence of extra Inert mass, which
contains cold fuel during startup, can make starting difficult^98).
Presumably these two problems could be overcome by engineering modifications.
A more difficult problem to solve 1s that of Increased smoke generation 1n
dies el engines due to the reduced excess air available during combustion.
Unlike engines with external EGR or air box bleed, those with valve overlap
cannot reduce the redrculatlon rate as the load Increases. Therefore, they
tend to emit smoke at high power settings. Derating the engine to the point
where smoke becomes acceptable appears to be the only solution.
Figure 4-36 shows the effects of Internal EGR on a naturally aspirated
gas engine, a blower-scavenged spark (gas) and a turbocharged diesel. For
these three engines Internal EGR reduced NOX emissions from 4 to 37 percent.
External EGR
External EGR is accomplished by ducting some of the exhaust gas from
the exhaust system to the intake manifold. Its effectiveness can be
increased by cooling the hot exhaust gases before they are introduced into
the inlet manifold. Generally this is done by attaching cooling fins to the
recirculating duct to transfer heat from the gases to the environment. The
4-115
-------An error occurred while trying to OCR this image.
-------
data available from tests of external EGR on large engines are Included 1n
Figure 4-36, together with the results obtained with Internal EGR. The
results from three tests on gas, dual fuel, and dlesel turbocharged
models show reductions varying from 25 to 34 percent. These reductions were
obtained with exhaust gas redrculatlon rates of 6.5 to 12.0 percent. The
effect of applying different amounts of EGR on NOX emissions and fuel
consumption 1s shown 1n Figure 4-37. NOX reductions ranged from 10 to 22
percent with 6 percent EGR. In general, fuel consumption remains unchanged
for EGR rates less than 12 percent.
Because the available oxygen decreases with Increasing load (I.e.,
Increasing fuel flow but nearly constant airflow), 1t 1s most efficient to
vary the EGR rate with load. Therefore, 1f the engine 1s expected to be
faced with varying load demands, 1t should have a control valve 1n the EGR
system regulated by a load sensor. Such a system 1s pictured schematically
1n Figure 4-38. Air-to-fuel ratios are typically close to sto1ch1cmetr1c on
dlesel truck engines when they operate at rated power. Therefore, external
EGR systems that have been Installed on truck engines to meet California
standards for heavy duty vehicles have set their proportional EGR controls to
shut the redrculatlng passage when the engine 1s run at full throttle.
Large-bore engines, on the other hand, run much leaner, even at rated
conditions; therefore, they could benefit from EGR though they normally
operate at rated power.
The primary durability consideration for external EGR systens,
especially when applied to dlesel engines, 1s the accumulation of solid
exhaust products 1n the redrculatlng system. When EGR 1s applied to
naturally aspirated engines, these deposits build up 1n the ducts, on any
valves used to control the redrculatlon rate, and possibly on the Intake
4-117
-------An error occurred while trying to OCR this image.
-------
Load
sensor
Muffler
Hot exhaust
gas
Control/cutoff
valve
Air or liquid
cooled heat
exchanger
Turbocharger
Cooled
exhaust
gas
Inlet air
Figure 4-38. Simplified schematic of EGR system for turbocharqed
intercooled engine.
4-119
-------
valves. However, when external EGR is used in conjunction with a
turbocharged and intercooled engine, the problems are magnified. Since the
inlet charge, after the compressor, is at a higher pressure than the exhaust,
one either has to provide a separate compressor and inter cooler for the
recirculating stream or one has to mix the recirculated gases with the
incoming air before they pass through the turbocharger. Both approaches have
similar problems, namely fouling of the compressor blades and the heat
exchanger surface^0^'^^). If the compressor is designed to operate close
to its optimum condition, its performance is very sensitive to the shape of
the blades, and these would be changed slightly by deposit build-up.
Similarly, the effectiveness of heat exchangers is reduced greatly by any
slight coating on the surface^102). In fact, moderate deposits can make the
heat exchanger virtually useless. Such deposition problems do not
necessarily preclude the use of EGR on these types of engines, but they would
require significantly increased maintenance by the user.
Two studies have been conducted which evaluated the deterioration of
EGR systems in heavy duty diesel and gasoline engines(103»104). The external
EGR system evaluated in a diesel engine showed no increase in NOX emissions
over 1000 hours of use. In contrast, the external EGR system of a gasoline
engine fouled the carburetor and EGR passages after 700 hours of use. The
significant difference between these systems was the relatively high
temperature of the recycled exhaust gas (approximately 600°F) in the diesel,
which prevented the condensation of exhaust products in the air charging
system. However some condensation will always occur, and, therefore,
trouble-free operation of any EGR system will involve increased maintenance.
One engine manufacturer has also expressed concern over the con-
tamination of lubrication oil in EGR equipped large engines(105). This
4-120
-------
problem has not occurred in current production automotive EGR systems,
presumably because of their lower cylinder pressures and their use of a fuel
with a lesser tendency to form sludge (sediment in engine oil).
4.4.7 Water Induction
The addition of water to the fuel-air charge has much the sane effect
as EGR. It increases the inert mass of the charge and results in lower NOX
production through decreased peak combustion temperatures. Although it is
easier to insure uniform mixing if steam is used, the water can be more
effective if it is injected as a liquid. Energy extracted from the products
of combustion to vaporize the water causes an additional reduction in the
peak cylinder temperature. Furthermore, steam is usually not readily
available, and no test data have been reported on the use of steam.
Therefore, the remainder of this discussion is restricted to water in the
liquid phase.
Nitrogen oxide reductions that have been achieved with water induction
are shown in Figure 4-39. The effectiveness of the technique appears to
correlate with an absence of excess air. Since the purpose of adding water
is to reduce the peak temperature by increasing the thennal mass in the
cylinder (amount of material that is heated by the energy released during
combustion), this technique should be more effective in a low excess air
system than in one with much excess air and, hence, much thermal mass. As
shown in Figure 4-39, the gas engines, which were the most responsive to the
addition of water, operated at lower air-to-fuel ratios than did the diesel
or dual-fuel units. Presumably the excess air in the latter two types of
engines absorbs all the heat that can be transferred to a fluid in the short
time between ignition and peak temperature. The combustion reactions for HC
4-121
-------An error occurred while trying to OCR this image.
-------
are inhibited by the lower temperatures, and, therefore, HC emissions
increase with water induction^**), increased quench layer effects at these
lower cylinder temperatures may also contribute to the higher HC emission
rate. CO emissions are largely unaffected by water Induction.
Water induction can be accomplished in a variety of ways ranging from
introducing it with the intake air to injecting it directly into the
cylinder. These systems vary in both complexity and effectiveness. The
effectiveness of the system is strongly a function of the degree of
atomization and mixing of the water within the combustion charge. The best
systems are those which provide for induction prior to the turbocharger, if
present, or spray injection into the inlet manifold or cylinder directly. To
control the water flow at various loads, a water pumping and metering system
(similar to the fuel system) must be provided.
The reported effectiveness of this control method in reducing NOX
emissions depends almost linearly on the rate at which water is added(107).
As shown in Figure 4-40 reasonable reductions (e.g., more than 30 percent)
for large-bore engines are obtained only if the water flowrate is at least
one-half the fuel flowrate, and significant reductions require that the rate
of water addition approximately equal the fuel flowrateUOS). Data with
water-to-fuel ratios greater than 1.0 are available only for high speed,
truck-si zed engines, but the same trends should apply to large-bore engines
as well. As Figure 4-4l(109) illustrates, the effectiveness of water
induction (percent NOX reduction per mass of water) decreases as the water
flowrate is increased beyond 1.0. This figure also confirms the earlier
statement that the effectiveness of water induction is inversely related to
the amount of excess air present; greater NOX reductions are achieved (at any
4-123
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
given water-to-fuel ratio) with the engine running at higher load and,
consequently, lower air-to-fuel ratio.
Since many engines are used 1n remote, and frequently arid, locations,
water availability may be a problem. Assuming the use of a 1.0 water-to-fuel
ratio, a typical large-bore 600-hp engine would consume water at the rate of
0.46 gallons per minute (1.75 liters per minute), or about 660 gallons per
day in continuous operation, and a 4300-hp unit at the rate of 2.9 gallons
per minute (11.0 liters per minute), or about 4200 gallons per day in
continuous operation.
All of the large-bore engine manufacturers who provided results from
tests with water induction (General Motors-END, Ingersoll-Rand, Cooper-
Bessemer, and White Superior) reported serious concern about the adverse
effects of this technique on engine durability. This concern was based on
observations of water in the crankcase, which contaminated the lubricating
oil(110)t ^d on rapid build-up of mineral scale around the valves, water
Injection nozzles and other components (e.g., the turbocharger) through which
the water flowed^11). The one independent researcher who applied this
technique to a large-bore engine experienced the same oil contamination^2'.
Only one long duration test (23 hours) of the use of water induction in a
large-bore engine has been conducted to date, and this program used untreated
water (400 ppm calcium carbonate)(113). All other tests were short —
usually less than 2 hours of engine running time.
4.4.8 Combustion Chamber Geometries
Combustion chamber redesign is the control technique having the
greatest potential for reducing NOX emissions from large-bore engines with
little or no loss in efficiency. It is probably also the technique requiring
4-126
-------
the greatest amount of research and development, and could result in
manufacturers changing their engine designs significantly, incurring
substantial retooling costs. Several years would be required by manu
facturers to perform research and development work and to retool to build
engines having redesigned combustion chambers before they could be used to
meet a standard of performance for new stationary engines, Some combustion
chamber designs which might be applied to reduce NOX emissions without
significant Increases 1n emissions of HC, CO and smoke, or fuel consumption
will now be discussed.
The combustion process can be Improved by redesigning chamber
geometries to increase turbulence, as in high swirl engines(114), by staging
as in engines with stratified charge^1*), or piston head cavit1es(116), or
by a combination of both, as with "squish I1ps"(117). High swirl systems use
modifications to the design of the cylinder and Inlet ports to produce
circulatory air motion in the cylinder during combustion. Since these flow
patterns are conducive to good air-fuel mixing and, hence, efficient
combustion, delayed ignition timing can be used to reduce peak temperatures
below those necessary for NO formation with less production of unburned HC
and smoke. Unfortunately, no data are available to compare the NOX emissions
of a low- or medium-swirl engine with those fron a high-swirl unit when both
are retarded as far as possible, I.e., until smoke becomes excessive. The
NOX emissions from the high-swirl engine should be higher than those from the
low- or medium-swirl units with corresponding ignition timing and could very
well be harder to control; the improved mixing promotes rapid, early
combustion and hence high temperatures for a long period of time(118).
All of the staged combustion designs (precombustion chambers, the M-
systems, and stratified charge) as shown in Figures 4-42 to 4-44(119-122)
4-127
-------
Figure 4-42. Schematic of a current production precombustion
chamber for a diesel engine (Reference 119).
4-128
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
Main
chamber
Prechamber
Figure 4-44(b). Schematic of a prototype stratified charge gasoline
engine (General Motors Corp. design — Reference 122),
4-131
-------
operate in basically the same manner. The fuel charge is introduced into a
cavity as a rich mixture and ignited. Since it burns in the absence of
excess oxygen, NO formation is inhibited. This burning mixture then expands
into the main chamber where it mixes with additional air at reduced
temperatures. Therefore combustion is completed in an overall lean mixture
at temperatures which are adequate for combustion but below those required
for NO formation. The stratified charge concept has been applied only to
gasoline engines, and no reports exist that document attempts to use any of
the other designs on the very large engines. However, there is no reason to
believe that the application of the basic combustion principles are not
applicable to the very large engines. The squish lip (Figure 4-45(123))
design appears to reduce emissions both by aerodynamic effects and by staged
combustion(124). The lip-shaped cavity in the piston head increases
turbulent mixing and also creates a vertical flow pattern which recirculates
burnt gases through the combustion zone within the cavity. Thus it
incorporates a form of internal EGR'^5).
The emission levels that have been reached by engines with these
various combustion chamber geometries are shown in Figure 4-46(a) (NOX
emissions) and Figure 4-46(b) (HC and CO emissions). Included are results
from some truck-sized engines and other commercially available units, such
as engines with precombustion chambers. Based on these data, NOX levels ranged
from 2.7 to 7.5 g/hp-hr and CO levels from 1.3 to 4.1 g/hp-hr. HC emissions
were clustered between 0.2 and 0.6 g/hp-hr for all of the engines. Since
these configurations result in approximately 60 percent lower NOX levels than
observed in open chamber engines, and since no large-bore units are currently
built with these designs, they can be considered a potential "control technique"
for the large engines. Although members of the DEMA have expressed concern
4-132
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
CO
01
-C
1
C Q.
O-d
at en
in
UJ 0»
i
4
3
2
1
0
Control
A1r Charging
Horsepower
Engine Number3
_ B _
- -
- Q
Q
Q
— CO = 50 ppm —
HC = 100 ppm
& A
A
PC PC PC PC VT
TC TC TC TC TC
750 850 970 850 1410
30 31 32 33b 34
A HC
Q CO
PC — Precombustion chamber
VT — Variable throat chamber
Footnote
a. Engine 34 is large bore; all others are
medium bore.
b. 13 mode composite; all other data at rated
conditions.
Figure 4-46(b). HC and CO emissions from four-stroke diesel engines with modified
combustion chambers.
-------
about their ability to adapt such designs to their engines^26), the problems
appear to be mainly of the kind that can be solved by application of current
technology rather than of the kind that are technical barriers. In fact, one
manufacturer of squish lip engines has stated that he sees no reason why
squish lips could not be used on pistons of large-bore engines^2?).
Furthermore, the open literature contains at least one report of a large-bore
engine that has been modified to incorporate something akin to one of these
staged combustion designs, namely a variable geometry precombustion chamber
(Engine No. 34 in Figure 4-46). However, the report presented results on a
full-scale engine at rated loads only under laboratory conditions^28).
It should be noted that most diesel engine models manufactured by the
Caterpillar Tractor Co. have precombustion chambers. However, they consume
5 to 8 percent more fuel than similar size direct injection engines, when
both, are designed by the same staff (i.e., equal engineering sophistication)
to operate at maximum efficiency and to meet the same emission
standards'^). Precombustion chamber designs were introduced during a time
when fuel economy was a less significant factor than maintenance costs and
adaptability to a wide variety of fuels. The pressure rise rate is lower in
these engines than it is in the direct injection versions because the
combustion is extended in time and space. Therefore, mechanical stresses are
also less. Furthermore, it is easier to initiate combustion in the rich
mixture within the small volume of the precombustion chamber than in the lean
mixture within the large volume of the main chamber. Hence the fuel spray
pattern is not so critical, and, therefore, a wide variety of fuels can be
burned without having to adhere to a strict maintenance plan. These
advantages are now outweighed by the need to conserve fuel. Therefore,
Caterpiller is planning to phase out precombustion chamber engines within the
4-136
-------
next few years, and their competitors plan to meet more stringent standards
(e.g., California's 1975 and 1977 heavy duty vehicle standards) without
resorting to this design.
4.4.9 Catalytic Converters
The automotive industry has expended considerable research and
development during the past 5 years for exhaust catalysts to reduce HC and CO
emissions. These catalysts, however, are not effective in reducing NOX
emissions. Catalytic converters that reduce NOX emissions have been
developed and applied only recently to 1977 model year cars(13°). These
systems are called "three-way conversion (TWC) catalysts since they reduce
HC, CO, and NOX emissions simultaneously. These TWC catalysts closely
resemble the earlier oxidizing catalyst (precious metals applied to monoliths
or pellets) with the addition of rhodium which selectively reduces NOX to N2-
To operate properly, the air-to-fuel ratio of the automobile exhaust under
operating conditions must be controlled to a narrow range about stoichio
metric. The recent availability of durable ceramic exhaust gas oxygen
sensors to determine A/F conditions and control logic systems to regulate the
A/F conditions in the exhaust have made possible the implementation of this
control system U31). Figure 4-47 illustrates the approximate location of
this device on an engine.
These catalytic systems, however, are not applicable to large-bore
engines for two reasons: (1) HC and CO emissions from large-bore engines are
relatively low in comparison with vehicle level (typically 0.1 to 0.5
g/hp-hr nonmethane HC and 0.2 to 5.0 g/hp-hr CO for a large engine versus 1
to 6 g/hp-hr reactive HC and 30 to 90 g/hp-hr CO for a 100- to 200-hp
gasoline fueled engine) (2) A/F conditions in the exhaust of large engines
4-137
-------An error occurred while trying to OCR this image.
-------
are typically much greater than stoichicmetric. Consequently, other approaches
which can selectively reduce NOX emissions in the presence of excess oxygen
must be considered.
One approach is to inject ammonia, hydrogen, CO, or natural gas into
the exhaust to create the required reducing atmosphere(^2,133)t Ammonia
appears to work best because it has an affinity for the NO rather than the
excess oxygen. That is, one needs to add only slightly more ammonia than
would be required for stoichicmetric reaction with the NO, whereas with the
other three substances, additional quantities are required to deplete the
excess oxygen in the exhaust. Because of the high cost (see Section 8.2 1n
Chapter 8), even with ammonia, no engine manufacturer is investigating the
reduction of NOX in 1C engines using ammonia(^).
One research laboratory has experimented with a combination converter
scrubber exhaust treatment system for a two-stroke diesel engine commonly
used in busesU35)t NQX scrubbers are actually N02 scrubbers and are
ineffective for NO. Since NO is the primary constituent of NOX in the
exhaust, scrubbers cannot be used on engine exhaust unless the NO is
converted to N0£ first. Therefore, this laboratory converted the NO to N0£
by passing the exhaust through a catalytic converter which also oxidized the
CO, HC, and oxygenated HC (odorous gases) to CO^ and water. Since the
exhaust gas temperature was below the 700°F required by the converter for
proper operation, the exhaust was first passed through a heat exchanger to
raise its temperature. Finally the oxidized exhaust gas entered a molten
carbonate scrubber where the NOX was reduced by approximately 50 percent
(reduction from 17.35 g/hphr to 9.2 g/hp-hr). This reduction fell short of
the desired 70-percent reduction (to meet a 5 g/hp-hr standard). In
addition, the cost of the system was judged unacceptable due to the high
4-139
-------
price of the platinum catalyst ($15/lb) used. With a system requirement of
60 Ib of catalyst, the platinum catalyst, alone, would cost approximately
$900.
In conclusion, then, catalytic reduction of NOX has not, as of the
present, been demonstrated as a viable control technology for large-bore
stationary reciprocating engines.
4.4.10 Combinations of Controls
Several manufacturers have conducted experiments with more than one
control system at a time. The results of these tests are presented by fuel
in Figure 4-48. Figure 4-48(a) indicates that the controls are nearly
additive for diesels. Although a greater reduction is achieved on gas and
dual fuel engines by the combination of controls than by any one of the
individual component techniques when used alone, the reduction is less than
the sum of the effects from each technique.
Thus, for the large-bore diesel engine shown in Figure 4-48(a) the
maximum NOX reduction for a single control (retard) is 1.7 g/hp-hr. When all
the controls tested (retard, reduced inlet manifold air temperature, in
creased air-to-fuel ratio, and water induction) were applied simultaneously,
NOX was reduced 3.0 g/hp-hr. This is shown on the figure as an uninterrupted
downward arrow. For comparison, to the left of this, is a multiple arrow
line that represents the depiction in series of all the separate control effects,
as they were individually measured. The difference between the length of
these two lines is a measure of the relationship between the additive effects
of the controls when applied simultaneously and the sum of their individual
contributions. As noted earlier, for this case the controls do appear to
be additive.
4-140
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
c
o
4-> t-
I.& 75°°
o ^
O)
•2 6500
14
12
t.
o.
.c
"Si 10
>
ox 8
6
4
Control
Engine Number
Stroke-Conf i gurat
"1
—
—
1
—
—
U v 1
|.
V
r
t
of.
•\
z:
f
i
\
1 r
1 t
ex.
\
f
z
m
P
»
t
p
r
RAM RM RMA
1
2-BS
t A A t t
~_
LEGEND
Uninterrupted line — NOX
reduction achieved
I
1
f
i
QL
\
z:
f
*
>
s
Of
w
' -
r
-J
by simultaneous ap-
pl ication of several
control techniques.
Line with intermediate —
arrows — Additive combin-
ation of lines in-
dicating reductions
when each control ap-
plied individually.
—
—
f
RAM RM RMA
2
2-TC
sf
Nl
1
Figure 4-48(c). Additive effects of controls for large-bore gas engines.
-------
Figure 4-48(5) shows that the combination of retard and manifold air
temperature reduction is nearly additive in dual-fuel engines, but the
additional use of increased air-to-fuel ratio does not decrease NOX as much
below the levels obtained by the first two controls as one would expect from
the results with air-to-fuel ratio alone. Furthermore, the addition of water
induction to these three controls has no effect, whatsoever. If moderate N0¥
/\
reduction were required (e.g., 25 to 30 percent) for this engine, one would
change the air-to-fuel ratio; however, if a greater reduction were neces-
sary one would use the combined controls of retard, manifold air temperature
decrease, and air-to-fuel ratio modification.
Gas engines also do not respond additively to the simultaneous
application of several controls (Figure 4-48(c)). Here reduced manifold air
temperature decreased emissions from the blower-scavenged engine by 36
percent, whereas the application of reduced temperature plus retard reduced
emissions only 29 percent. Only the combination of the above two techniques
with increased air-to-fuel ratio could reduce emissions below the level
obtained by reduced manifold air temperature alone.
The turbocharged engine reached its lowest level (7.4 g/hp-hr) with an
individual control when the air-to-fuel ratio modification was used.
Although this engine was controlled to 4.5 g/hp-hr by the simultaneous
application of all three controls, if these controls would have had an
additive effect on the emissions, the engine would have emitted no NOV
A
whatsoever.
All but two combinations of controls resulted in a greater reduction
than any individual technique. In one gas engine the combinations without
increased air-to-fuel ratio were not as effective as increased air-to-fuel
ratio by itself. Recent tests on installed gas pipeline engines have shown
4-144
-------
that increased air-to-fuel ratio is the most significant source of NOX
reductions from large-bore gas engines(^6)t jn tne other case, the
combinations without increased air-to-fuel ratio were not as effective as
increased aftercooling by itself.
4.4.11 Summary of NOX Reductions
This subsection summarizes the data presented in Sections 4.4.1 through
4.4.10 for control techniques that have been shown to be technically viable
and effective in reducing NOX emissions. These techniques include: derate,
retard, air-to-fuel changes, manifold air temperature reductions, exhaust
gas recirculation (EGR), and combinations of these controls. With the exception
of EGR, all of these techniques consist of a change in an engine operating
condition.
Although water induction has been shown effective in reducing NOX
emissions, no data have been summarized, since this technique has serious
technical limitations (see Section 4.4.7). Furbhermore, modifications to
combustion chamber geometries (Section 4.4.8) and the application of
catalytic converters (Section 4.4.9) have been excluded from this summary
since only limited data for large-bore engines are available for these
control approaches. Nevertheless, both of these approaches have been used
successfully on smaller bore, mobile engines. In general, however, these
techniques will require considerable development before they can be considered
technically viable considering costs. Turbocharging (see Section 4.4.4) has
not been listed as a control approach since most large-bore engines are already
turbocharged, and engine users are showing increasing preference for turbocharged
units because of their lower initial costs
4-145
-------
A summary of the NOX reductions achieved by the candidate control
techniques discussed above is presented on Figure 4-49. The arrows on these
figures begin at the uncontrolled NOX level and end at the controlled level.
Dashed arrows indicate the same reduction after correction to standard
ambient condition (85°F, 75 grains/ft3 humidity) (see Section 4.2.1.) The
changes in fuel consumption after application of NOX controls are also shown
in Figure 4-49. Figure 4-50 shows uncontrolled NOX levels and the lowest
controlled level for each manufacturer of diesel, dual-fuel, and gas engines.
Note in Figure 4-50 that some of the engines show only uncontrolled emissions
since no control techniques had been applied.
A NOX control data summary appears in Table 4-13 based on Figures
4-65 through 4-70. This table also shows uncontrolled levels and the lowest
controlled level for each kind of control technique as applied to each engine
type and fuel combination. These data are summarized in a different way in
Table 4-14.
t Uncontrolled levels for dual-fuel engines are generally lower than
for diesels, which are generally lower than for gas engines
• Diesel, dual-fuel, and 2-TC gas engines more consistently achieve
lower NOX levels than do 4-TC gas engines. For example, at least
one turbocharged diesel, dual-fuel, or two-stroke gas engine is
capable of reaching a NOX level of 6 g/hp-hr. In comparison, the
lowest NOX level (corrected for ambients) demonstrated by any for
which data are available is 9 g/hp-hr.
• In general, the data show that two-stroke turbocharged gas and
diesel engines have reached lower levels than blower-scavenged or
naturally aspirated engines. No trend is apparent for the same
comparisons of four-stroke diesel and gas engines.
4-146
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
I
I—1
<£>
SuiBBry
9000
Brake specific ^^
Fuel consumption 800°
(Btu/hp-hr)
7000
6000
22
20
18
16
14
NOx level
(g/hp-hr) 12
10
8
6
4
2
Uncontrolled brake
specific fuel
consumption (BSFC)
Engine nunfcer
Stroke — configuration
4
-
A V
*
1
r
i i
A
M
1
1
ktT
1
r
27
1 1
Z-«S
^
i
*
*
7«n
1 71
T! ** T
t**Att*l UA* A^A*'
.
1! ! 1
:! ! , • i i
[|| |*i {
V V '
6775 UM 0H 72.B C7M fOi 39
2-TC
I
I
r*
5
C Ml
Bu
k
i
i
li-
15
11
i
V
n
'tt^t
i
«
IB S13i
1
-•
s
1
1
Ir
77f
4-1
',tl
11
,
* .
• am
m
k
1
f
^->
i
t,
p
f
* .. ^ **+
A AA A"
-
"
•
|| 1;
1 f
• r
-
4374 730D 74*0 7200 «l» 7S9D 7500 iiOO
*-TC
Figure 4-49(c). Sunmary of NOX reductions achieved by gas engines.
-------
s
Lowest Controlled
S unary
Brake specific 900°-
Fuel consiaptlon
(Btu/hp-hr) 8000
7000
6000
22
20
18
16
14
NOx level
(g/hp-hr) 12
10
8
6
4
2
Engine nunber
Stroke configuration
Percent BSFC increase
Percent N0x reduction
Control
.*.
(D
Q
v T T *
1
*
10 11 17 42 43
2-BS
2.6 4.1 0.0 --
23 20 28 — —
D D R - -
Q Uncontrolled BSFC
O Uncontrolled NO Darkened synbo
V Controlled HOX "Corrected for art
I. .*..
Q
T 9
1 ° 1 1
4
3 12 18 19 41 72 73 74 77
2-TC
7.4 0.0 -- — 3.2 — 2.5 4.5
- 43 23 - - 30 - 22 35
— R R AN — R RA
0°
O
*
24 35
4-m
-2.2 -
-- --
A j
s
V
IX o
• m O r» ®
T • T ^P
A 00*
i
5 8 ' ls » » 36 47 51 52 55 56 57 58 59 60 62 68 75
4-TC
-1.4 3.6 2.8 1.5 6.4 00 — ~ 11 18 7 2 3 0 0
Zl 16 18 16 40 4 - - - - 25 - - .. .. M.3 „ „ ..
"" D R R R D _--__«_ „ .. „
Figure 4-50(a). Uncontrolled and maximum controlled NOX levels for diesel engines.
-------
I
I—»
(71
Lowest controlled
Brake specific
Fuel consumption
(Btu/hp-hr)
NOx level
(g/hp-hr)
Engine Huaber
9000
8900
7000
6000
22
20
18
16
14
12
10
8
6
4
2
Percent BSFC increase
Percent NOX reduction
Control
Stroke — configuration
Dual fuel
- Q Q
B
©
9
: I
*
13 14 78
10.1 — 3.1
74 ~ 56
R — R,A
2-TC
W Q Cl m E* A
*-* ^ Note: All data corrected for a«bients.
& Controlled BSFC
D Uncontrolled BSFC
O Uncontrolled NOX
V Controlled MOX
1
©
© $
\f 1 f ^
|
*,
6 7 16 50 61 76
3.0 3.1 2.8 — 6.1 0.9
41 16 22 — 9 49
RH D A— D R,A
4-TC
Figure 4-50(b). Uncontrolled and maximum controlled NOX levels for dual-fuel engines.
-------
.£»
I
tn
ro
Lowest controlled
suHBry
9000
Brake specific
Fuel consumption 8000
(Btu/hp-hr)
7000
6000
22
20
18
16
NOX level 14
g/hp-hr
12
in
IU
8
6
4
2
Engine number
Percent BSFC increase
Percent NOX reduction
Control
Stroke — configuration
.
^
. e
.
.
-
-
.
<
•
-
^
•
1 7
2.6 4
75 61
0 D
2-BS
&
>
t
1
.1
0
t
ffl m 4
?
^
i
V
2 63 »
6.4 22.6 0.
66 48 21
Ml CS.A.R D
4
!> 1
<
N
)
r
\
\ 6
3 3.
r 4
R
ft
3 t
)
(
1
P
5 61
4 0.
} 2
A
2-TC
L
Uo [
)
r <
O
\
6 67 6<
6 0.0 14
t — 4<
A 0
L
b {
)
c
t \
} 7(
.4 5.
) 4<
EE
Jk
&
)
^
J
3
I
.0
<
f
\
^
2
4.
4,
IE
D
4
29.
)
I«
f
22 39 4
2 4.8 --- 4-
? 43 ~ 79
,A D.IM — D
4-NA
Gas
^
»
00
•
r
1 79 80
] — —
A Controlled BSFC
D Uncontrolled BSFC
O Uncontrolled NO Darkened syrib
unconnected f
V Controlled N0x art>ients
i
A T Q Q . ;
A A Q Q Q D
*— •
G>
• 0 0
• o -
t * -
T A •-
T t '•• 1
~ ~
^ v 0 *
-
-
1 r
4 20 23 25 28 29 37 38 40 45 46 48 49 53 54 81 82
2340176402139 -3555
31 42 38 61 6 88 — — — 55 27 — — -
A R 0* D R A - — — OR
4-TC
Figure 4-50(c). Uncontrolled and maximum controlled NOX levels for gas engines.
-------
TABLE 4-13. SUMMARY OF LOWEST NOX LEVELS (g/hr-hr) FROM CONTROLLED LARGE-BORE ENGINES
Fuel
Strokes/Cycle
Air Charging
Uncontrolled
Derate, D
Retard, R
A/F, A
MAT, M
Internal EGR, IE
External EGR, EE
R+M
R+A
D+M
A+M
R+M+A
Best Control
Diesel
2
BS
14.6-17.8
13.8
10.6
~
~
—
~
~
~
~
~
~
10.6
TC
8.5-14.7
9.0
5.6
—
—
—
6.6
—
7.0
~
8.0
~
5.6
4
NA
8.7-10.0
10.7
—
—
—
—
—
—
--
—
—
—
10.7
TC
5.2-15.2
7.1
3.0
6.3
6.2
7.1
—
6.2
2.6
14.2
6.3
8.7
2.5
Dual Fuel
2
TC
6.9-13.4
—
2.5
6.3
5.0
—
6.5
—
—
—
—
~
2.5
4
TC
4.4-12.1
7.3
8.5
5.8
7.7
~
—
7.0
—
—
--
5.5
3.8
Gas
2
BS
11.5-15.1
3.8
12.8
14.7
9.6
9.5
~
10.6
—
~
—
8.7
4.4
TC
8.6-17.7
4.4
11.8
7.4
8.3
—
5.8
8.1
9.8
—
—
4.6
4.4
4
NA
14.0-29.0
6.0
27.9
20.5
~
14.4
~
~
10.9
~
--
~
6.0
TC
7.8-22.3
8.6
10.0
2.0
12.0
—
—
~
—
--
—
~
2.0
cn
co
-------
TABLE 4-14. NOX EMISSIONS FROM VARIOUS GROUPINGS
OF ENGINES
Group
Air charging
Turbocharged
Nonturbocharged
Strokes/cycle
2
4
Fuel
Diesel
Dual Fuel
Gas
Emissions, g/hp-hr
Uncontrolled
4.4-22.3
8.7-29.0
6.9-17.8
4.4-29.0
5.2-17.8
4.4-13.4
7.8-29.0
Controlled
2.0-5.6
3.8-10.7
2.5-10.6
2.0-10.7
2.6-10.7
2.5-5.5
2.0-6.0
4-154
-------
Since emission levels from uncontrolled engines vary so widely, the
environmental impact depends more on the percentage NOX reduction it produces
than on absolute NOX level it achieves. Therefore, the data shown in Figure
4-49 have been recost in terms of percentage NOX reductions and summarized
in Table 4-15. A number of techniques, either alone or in combination, are
capable of achieving NOX reductions of 20 to 40 percent with moderate increases
in fuel consimption (less than 10 percent). In general, retard, air-to-fuel
ratio changes, or combinations of these two techniques achieved greater than
40 percent NOX reductions.
In addition to these techniques, combinations of retard and manifold
air cooling produced similar results for gas and dual-fuel engines. As shown
in Figure 4-49, a number of turbocharged diesel, dual-fuel and gas engines
achieved 20 to 40 percent NOX reductions with less than a 10-percent increase
in fuel consumption. Thus, several techniques can effectively reduce NOY
A
emissions from diesel large-bore. With the exception of controls applied to
dual fuel engines, however, maximum NOX reductions are accompanied by large
increases in fuel consumption.
Figure 4-51 presents an overview of the fuel penalties associated with
the various NOX control techniques. Based on the data shown in Figure 4-51,
large-bore diesel engines typically suffer a fuel consumption increase of
somewhat less than 1 percent for each g/hp-hr of NOX reduction. The major
deviations fron this average figure are manifold air temperature reduction,
which in some cases actually causes fuel consimption to decrease, and
derating which is accompanied by higher than average fuel penalties.
Retard, whether used alone or in combination A/F ratio changes, causes the
largest fuel penalties, but also achieves the lowest NOX levels.
4-155
-------
TABLE 4-15. ALTERNATIVE CONTROL TECHNIQUES - SUMMARY OF SELECTED CONTROLS
-o
I
Ul
01
% NOX Reduction
£20
20 and <40
>40
Maximum
reduction (%)
Diesel
Control
A, M, IE,
RMA, D, R,
EE, RM, RA,
AM
D, EE, RM,
AM, R, RA
R, RA
RA (65)
ABSFC, %
1 to 8
1 to 16
8 to 26
26
Dual Fuel
Control
EE, D, R,
A, M, RM,
RA, RMA
D, A, M,
RM, RMA, R,
RA
R, RA
R (73)
ABSFC, %
0 to 4
1 to 7
2 to 5
3
Gas
Control
D, R, A, M,
IE, EE, RM,
RA, RMA
R, M, IE,
EE, RM, D,
A, RA, RMA
D, A, RA,
RMA
D (90)
ABSFC, %
0 to 5
0 to 8
5 to 12
12
-------
9200
8800
8400
L 8000
-I.
-p.
I
en
7600
o.
£
3
7200
6800
6400
6000
#10
Diesel
Q 2-BS
Q 2-TC
& 4-NA
^7 4-TC
D Derate
R Retard
A Air-to-fuel ratio
M Manifold air temperature
IE Internal EGR
EE External EGR
CR Compression ratio
10 12 14
NOX level, g/hp-hr
16
18
20
22
24
Figure 4-51(a). Fuel consumption versus controlled NOX level for diesel enqi
engines.
-------
8ST-1?
Fuel consumption, Btu/hp-hr
O
O
O
§
o
i
ro
§
en
cr
c
n>
o
o
•o
o
I/)
c
V)
o
o
rl-
3
(B
<
ro
fD
a.
o
x
ro
n>
a.
00
3
PO
r>o
m
m
Cl
33
I
jf »
Bi -i. (B (5
2. 7 ST 3 o
? s 2. sr
£ ?
a» m
wli —I
* 2
s <*
2
0 i1
a
no
i
ro
us
INJ
4k
A-
-------
9200
8800
8400
8000
7600
_ 7200
1
8800
6400
6000
EE
R.A.CR
Gil
O 2-8S
Q 2-TC
0 Dtritt
R Rltird
A A1r-to-fu«l rit(o
N Mvrifold ilr tMip«riturt
EE Exttrnil EGR
IE Inttrml EGR
CR CMiprtiiton ritlo
163
0 I? 14
NO, livtl, g/hp-hr
20 22
9200
8800
8400
8000
= 7600
- 7200
6800
6400
6000
RtAl
29..144
811
& 4-NA
7 4-TC
D Dtrit*
R Rttird
A A(r-to-fMl ritlo
H Ntnlfold ilr U«p«rtturi
IE Intirnil EOR
29.,144
NO, Itvil, g/hp-dr
18 20 22 24 26
Figure 4-51(c). Fuel Consumption versus controlled NOx level
for gas engines.
4-159
-------
Fuel consumption increases slightly more than 1 percent per g/hp-hr
of NOX reduction in dual-fuel engines. Derating again results in the worst
fuel consumption penalty per unit NOX reduction. Large-bore gas engines show
the widest range of fuel consumption variation with the use of controls.
Fuel penalties range from slight improvements to a maximum of 1.8 percent
per g/hp-hr of NOX reduction. In general, reduced manifold air temperature
causes the least fuel consumption increase per unit NOX removed. Air-to-fuel
ratio changes and combinations of A/F change and retard cause the largest
fuel penalties per unit of NOX reduced.
The techniques summarized here and their effectiveness, their fuel
penalties, and the time required to implement them will be discussed further
in Chapter 6.
4.4.12 Effects of NQy Controls on HC. CO and Smoke
This section discusses the effect of the application of -;he NOX
controls on the criteria pollutants HC, CO, and smoke. The purpose of this
discussion is to insure that NOX emissions can be reduced by the techniques
discussed above without unacceptable increases in emissions of these other
criteria pollutants.
The effect of NOX control systems on hydrocarbons is shown graphically
by fuel in Figure 4-52. It should be noted that only about 10 to 25 percent
of the hydrocarbon emissions from gas and dual-fuel engines are reactive
(i.e., criteria hydrocarbon pollutants), the remaining 75 to 90 percent being
unburned methane (see Appendix C.4 and Reference 138). In contrast, over 90
percent of the total hydrocarbon emissions from diesel engines are
reactive(^9)j S1-nce ,j-jese] fuel is composed of higher order hydrocarbon
molecules (i.e., there are many hydrogen and carbon atoms in each molecule).
4-160
-------
cr»
.8
.7
.6
£ -5
i
OL
-C
cn
,-" .4
O>
o
re
.3
.2
.1
IE §27
'fn
R&A
10 12 14 16
NOX level, g/hp-hr
18
Diesel
O 2-BS
Q 2-TC
A 4-NA
^ 4-TC
D " Derate
R Retard
A Air-to-fuel ratio
M Manifold air temp.
IE Internal EGR
EE External EGR
CR Compression ratio
i
Figure 4-52(a). HC levels versus controlled NOX levels for diesel
22
engines.
24
-------
t 5
!
- R
R&M
Dual Fuel
Q 2-TC
^T4-TC
D Derate
R Retard
A Air-to-fuel ratio
M Manifold air temp.
EE External EGR
8 10
12 14 16
^ level, g/hp-hr
18 20 22
24
Figure 4-52(b). HC levels versus controlled NOX levels for dual-fuel engines.
-------
I 4
EEU
til
0 MS
Q 2-TC
D Dtritt
R Retard
A Atr-to-fuil ntto
M Mint/old Itr temp.
EE External COR
IE Inttrml EAR
10 U 14
NOX livtl, g/hp-hr
16 IB 20 22
24
Gas
& 4-NA
^ 4-TC
D Derate
R Retard
A Atr-to-fuel ratio
M Mantfold atr tmp.
IE Inttrnal EGR
10 12 14 16 18
NOx level, g/hp-hr
20 22
Figure 4-52(c).
HC levels versus controlled NOX levels
for gas engines.
4-163
-------
Therefore, the reactive emissions from most engines range between 0.1 and 0.5
g/hp-hr, independent of fuel. These values are nearly an order of magnitude
lower than corresponding numbers for uncontrolled automobile engines and
within a factor of two from the ultimate statutory limit of 0.41 gHCt/mile.-'
Furthermore, as discussed in Appendix C.4, the ratio of reactive
hydrocarbons to total hydrocarbons remains, in general, unchanged after the
application of derate, air-to-fuel, retard, manifold air cooling, and EGR.
Thus, reactive hydrocarbon levels respond in the same manner as the total
hydrocarbon levels illustrated in Figure 4-52. Therefore, the remaining
discussion will consider only total hydrocarbon emissions, since these are
what most manufacturers have measured and reported.
With the exception of three four-stroke, turbocharged engines (Engines
No. 6, No. 16, and No. 4), one gas and the other two dual-fuel, the
hydrocarbon emissions do not change much due to the application of NOX
controls. Thus, among the diesel engines, HC emissions generally change by
less than 0.1 g/hp-hr. These changes can be either increases or decreases
from the starting levels of 0.15 to 0.5 g/hp-hr prior to the application of
NOX controls (Figure 4-52(a)). The two-stroke dual-fuel unit (Engine No. 13)
with low emissions of HC experienced at most a 1 g/hp-hr increase in HC
emissions from a baseline level of about 1.5 g/hp-hr. Again the application
of some NOX controls did not cause the HC emissions to increase (Figure
4-52(b)). Control of NOX has a similar effect on HC emissions from gas engines
(Figure 4-52(c)) as it does on these emissions from dual-fuel units. As will
be mentioned in the next section (4.5), methods exist for reducing
— Conversion from g/roile to g/hp-hr based on automobile fuel consumption
rates of 15 miles/gallon on the road and 0.6 Ib/hp-hr on an engine
dynamometer.
4-164
-------
hydrocarbons from large-bore engines. Therefore, it should be possible to
reduce the HC emissions from the three high emission four-stroke,
turbocharged engines.
Similar graphs are shown for CO variations due to the application of
NOX controls in Figure 4-53. These graphs show that the incremental changes
in the CO emissions generally are less than 2 g/hp-hr. Thus, for the diesel
engines shown in Figure 5-27(a), CO emissions increase by less than 0.5 g/hp-
hr when the uncontrolled level is low (i.e., around 1 g/hp-hr) and by 1
to 2 g/hp-hr when the uncontrolled levels are higher (3 to 7 g/hp-hr). One
exception to this trend is engine No. 60, which is turbocharged. Reductions
in its air-to-fuel ratio caused a large increase in CO emissions.
Similarly, dual-fuel engines that are relatively low emitters of CO
experience little changes in these emissions due to the application of NOX
controls whereas those units that start from a higher level show greater
changes. Specifically, the three engines whose controlled level is around
1.5 g/hp-hr show changes that range from no increases to increases of less
than 0.5 g/hp-hr. In contrast, the two four-stroke units whose uncontrolled
emissions are 3.6 to 4.5 g/hp-hr react to the application of NOX controls
with CO changes that range from a decrease of about 2 g/hp-hr to an increase
of 3 g/hp-hr. With the exception of Engine No. 21 (naturally aspirated),
all the gas engines shown in Figure 4-53 experience CO increases less than 1
g/hp-hr when controlled to reduce NOX.
Since the baseline CO emissions from large-bore engines are low, at
least by comparison with automobiles, these incremental changes can represent
large percentage increases, frequently as high as 100 percent. The highest
CO values recorded probably can be reduced by increases in the air-to-fuel
ratio along with the control techniques used (usually retard and/or increased
4-165
-------
i
i-»
cr>
$
o.
JC
en
H 3
-------
Dual Fuel
O 2-TC
t» 4-TC
D Derate
R Retard
A Air-to-fuel ratio
M Manifold air temperature
EE External EGR
20
22
i
<
24
NOX level, g/hp-hr
Figure 4-53(b). CO emissions versus controlled NOX level for dual-fuel engines.
-------
10
s
Gas
O 2-B
Q Z-TC
0 Otratt
ft Rttird
A A1r-to-fu«l ratio
M Hanlfold air temperaturi
EE Enttmal EGR
IE Intemil EGR
10 12 14 It
NO, levtl, g/hp-hr
18 20 22
77.
16
10
_• e
I
129
129
& »-29.
Gas
& 4-NA
^ 4-TC
0 Derate
R Retard
A Air-to-fuel ratio
N Manifold Air Temperaturi
IE Internal EGR
to 29.0, «44
Figure 4-53(c).
10 12 14 16 18 20 22 24
NOX level, g/hp-hr
CO emissions versus controlled NOX
level for gas engines.
4-168
-------
Inlet air cooling). With this in mind, and if one excludes the two-stroke,
blower-scavenged diesel (Engine No. 17), the four-stroke turbocharged diesel
(Engine No. 60), the four-stroke turbocharged dual fuel (Engine No. 16), and
the four-stroke naturally aspirated gas engine (Engine No. 21) that achieve
neither low NOX nor low CO levels (excluding the dual fuel unit), all the
data points for engines with NOX controls fall below 4 g/hp-hr.
Presumably CO emissions would be reduced from the four-stroke,
turbocharged engine (all fuels) as a consequence of any attempt to reduce HC
from them. However, if one excludes the data points for Engine Nos. 5, 6,
15, 16, 28 and 60, all other turbocharged engines could meet a 2 g/hp-hr
limit on CO, even when controlled for NOX. These figures are significantly
lower than the federal standards of 40 g/hp-hr on heavy duty vehicles^140) or
even than the proposed 1977 California standards of 25 g/hp-hr(141).
The plumes from most well maintained large-bore engines are virtually
invisible when the engine is operating at a steady state. However, excessive
retard or large EGR rates will cause diesel and dual-fuel engines to emit
smoke. Visible emissions usually were not measured during experiments with
NOX control systems. In most cases the manufacturers observed the exhaust
visually and simply did not measure emission rates after the plimes become
significantly visible. If the manufacturer did report smoke readings, along
with NOX reduction due to the application of potential controls, the NOX
control results were used in this document only if the plune did not exceed
10-percent visibility.
Figure 4-54 shows the relationships between smoke emissions and N0¥
A
reductions for those engines where data were reported on both pollutants.
All of the engines were diesels with the exception of one dual-fueled unit
(Engine No. 16). The smoke levels ranged from 1.5 to 8 percent opacity or
4-169
-------
4-NA
Dual fuel
Diesel
4-TC
V
o
a.
s
o
to
CO
o
00
D Derate
S Increased speed
A Alter air-to-fuel
M Manifold air temperature
E External EGR
H Water induction
I Internal EGR
CR Reduced compression ratio
R Retard ignition
BS Blower scavenged
NA Naturally aspirated
TC Turbocharged
NOTES:
1. Engine code number denotes
initial point. Control code
denotes level after control.
2. Bacharach and Bosch meters
are filter-type instruments.
3. Smoke levels for engines #8-#12
were measured with a Bosch meter.
#27
10
%«
>i
ex
o
#34
J
D
CR 18 19
#10
6 8 10 12
NOX level, g/hp-hr
14
16
18
20
Figure 4-54. Smoke levels versus NOX levels for large-bore diesel engines.
-------
Bosch Smoke Spot Number 2 to 3 for uncontrolled engines.9/ Since NOy
^™ rt
controls which caused smoke levels to exceed 10 percent opacity were not
considered as acceptable, none of the data points for controlled engines are
K
above this} value. However, the effect of progressive application of retard
or EGR on smoke is demonstrated best by data which include higher smoke \
levels. Such data are presented in Table 4-ie(142) for two different diesfl
engines (b|oth two-stroke) and substantiate the above assertion that smoke ''
. increases 'as either the percentage of exhaust gas recirculation or the degree
of retard is increased. The table also shows the effect of engine (i.e.,
stack) size on opacity.
4.4.13 Combustion of Nonstandard Fuels
Although most stationary 1C engines burn No. 2 distillate or pipeline
quality natural gas, both of which contain less than 0.3.percent sulfur (by
weight), engines are occassionally sold for applicationsvwhich use residual
oils or unprocessed natural gas. The primary air pollution concern with the
use of these fuels is their potential for causing high SOX emissions, since
these fuels typically contain more sulfur then do the distillates and
^f
processed natural gas. This section briefly reviews those applications that
use nonstahdard fuels and their missions.
9/
_ Since the Bosch Smoke Spot Number is a measure of the soiling
characteristics of the particulate and is measured by the darkness of
the spot, of particulate collected on a piece of filter paper, its value
in any g.tven test depends upon such factors as stickiness, color, and
reflectence of the particulate. Opacity, on the other hand depends on the
size distribution of the particulate, which affects the pi line's-ability ,'
nlh- *!*? the Fussage of light' Therefore, there is no general
nship between the two systems.
4-171
-------
TABLE 4-16. RELATIONSHIP BETWEEN SMOKE, EGR, RETARD,
AND ENGINE SIZE
Engine Type3
2-BS-D
2-TC-D
Control b
None
10% EGR
20% EGR
30% EGR
4° advance0
None
4° retard
None
4.9%
8.4%
12.1%
Opacity, %
4.7
12
27.5
59
2.7
4.6
10
6 Cylinder 12 Cylinderd
7.5 11.0
10.0 14.0
11.5 16.7
14.8 21.4
Blower-scavenged engine is two-cylinder test engine with
needle valve injector (from Reference 142). Turbocharged
unit is Engine No. 12.
DA11 EGR data based on hot EGR.
'Injection advance is not a control; data included to
show trend.
Computed from six-cylinder data using Beer Lambert law
(see Figure 8-1), double the stack cross sectional area
(to accommodate doubling of exhaust volume), and the same
particulate concentration (mg/m3) as in six-cylinder unit.
4-172
-------
Seme of the more common applications of engines using unprocessed natural
gas are sewage treatment plants. The engines at these facilities often burn
sewage digester gas. Although the sulfur content of this gas varies, typical
values range from 100 to 150 ppm of hydrogen sulfide (H2S)(143). This level
translates into an emission rate of less than 0.5 g/hp-hr of S02 (assuming
600 Btu/ft3 gas and a fuel rate of 7500 Btu/hp-hr). Emissions data shown
in Table 4-17(144) for a sewage plwt jn LQS Angeles> however> indicate that
some engines emit as much as 1 g/hp-hr of S02U45). These levels are well
below the estimated 2 g/hp-hr emitted from a diesel engine burning 0.5 percent
sulfur distillate (see Section 4.1.6).W
Stationary reciprocating engines occasionally are sold with the
capability of burning residual or crude oils. The only known domestic site
where stationary 1C engines are burning residual oil is the Anamax Mining
Company diesel electrical generation plant near Tucson, Arizona^). This
installation was designed to take advantage of the lower cost of these heavy
oils as well as their availability at that site. Distillation of crude oil
produces distillates which are relatively low in sulfur content and, hence,
concentrates any sulfur present in the crude into the residual oil. The
sulfur content of crude and residual oils depends on their source and, in the
case of residual oil, on how severely the parent crude is distilled. Table
4-18(147) ir)d1cates sulfur contents of crude oils from various sources.
The naturally low sulfur, or "sweet", crudes and their residuals may
have sufficiently low sulfur content to present no S02 emission problem, but
are in short supply and command higher prices than the high sulfur fuels. The
_ The No. 2 distillate available to domestic markets has a sulfur cont
whTch averages 0.25 percent (by weight) and ranges from o!01 to 0 7
4-173
-------
TABLE 4-17. EMISSIONS FROM ENGINES BURNING SEWAGE DIGESTOR GAS
(from Reference 144)
Engine #
1
2
3a
4*
HP
1100
1100
1600
1600
# cyl
10
10
8
8
rpm
330
330
352
350
Fuel
% CH4
62
60
64
64
Heating
Value
580
557
Emissions, g/hp-hr
N0x
3.14
8.43
2.21
2.37
CO
<0.1
<0.1
<0.1
<0.1
HCT
0.96
0.91
9.16
5.65
S02
0.10
0.12
1.08
1.03
^ual-fuel operation, 4.5 gallon/hr No. 2 diesel; 14,600 scf/hr gas.
-------
TABLE 4-18. SELECTED APPROXIMATE PROPERTIES OF CRUDE OILS CURRENTLY GENERATING
U.S. RESIDUAL OIL SUPPLY (Reference 147)
en
Origin
Venezuela
Libya
Mid-East
Indonesia
Alaska (North Slope
and adjacent)
California
Vanadium
(ppm)
100-900
7
20-50
—
15-30
100-200
Property
Nickel
(ppm)
30-100
10
5-15
20
5-15
100
Sul fur
(wt %)
1-6
0.3
1-3
0.1
1
1-5
Nitrogen
(wt %)
0.2-0.6
0.15
0.1
0.1
0.2
0.6-0.9
Residual
Percent
(@ 787°F)
30-60
25
30
50
35
60
Residual
Carbon
(wt %)
10-15
8
10
5
10-15
10
-------
sulfur content of residual oil can be reduced by catalytic hydrodesulfuriza-
tion, a petroleum refining technique. Costs for this refining are discussed
in Section 8.2 for different amounts of fuel sulfur.
Residual oils, blends, some crudes, and heavy distillate also contain
higher amounts of fuel-bound nitrogen than the lighter distillates.
Therefore, the combustion of these fuels may result in greater NOX emissions
than if lighter distillates were burned. However, a byproduct of catalytic
hydrodesulfurizatlon is a reduction 1n both the metal and nitrogen content,
as well as the sulfur, of the residual oils. Thus, fuel oil desulfurlzatlon
processes can reduce fuel nitrogen. In one case, using a high sulfur, high
nitrogen residual fuel feedstock, 78 and 94 percent sulfur removals were
accompanied with 40- to 70-percent reductions in nitrogen, respectively^48).
At this time, however, there are no known emissions data for stationary 1C
engines burning heavy fuels. Thus, the effect of heavy fuel combustion on
NOX and other emissions is unknown.
4.5 EMISSION REDUCTION TECHNIQUES PRIMARILY FOR HC AND CO
As mentioned in Subsection 4.4.12 hydrocarbon (HC) and carbon monoxide
(CO) emissions from large-bore engines are generally low, especially when the
engines are operated at steady state. Therefore, very little effort has been
expended by manufacturers to reduce the emission rate of these pollutants,
except where ancillary to a smoke reduction program. One foreign
manufacturer of medium-bore, high-speed engines has developed a relatively
low HC and CO engine for use 1n mines (0.4 and 1.8 g/hp-hr, respectively,
based on the 13 mode composite test cycle)(149), and at least one domestic
manufacturer of catalysts markets a HC and CO catalytic oxidation unit for
truck sized dlesel and gas-fueled engines^150). When Installed on a diesel,
4-176
-------
these catalysts bring about reductions of 70 percent in CO emissions and 45
percent in HC emissions while fuel consumption, peak smoke, and NOX emissions
remain essentially unchangedUSl). Furthermore, catalytic systans to reduce
HC and CO emissions from heavy duty gasoline and diesel have remained 80
percent to 95 percent effective for both gasoline and diesel engines
throughout the 1000-hour test(152,153). No attanpt to extend e1ther Qf thfise
approaches to large-bore engines has been reported, Therefore, the
description of the HC and CO control techniques which are presented below are
mostly qualitative and frequently draw upon experience gained with medium-
bore, medium-to-high speed engines.
In general, the combustion process is enhanced by high temperatures,
freely available oxygen, good mixing of the fuel-air mixture, and long
residence times. Since these conditions also result in the highest NOX
formation rates, care must be exercised in a balanced emission reduction
effort not to reduce HC and CO at the expense of NOX. Therefore, attempts to
reduce HC and CO emissions must concentrate on quench layer effects, air-to-
fuel mixture inhomogenelties, fuel droplet size, exhaust gas treatment, and
the like.
The three major approaches to control of HC, (1) modifying air and
fuel handling systems, (2) modifying combustion chamber designs, and (3)
treating the exhaust, are discussed In detail in the following subsections.
4.5.1 Modifications to A1r and Fuel Handling Systans
This category of HC and CO emission control techniques can be further
divided into the three subcategorles listed below:
4-177
-------
1. Air-to-fuel mixture
2. Liquid fuel injection (compression ignition engines only)
3. Cylinder scavenging
Since HC and CO emissions decrease as the combustible mixture becomes
leaner, at least up to a point (see Figure 4-28), increasing the air-to-fuel
ratio can be a control technique for these pollutants. On a diesel engine
this usually requires the addition of a turbocharger, if the engine is not
already equipped with one, or an increase in the capacity of the existing
turbocharger unit. In a carbureted spar-k-ignition engine, the air-to-fuel
ratio can be changed by alterations to the Venturis and fuel nozzles in the
carburetor.1 Unfortunately, as shown in Figure 4-28, both minimum HC and
maximum NOX emissions frequently occur at about the same air-to-fuel ratio.
Hence this approach has limited utility in a total emission reduction
program, unless coupled with other controls.
Some carbureted gas engines, however, are adjusted to operate near
stoichiometric, on the average, to overcome the problem of poor distribution
of fuel and air among the cyl inders(154). These same engines would have
excessively lean mixtures in some cylinders if the carburetor were not set
rich, and these lean cylinders could not support combustion (i.e., they would
exceed the lean misfire limit). This problem can be ameliorated by improved
aerodynamic; design of the carburetor and intake manifold system or by
converting the engine to a fuel injection system. In fact, most large-bore,
gas fueled-engines already use fuel injectors.
The [Injector in a liquid fueled compression ignition engine is
required to deliver fuel only when specified, to atomize the fuel into
droplets of'the smallest possible diameter, to distribute these droplets
uniformly throughout the combustion volume, and to avoid impinging fuel on
4-178
I /' *
\
-------
the relatively cool cylinder walls (potential quench layer). Therefore,
attempts to reduce hydrocarbon (and smoke) emissions have traditionally
concentrated on higher injection pressures to improve atomization and on
changes to the design of the injector tip, in particular, hole size,
location, and number(155). Frequently these efforts have gone hand-in-hand
with changes to the shape of the piston head, because wall impingement
problems can sometimes be reduced by directing a portion of the fuel spray
onto the contoured surface of the piston head.
A major change to the injector that reduces HC emissions is the
introduction of the low sac nozzle (LSN) injector. As shown in Figure
4-55(156), older nozzles use a valve that is located at some distance from
the injector tip to control the flow of fuel into the combustion chamber. A
consequence of this design is the presence of a small volume, or sac, within
the injector that lies between the valve and the combustion chamber. When
the valve is open, the fuel flows through this sac on its way to the
conbustion chamber, where it mixes with the air, ignites, and contributes to
a rapid pressure rise in the cylinder. The high cylinder pressure prevents
any fuel that is in this little passageway when the valve closes from
entering the cylinder during the combustion process. However, as the piston
moves away from the cylinder head during the expansion process, this remaining
fluid "dribbles" out of the sac. Since the cylinder temperatures are too
low to support con bus ti on by now, these droplets remain un burned, or are only
partially oxidized, and eventually pass into the exhaust system in that form.
Several manufacturers have practically overcome this problem by
switching to a nozzle with a greatly reduced sac volume, i.e., a low sac
nozzle(157). A schematic representation of such an injector is shown in Figure
4-55. In some designs of sac volume is so small as to be virtually negligible.
4-179
-------
Fuel under pressure
Valve
00
o
-Holes through which fuel is
atomized into the cylinder
a.
Old standard production
spray tip.
b. Low sac volume spray
tip.
Figure 4-55. Schematic illustrations of standard production
spray tip and the low sac volume spray tip.
(after Reference 157).
-------
The use of such nozzles lead to 30- to 35-percent reductions in HC and C0(158).
The reductions were accompanied by a decrease in opacity to 4 percent (at
rated conditions); NOX emissions, however, rose nearly 35 percent, probably
as a result of changes in the injector tip design that accompanied the switch
to a lew sac volune. These changes could alter the fuel spray pattern and/or
degree of atomization.
Unburned or partially burned hydrocarbons that come from either
injector "dribble" or quench layers are usually concentrated in the last
volumes of gas to be expelled from the cylinder during the exhaust stroke.
Therefore, EGR, whether internal or external, serves to reduce HC emissions,
in addition to NOX, as long as the recirculation rate is not so great that
portions of the combustion reaction are quenched. Reductions in the
scavenging pressure in a two-stroke, blower-scavenged engine can also have
the same effect.
4,5.2 Combustion Chamber Modifications
When one discusses HC and CO control techniques for large-bore
engines, one must remember that these contaminants are present only at very
low quantities (typically below 2.0 g/hp-hr) in the exhaust from these units.
This is so mainly because of the presence of excess air during combustion. In
addition, the high temperatures and relatively long residence time (low engine
speed) facilitate the combustion of the fuel. Finally, the large voline, low
surface-to-volume ratio, and use of injectors gives the engineer of medium-
and-large-bore units more control on the following parameters than he has with
automobile-sized engines: air-to-fuel distribution within a cylinder,
variation in air-to-fuel ratio among cylinders, and fuel droplet impingement
on the cylinder walls. As a result, an average value for the total HC
4-181
-------
emissions from a variety of large engines is 2 g/hp-hr (see Figure 4-52)
which corresponds to 0.1 percent of the incoming fuel.£^/ Thus, even though
combustion characteristics are evidently very good from the point of view of
HC and CO, combustion chamber design can be important as one attempts to
oxidize the last 0.1 percent of the fuel.
A variety of chamber designs has already been presented in Section
4.4.8. Of these techniques, higher swirl, inlet valve and/or port redesigns,
and contoured pistonheads reduce HC and CO because they improve the mixture
rate and homogeneity of the fuel and air. The designs for staged combustion,
such as precombustion chambers and "squish lips", can reduce HC and CO, too.
These geometries cause the combusting mixture to expand the available oxygen
with the unburned or partially burned fuel. Unfortunately this mixing
process is also the source of some increased fuel consumption; as the burning
mixture expands into the main chamber, its pressure is decreased slightly,
resulting in reduced power output. The mechanical energy lost by the gas
becomes thermal energy and is seen as a slight increase in the gas
temperature. Of the 5- to 8-percent energy loss attributed to precombustion
chambers (see Section 4.3.8), however, most is probably due to the heat loss
in the precombustion chamber.
Another technique that is available to decrease HC emissions, especially
from diesel engines without affecting NOX emissions, is to reduce the "quench
layer", or region immediately adjacent to the combustion chamber surfaces.
comon assumption used in calculating mass emissions of total hydro-
rbons is that a typical hydrocarbon tr
is the same as a typical fuel molecule'
carbons is that a typical hydrocarbon molecule emitted in the exhaust
4-182
-------
Heat transfer outward through the cylinder walls causes the mixture
temperatures in this layer to be too low to support complete combustion.(16°)
Some unburned hydrocarbon emissions in diesel engines result from fuel
droplets that were injected into this "quench layer". Attempts to reduce
this source of HC emissions rely primarily on injector design to avoid
impinging the fuel droplets onto the cylinder walls and on relocation of the
uppermost piston ring as close to the piston head as possible to minimize the
relatively cool space between the piston and the cylinder adjoining the
combustion volume^6*).
These quench layer hydrocarbons could theoretically be reduced by
using higher cooling water temperatures. However, the increase in wall
temperatures may give rise to structural problems (current cooling systems
are presumably designed to avoid such problems) and, in any case, the
resulting higher combustion temperatures would result in an increase in N0¥
A
emissions.
Since high temperatures are conducive to complete combustion and,
hence, low HC and CO emissions, these pollutants can be minimized by raising
the peak combustion temperature through an increase in the compression ratio.
This parameter is the ratio of the largest volume in the cylinder to the
smallest — i.e., the ratio of the volume when the piston is at the bottom of
its stroke (bottom dead center or BDC) to that when it is at the top (TDC).
Increasing this ratio increases the volume change and hence the temperature
rise of the fuel and air during the compression stroke. This approach
also increases the efficiency of the engine, but unfortunately at the expense
of higher NOX emissions.
4-183
-------
4.5.3 Exhaust Treatment
Potential methods of treating the exhaust to reduce NOX emissions
have been discussed in Section 4.4.9. Here we are concerned only with
catalytic converters, thermal reactors, and exhaust manifold air injectors,
which are designed to reduce HC and CO emissions. Although these systems
have been tested on small- and medium-si zed engines, to our knowledge they
have not been used on large-bore units. Application of these techniques to
large engines require careful design to insure that proper operating
temperatures are maintained in all parts of the larger units. The oxidation
reactions which convert the HC and CO to water and COg take place only above
a certain temperature, and this high-temperature environment 1s maintained
inside the exhaust treatment unit by the heat from earlier reactions. It 1s
more difficult to Insure the existence of a homogeneous high-temperature
environment throughout the combustion volume of a large unit than of a small
one because the heat release from localized reactions is absorbed by more gas
In the large system than in the small one. That is, the heat generated may
not affect the entire volume in the large unit. Nevertheless, there does not
seem to be any technical reason why such units cannot be designed.
The catalytic converter has been proven on mobile gasoline engines and
functions by causing the oxidation reaction to continue at reduced
temperatures. It is a passive device which consists of a container that
houses a porous structure covered with the catalyst material. The converter
replaces a portion of the exhaust tail pipe. The chemical reaction then
takes place on the catalyst-coated surface as the exhaust passes through the
converter. If automotive type platinum catalytic converters are used, only
unleaded fuels can be burned to avoid inactivating the catalyst. In
addition, fuels with low sulfur content should be used to suppress sulfurlc
4-184
-------
acid formation (HgSC^). This technique is not limited to small-bore engines
and could, therefore, be applied to large-bore units as well. The need to
design the converter to operate at uniform elevated temperature has already
been discussed; cost implications are considered in detail in Section 8.2.
In the thermal reactor air is added to the exhaust stream in a
container which is located in the exhaust pipe system and is specially
designed to maximize both the residence time of the charge and its
turbulence. The purpose of the device is to provide a site where oxidation
can proceed at elevated temperatures. These are maintained by the heat
released from the oxidation reaction of CO and HC with excess air. Air
injection into the exhaust manifold functions to the same end as the thermal
reactor. However, it is not as effective an oxidizer of HC and CO as the
thermal reactor because it utilizes the existing shape of the exhaust systen.
This volume is generally not optimized for maximum residence time, heat
retention, or turbulent mixing.
4.6 SMOKE
In general smoke in well operated and maintained large-bore engines is
a problem only for compression ignited engines when fired with nothing but
liquid fuel. Attempts to reduce visible emissions aim at reducing locally
rich zones by improving air-to-fuel mixing and distribution throughout the
cylinder. The following techniques are most commonly used for this purpose:
t Improved combustion through chamber redesign, particularly piston
head shape and increased swirl
t Leaner air-to-fuel ratio (by carburetor adjustments, inlet
manifold design, or turbocharglng)
4-185
-------
• Improved fuel atomization and dispersion
• Low-sac volume injectors
A number of manufacturers have successfully reduced smoke emissions over the
operating range of large diesel engines through the use of improved needle
and low-sac fuel injector designs. Visible emissions from most current,
properly maintained units are well below 10-percent opacity during continuous
load application^62*163).
Figures 4-56(164) and 4-5?(165) illustrate the improvements obtained
by the Electromotive Division (EMD) of General Motors on both blower-
scavenged and turbocharged versions of their two-stroke Model 645 diesel
engine (although this engine is used mainly in locomotives, it is also sold
for stationary applications). Smoke reductions were also achieved on these
engines by widening the size of the ports in the cylinder liners 12-1/2
percent to improve charging and scavenging processes and by moving the top
piston ring ("fire ring") three-quarters of an inch higher. Sea level smoke
emissions at rated conditions are now about 6-percent opacity for both
engines, having been reduced from 20 percent for the turbocharged unit and
30 percent for the blower scavenged one. The effect of derating is shown in
Figure 4-56 to illustrate that this technique has to be used with blower-
scavenged engines at high altitude in order to comply with local regulations
that restrict visible emissions to a Ringelmann No. 1 (20-percent opacity).
Although not recommended by manufacturers, additives are sometimes
mixed with the fuel to reduce smoke. Barium is a common additive and
4-186
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
reportedly reduces the particulate mass emission ratej£/ in medium-bore
dies el engines by as much as 50 per cent (166~168). This is theorized to be
due to the inhibiting effect of the barium on the formation of carbon in the
vicinity of the burning fuel droplets, or to its action as a catalyst
accelerating the oxidation of the carbon formed(169).
In addition to reportedly decreasing the mass of particulate emitted,
the character of the particulate composition is altered as well by the
additives^7^). jnjs was observed in one case where carbon comprised
96 percent, hydrogen 2 percent, iron 1 percent, and zinc 1 percent of the
particulate sampled when untreated fuel was used. In contrast, carbon
comprised only 36 percent and hydrogen 1 percent of the particulate mass
emitted when a barium treated fuel was substituted. Barium sulfate (83804)
now represented about 60 percent of this mass, and iron and zinc were present
in trace quantities only.
Despite the effectiveness of this approach, it is not favored by
industry because of concern over possible build-up problems that might occur
at the cylinder ports as a result of the increased metallic content of the
fuel(171>172K Furthermore, BaS04 and tne other soluble barium compounds
produced may represent a health hazard when emitted into the atmosphere in
substantial quantities.
Although no particulate scrubber is known to have been used on large-
bore engines, probably because they usually operate at steady state and
generally emit little smoke, one automobile-oriented service industry has
developed such a device to remove lead particulates from automobile
12/
The level of smoke opacity was determined using a Hartridge Smokemeter.
The weight of solid particulates was measured by filtering a precise
volume of exhaust gas through glass filter papers and weighing the
deposited soot.
4-189
-------
exhaust(173). In tests on four cars equipped with these particulate traps
the average air-suspended lead emissions were reduced about 67 percent from
26.6 to 10.1 percent of the lead contained in the fuel. The same particulate
trap also reduced total particulate emissions (one-fifth of which is lead) by
74 percent.
Presumably, a similar system, or one of the many kinds of particulate
scrubbers available, could be used if it were decided to reduce NOX
significantly by means of large amounts of retard or EGR. Both these
techniques reduce NOX effectively, but, beyond a certain point, only at the
expense of smoke.
4-190
-------
REFERENCES FOR SECTION 4
(1) Lachape le, D. G,, J. S. Bowen, and R. P. Stern (EPA). Overview of
Environmental Protection Agency's NOX Control Technology for Stationary
Combustion Sources. (Presented at 67th AICHE Annual Meeting. December 4
1974).
(2) Durkee, K., E. A. Noble, F. Collins, and D. Marsland (EPA). Draft
Standards Support Document: An Investigation of the Best Systems of Emission
Reduction for Stationary Gas Turbines. Environmental Protection Agency
August 1974. p. VI-17.
(3) Ibid.
(4) Ibid.
(5) Urban, C. M., et al. Emissions Control Technology Assessment of Heavy
Duty Vehicle Engines. EPA-460/3-74-007. December 1973.
HL ^]emin9>uD< R'na!?d I: R-,Frencn- Durability of Advanced Emission
Controls for Heavy Duty Diesel and Gasoline Fueled Engines. EPA-460/3-73-010
September 1973.
cA R: B;.and A- J- Rogowski. The Interaction of Air Motion, Fuel
Spray, and Combustion in the Diesel Combustion Process. Journal of
Engineering for Power. 94, Serial A, No. 1. January 1972.
(81 T!!aJterS0?' ?' J; and ?• A> Hene1n- Emissions From Combustion Engines
and Their Control. Ann Arbor, Michigan. Ann Arbor Science Publishers, Inc.,
iy '" • . bcl) .
(9) Savery, C. W., R. A. Matula, and T. Asmus. Progress in Diesel Odor
Research. SAE Paper 740213. February 1974. "uiweiuaor
(10) Springer K. J. and C. T. Hare. Four Years of Diesel Odor and Smoke
control Technology Evaluations — A Summary. ASME Paper 69-WA/APC-3
November 1969.
(11) Springer, K. J. and R. C. Stahman. Control of Diesel Exhaust Odors
' NeW Y°rk' New Y°;k
c ?f,0dor Components in Diesel Exhaust. Arthur D. Little.
September 1973.
(13) Springer and Stahman, Op. Cit.
(14) Patterson, D. J., Op. Cit.. p. 132.
(15) Ibid., pp. 32, 267.
^"™~^^^~~
(16) Ibid., pp. 219, 286.
4-191
-------
(17) Ibid., p. 272.
(18) Roessler, W. U., A. Muraszew, and R. D. Kopa. Assessment of the.
Applicability of Automotive Emission Control Technology to Stationary Engines.
EPA-650/2-74-051. July 1974.
(19) Turley, C. D., D. L. Brenchley, and R. R. Landolt. Barium Additives
as Diesel Smoke Suppressants. Journal of the A1r Pollution Control Association.
23_. September 1973.
(20) Bolt, J. A. and N. A. Henein, The Effect of Some Fuel and Engine Factors
on Diesel Smoke. SAE Paper 690557. August 1969.
(21) Patterson, D. J., Op_. C it., p. 272.
(22) Ibid.
(23) Offen, G. R. (Acurex/Aerotherm). Trip Report. November 26, 1974.
(24) Standards of Performance for Fossil-Fuel Fired Steam Generators. Federal
Register. 36 (247), Title 40, Subpart D. December 23, 1971.
(25) Jenkins, R. and G. D. McCutchen. New Source Performance Standards.
Environmental Science and Technology, j5. October 1972.
(26) Standard Practices for Stationary Diesel and Gas Engines, Diesel Engine
Manufacturers Association, p. 118. Sixth Edition. 1972.
(27) S. R. Krause, D. F. Merrion, and G. L. Green, "Effect of Inlet Air Humidity
and Temperature on Diesel Exhaust Emissions," SAE Paper 730213.
(28) Coordinating Research Council (CRC), "Effect of Humidity of Air Intake
on Nitric Oxide Formation in Diesel Exhaust," CRC Report 447, December 1971.
(29) "Revised Heavy Duty Engine Regulations for 1979 and Later Model Years,"
Federal Register, Volume 41, No. 101, May 24, 1976.
(30) Environmental Protection Agency (40 CFR Part 85, Subpart A), "Control
of Air Pollution from New Motor Vehicles and New Motor Vehicle Engine,"
(Notice of Proposed Rule Making), February 1971.
(31) Krause, Merrion, and Green, Op. Cit.
(32) S. R. Krause, "Effect of Engine Intake-Air Humidity, Temperature, and
Pressure on Exhaust Emissions," SAE Paper 710835.
(33) Private communication between C. L. Newton (Colt) and D. R. Goodwin,
August 2, 1976.
(34) Krause, Op. Cit.
(35) W. J. Brown, et al., "Effect of Engine Intake-Air Moisture on Exhaust
Emissions," SAE Paper 700107.
4-192
-------
(36) Scott Research Laboratories, "Effect of Laboratory Ambient Conditions
on Exhaust Emissions," Project No. 2846.
(37) J. A. Robinson, "Humidity Effects on Engine Nitric Oxide Emissions at
Steady-State Conditions," SAE Paper 700467.
(38) Coordinating Research Council (CRC), "Effect of Humidity of Air Intake
on Nitric Oxide Formation in Diesel Exhaust," CRC Report 447, December 1971.
(39) Reference 29, Op. Cit.
(40) Krause, Merrion, and Green, Op. Cit.
(41) Private Communication between R. W. Sheppard (Ingersoll-Rand) and
D. R. Goodwin (EPA), June 17, 1976.
(42) Krause, Op. Cit.
(43) Private communication between R. W. Sheppard (Ingersoll-Rand) and
D. R. Goodwin (EPA), June 17, 1976.
(44) Krause, Op_. Cit.
(45) Ibid.
(46) Krause, Merrion, and Green, Op. Cit.
(47) Krause, Op. Cit.
(48) Brown, Op. Cit.
(49) Reference 36, Op. Cit.
(50) Private communication between C. K. Powell (Ingersoll-Rand) and G. R.
Offen (Acurex/Aerotherm) August 30, 1974
(51) Urban, C. M. and K. J. Springer, Study of Exhaust Emissions from Natural
Gas Pipeline Compressor Engines. Southwest Research Institute. PR-15-61
February 1975.
(52) Krause, Merrion, and Green, Op. Cit.
(53) Reference 28, Op. Cit.
(54) Krause, Merrion, and Green, Op. Cit.
(55) Reference 28, Op. Cit.
(56) Krause, Merrion, and Green, Op. Cit.
(57) Ibid.
(58) Brown, Op. Cit.
4-193
-------
(59) Urban and Springer, Op. C1t.
(60) Coordinating Research Council (CRC) "Cooperative Study of Heavy Duty
Diesel Emission Measurement Methods," CRC Report 487, July 1976.
(61) Coordinating Research Council (CRC) "Cooperative Study of Heavy Duty
Diesel Emission Measurement Methods," CRC Report 487, July 1976.
(62) Recomnended Practice for Heavy Duty Engine Emission Measurement and
Test Procedures. EPA/Office of Air and Waste Management/Emission Control
Technology Division/Standards Development and Support Branch. February 28,
1977.
(63) Diesel Exhaust Emission Measurement Procedure for Low and Medium Speed
Internal Combustion Engines. Diesel Engine Manufacturers Association. 1974.
(64) Measurement of Carbon Monoxide, and Oxides of Nitrogen in Diesel
Exhaust — SAE J177A. Society of Automotive Engineers. March 1974.
(65) Private conmunication between T. M. Fisher (6MC) and D. R. Goodwin (EPA)
May 12, 1976.
(66) Holman, J. P. Experimental Methods for Engineers. McGraw Hill,
New York, N.Y. 1971.
(67) Juneja, W. K., D. D. Horchler, and H. M. Haskew. "A Treatice on Exhaust
Emission Test Variability," SAE Paper 770136, 1977.
(68) Youngblood, S. B. (Acurex/Aerotherm). Meeting report. April 22, 1975.
(69) Ibid.
(70) Private conmunication between C. L. Newton (Colt) and D. R. Goodwin
(EPA), August 2, 1976.
(71) Private communication between C. L. Newton (Colt) and D. R. Goodwin
(EPA), April 2, 1976.
(72) Private conmunication between G. P. Hanley (GMC) and D. R. Goodwin (EPA),
January 29, 1975.
(73) Ibid.
(74) Standard Practices for Low and Medium Speed Stationary Diesel and Gas
Engines. Diesel Engine Manufacturers Association. New York, New York. 1972.
p. 28.
(75) Schaub, F. S. and K. V. Beightol (Cooper-Bessemer). Effect of Operating
Conditions on Exhaust Gas Emissions of Diesel, Gas Diesel, and Spark Ignited
Stationary Engines. Unpublished Cooper-Bessemer Co. Report. Presented at
the Gas Compression Shortcourse, Norman, Oklahoma, Fall 1973.
(76) Patterson and Henein, Op_. Cit., pp. 267-268.
4-194
-------
(77) Roessler, Muraszew, and Kopa, Op_. C1t., p. 5-3.
(78) Fletcher, J. S. (Acurex/Aerotherm) Meeting report. April 22, 1975,
(79) Offen, 6. R. (Acurex/Aerotherm) and F. S. Schaub (Cooper-Bessemer).
Private cormiunication. October 13, 1975.
(80) Youngblood, S. B. (Acurex/Aerotherm) and M. P. Thompson (White Superior).
Private conmuni cation. February 20, 1976.
(81) Ewing, 6. H. (AGA) and D. R. Goodwin (EPA). Private communication,
September 24, 1976.
(82) Patterson and Henein, Op. Cit., pp. 125, 267.
(83) Reference 78, Op. Cit.
(84) Ibid.
(85) Patterson and Henein, Op. Cit., pp. 206-207.
(86) Cox, N. S. (Waukesha) and G. R. Offen (Acurex). Private communication.
June 16, 1975.
(87) Schaub and Beightol, Op. Cit.
(88) Reference 68, Op. Cit.
(89) Ibid.
(90) Patterson and Henein, Op_. Cit.. pp. 125, 267.
(91) Obert, E. F., Internal Combustion Engines and Air Pollution. Intext
Educational Publishers. New York, New York. 1973. p. 619.
(92) Reference 68, Op. Cit.
(93) Youngblood, S. B. (Acurex/Aerotherm) and J. Webb (Turbos, Inc.). Private
communication. June 5, 1975.
(94) Roessler, Muraszew, and Kopa, Op. Cit. pp. 4-82, 4-84.
(95) Storment, J. 0. and K. J. Springer. Assessment of Control Techniques
for Reducing Emissions from Locomotive Engines. Southwest Research Institute.
AR-884. April 1973.
(96) Roessler, Muraszew, and Kopa, Op_. C1t.
(97) Shaw, J. C. Emission Reduction Study on a Carbureted Natural Gas Fueled
Industrial Engine. Draft ASME Paper. White Superior Division. White Motor
Corporation. November 1974.
4-195
-------
(98) Schaub, F. S. (Cooper-Bessmer). Methods of Reduction of NO Emissions
from Large Slow Speed Diesel Engines. U.S. Army Corps of Engineers. DACA
87-70-C-0019. June 24, 1970. pp. 12
(99) Fleming and French, Op_. Cit.
(100) Ibid.
(101) Characterization and Control of Emissions From Heavy Duty Diesel and
Gasoline Fueled Engines. Fuels Combustion Research Group. Bartlesville Energy
Research Center, Bureau of Mines. EPA-IAG-0219(D). December 1972. NTIS:PB219052
(102) Bosecker, R. E. and D. F. Webster. Precombustion Chamber Diesel
Emissions -- A Progress Report. SAE Paper No. 710672. August 1971.
(103) Urban, Op. Cit.
(104) Fleming and French, Op. Cit.
(105) Youngblood, S. B. (Acurex/Aerotherm) and G. Hanley (General Motors).
Private communication. May 19, 1975.
(106) Bosecker and Webster, Op. Cit.
(107) Shaw, 0£. Cit.
(108) Bosecker and Webster, Op. Cit.
(109) Ibid.
(110) Offen, G. R. (Acurex/Aerotherm) and F. Schaub (Cooper-Bessemer). Private
comnunication. January 13-14, 1975.
(Ill) Shaw, Op. Cit.
(112) Storment and Springer, 0£. Cit.
(113) Shaw, Op. Cit.
(114) Walder, C. J. Can Diesels Meet 1975 California Emission Limits? Auto-
motive Engineering. 80:27-30. December 1972.
(115) Bascunana, J. L. Divided Combustion Chamber Gasoline Engines: A Review
for Emissions and Efficiency. Journal of the Air Pollution Control Association.
20 (7). July 1974.
(116) Urlaub, Alfred. How M. A. N. Cleans Up Diesel Engine Exhaust.
M. A. N. Research Engineering. Manufacturing. No. 4. October 1973.
(117) Bertodo, R., et al. Evolution of a New Combustion System for Diesel
Emission Control. Perkins Engines Company (England). SAE 741131. November
1974.
4-196
-------
(118) Walder, Op_. C1t.
(119) Bosecker and Webster, Op_. C1t.
(120) Diesel and Gas Turbine Worldwide Catalog. Diesel and Gas Turbine Progress.
Milwaukee, Wisconsin. 39. 1974.
(121) An Automotive Engine That May be Cleaner. Environmental Science and
Technology. ]_. August 1973. p. 688.
(122) Johnson, P. R., et al. (GMC). Vehicle Emission Systems Utilizing a
Stratified Charge Engine. SAE Paper 741157. October 1974.
(123) Bertodo, Op_. Cit.
(124) Ibid.
(125) Ibid.
(126) Durkee, K. R. (EPA/ISB) to J. C. Berry (EPA/Chief ISB). Meeting report
with DEMA representatives, September 3, 1974.
(127) Hartwell, N. J. E. (Perkins Engines, Farmington, Michigan). Letter
to G. R. Offen (Acurex/Aerotherm). November 18, 1974.
(128) Brisson, B., et al. A New Diesel Combustion Chamber -- The Variable
Throat Chamber. SAE Paper 730167. 1973.
(129) Henderson, R. D. (Caterpillar Tractor Company). Commentary on Diesel
Exhaust Emission Control System. November 11, 1974.
(130) Mooney, J. J., C. E. Thompson, and J. C. Dettling (Engelhard) "Three-
Way Conversion Catalysts: Part of the New Emission Control System," SAE Paper
770365. 1977.
(131) Bradow, R. L., F. D. Stump, (EPA) "Unregulated Emissions from Three-
Way Catalyst Cars," SAE Paper 770369. 1977.
(132) Anderson, H. C., et al. (Engelhard Industries). Catalytic Treatment
of Nitric Acid Plant Tail Gas. Industrial & Engineering Chemistry. 53.
March 1961. pp. 199-204. ~
(133) Gillespie, G. R., et al. (Engelhard Industries). Nitric Acid: Catalytic
Purification of Tail Gas. Chemical Engineering Processes. 68:72-77.
April 1972.
(134) Offen, G. R. (Acurex/Aerotherm) and A. W. Tamarelli (Engelhard Industries).
Private communication. December 3, 1974.
(135) Sudar, S. and Grantham. Diesel Exhaust Emission Control Program.
Atomics International Division. Rockwell International. Report No. Al-73-61.
January 1974.
4-197
-------
(136) Urban, C. M. and K. J. Springer. Study of Exhaust Emissions from
Natural Gas Pipeline Compressor Engines. Southwest Research Institute. AGA
Project PR-15-61. February 1975. p. 57.
(137) Durkee, K. R. (EPA) Meeting Report, September 3, 1974.
(138) Newton, C. L. (Colt Industries). Letter to G. R. Offen (Acurex/Aerotherm)
November 20, 1974.
(139) Henderson, R. D. (Caterpillar). Letter to R. D. Selffert (EPA).
January 17, 1975.
(140) 40 CFR 85, Subpart J -- Engine Exhaust Gaseous Emission Regulations
for New Diesel Heavy Duty Engines.
(141) New Vehicle Standards Summary. California Air Resources Board Fact
Sheet 6 (Revised). July 8, 1974.
(142) Storment and Springer, Op. C1t.
(143) Sacramento Area Consultants, Study of Methane Uses, Sacramento Regional
County Sanitation District, June 1976.
(144) Private coirniunication between G. Thomas (Los Angeles County APCD) and
J. S. Fletcher (Acurex/Aerotherm) December 2, 1974.
(145) Ibid.
(146) Madman, B., "Most Efficient Fossil Fuel Generating Plant?" Diesel
and Gas Turbine Progress. April 1976. p. 13-16.
(147) Whlsman, M. L. and F. G. Cotton. BuMines Data Promise Help 1n Identifying
Petroleum Spill Sources. Oil and Gas Journal. December 27, 1971. pp. 111-113.
(148) Durkee, Op. Cit., p. II-5.
(149) Diesel and Gas Turbine Worldwide Catalog. 1975 Edition, p. 10.
(150) Hansel, J. G. (Engelhard Industries). Letter to G. R. Offen (Acurex/
Aerotherm). April 14, 1975.
(151) Reference 101, Op_. Cit.
(152) Urban, Op_. Cit.
(153) Fleming and French, Op. Cit.
(154) Shaw, J. (White Superior). Letter to J. S. Fletcher (Acurex/Aerotherm).
November 14, 1974.
(155) Melton and Rogowski, pj>. Cit.
4-198
-------
(156) A Progress Report on Electro-Motive's Emissions Reduction Program
for Diesel Engines, Electro-Motive Division, GMC. September 1972.
(157) Ibid.
(158) Storment and Springer, Op_. CJt.
(159) Standard Practices for Low and Medium Speed Stationary Diesel and Gas
Engines. Diesel Engine Manufacturers Association. New York, New York. 1972.
p. 298.
(160) Patterson and Heneln, Op_. C 1t.. pp. 117-124.
(161) Bascom, R. C., et al. Design Factors that Effect Diesel Emission.
SAE Paper 710484. July 1971.
(162) A Progress Report on Electro-Motive's Emissions Reduction Program
for Diesel Engines. Electro-Motive Division, GMC. September 1972.
(163) Offen, G. R. (Acurex/Aerotherm). Trip Report. September 13, 1974.
(164) Reference 162, Op_. Clt.
(165) Ibid.
(166) Turley, Brenchley, and Landolt, Op_. C1t.
(167) Shamah, E. and T. 0. Wagner. Fuel Quality or Engine Design: Which
Controls Diesel Emission? SAE Paper No. 730168. January 1973.
(168) Golotham, D. W. Diesel Engine Exhaust Smoke: The Influence of Fuel
Properties and the Effects of Using Barium-Containing Fuel Additive. SAE
670092. 1967.
(169) Ibid.
(170) Shamah and Wagner, Op_. C1t.
(171) Ibid.
(172) Golotham, Op_. C1t.
(173) Adams, W. E., F. J. Marsse, and D. L. Lenane. "Lead-Compatible Emission
Controls — A Route to Improved Fuel Economy." (Ethyl Corp.) NPRA Paper
F&L-74-60. (Presented at Fuels and Lubrication Meeting, National Petroleum
Refiners Association, November 1974.)
4-199
-------
CHAPTER 5
MODIFICATIONS AND RECONSTRUCTIONS
5.1 BACKGROUND
This chapter identifies and discusses possible or typical changes to
stationary reciprocating internal combustion engines which could be termed
modifications or reconstructions. According to Federal Regulations, such
changes would subject an engine to a standard of performance. A modification
is defined as "...any physical change in, or change in the method of
operation of an existing facility which increases the amount of any air
pollutant (to which a standard applies) emitted into the atmosphere by that
facility or which results in the emission of any air pollutant (to which a
standard applies) into the atmosphere not previously em1tted"U). An
"existing facility" is defined as one which would be required to conform to
a standard of performance, if it were new, but which was, in fact,
constructed or modified before the date of proposal of the standard of
performance.
The regulation requires the owner or operator of any engine classified
as an "existing facility" to notify the EPA of changes which could cause an
increase in emissions of an air pollutant for which a standard of performance
applies(2). These changes would not be termed "modifications" — i.e., the
existing facility would not have to meet the standards of performance -- if
5-1
-------
the owner or operator could demonstrate that no Increase in emissions for
which a standard applied resulted from the alteration.
In addition, the regulation also defines the term "reconstruction,"
namely the ",.,replacement of a substantial majority of the existing
facility's components Irrespective of any change of emission rate"(3).
Therefore, 1f substantial numbers of parts are changed 1n an engine which
belongs to the class of existing facilities, this engine will be considered
to have undergone a reconstruction and will have to conform to standards of
performance as 1f 1t were a new engine. The purpose of this provision is to
discourage the perpetuation of a facility which, in the absence of a
standard, would normally have been replaced(4). Again, the regulation
requires the owner or operator to provide Information concerning the
construction or reconstruction of an existing facility to EPA(5),
Certain practices, however, are exempted from being considered as a
modif1cation(6). These exemptions are listed below, in terms that are
appropriate for stationary engines.
1. Routine maintenance, repair, and replacement of parts
2. Increases in power generation, provided such increases do not
exceed the rated power of the engine
3. Increases in hours of operation
4. Use of an alternative fuel, if specified by the manufacturer for
that particular engine
5. Addition of a control device which reduces emissions or
replacement of a control device by another of at least equal
efficiency
6. Relocation or a change in ownership of an engine
5-2
-------
The addition of a new, modified, or reconstructed unit to a multiple-
unit installation subjects only the addition to a standard of performance and
not the remainder of the existing facility(7).
Specific changes to stationary, reciprocating internal combustion
engines that could be classified as modifications or reconstructions will be
discussed separately in the following two sections.
5.2 MODIFICATIONS
5.2.1 General
As discussed in Section 5.1, the law regarding modifications to
stationary reciprocating engines is very explicit; that is, any physical or
operational changes to an engine that result in increased emissions (for
which a standard applies) subjects that engine to a standard of performance.
This section will discuss those physical and operational changes to
stationary engines that could result in increased emissions. Alterations
which are likely candidates to be exempted as modifications are noted.
Physical changes are alterations to either the engine's structure, its
components, or its auxiliaries, and operational changes include variations of
such parameters as fuel, air-to-fuel ratio, ignition or fuel injection timing,
and manifold inlet air temperature.
5.2.2 Physical Changes
In general, stationary engines are "customized" for a particular
application, and major changes are not made to their hardware and components
during their lifetime other than routine replacement of wearing parts(S-ll).
Practically, then, modifications will usually occur only when a replacement
part has been altered from the original design, and the use of this part
causes the engine to produce greater emissions than the original
5-3
-------
configuration. Manufacturers have, in fact, made changes in recent years,
primarily to improve fuel economy and smoke characteristics of their
engines(12,13). For example, one manufacturer of medium-bore diesel-fueled
engines has made four injector design changes since 1962(14). Although many
of these changes may reduce HC, CO, or smoke levels, they frequently will
increase NOx levels. Hence, if a user replaces an old part with a newer,
modified one, he may be making a change that is a "modification" as defined
by Federal Regulations. Examples of such changes are discussed in the
following paragraphs.
The design of various engine components can significantly affect
emissions from an engine. For example, the cylinder head shape, the
compression ratio, or the shape of the piston may be altered to enhance
combustion(15). Since improved combustion frequently results in higher
temperatures within the cylinder, these design changes can cause NOx
emissions to increase. Another alteration is to move the uppermost piston
ring to a higher location on the piston. This change reduces the quench
layer, and hence hydrocarbon formation, but generally does not affect
emissions of the other pollutants. One study, however, indicated that moving
the uppermost piston ring higher on the piston and increasing the injector
rack setting substantially reduced CO in addition to decreasing HC and caused
the NOX level to increase(16).
Engine components used for the intake of fuel and air or the exhaust
of combustion products may also influence emissions. The number and size of
the holes in a fuel injector as well as the sac volume (see Figure 4-55) have
an effect on emissions(17). If the fuel injector change improves the
combustion characteristics, NOx emissions may increase. Decreasing the sac
volume, in general, reduces HC emissions without affecting the other
5-4
-------
pollutants. Fuel pump characteristics and intake valve camshaft design
determine injection rates, which in turn affect combustion characteristics
and, hence, emission rates (see the discussions on injection timing and
internal EGR in Chapter 4).
In spark-ignition engines the carburetor, intake manifold, and intake
valves (location and number) affect the mixing of fuel with air (both local
and overall air-to-fuel ratio, see Sections 3.2.1 and 4.1 through 4.4).
Consequently, changes to these systems may increase the emissions from the
engine, particularly NOX. The location, size, and number of ports in the
liners of a two-cycle engine influence the amount of air in the cylinder and
its manner of mixing with the fuel(18). Similarly, blowers, turbochargers,
and aftercoolers determine the delivery rate and temperature of the inlet air
or the air and fuel mixture (see Sections 3.2.1 and 4.1 through 4.4).
The number and location of exhaust valves also affect emissions. For
example, a manufacturer of two-cycle engines uses heads with four exhaust
valves on engines that must meet EPA or California heavy-duty vehicle
emission standards, but retains the original two-valve design for the
uncontrolled stationary counterparts of these truck engines(19). Variations
in valve overlap (four-cycle) or port overlap (two-cycle) can alter the
emission characteristics of an engine, too. The amount of overlap affects
the amount of oxygen available for combustion as well as the temperature of
the air or the air and fuel mixture. Most of these items could be changed by
the user on an existing engine and, therefore, might be a modification in the
present context.
Large and medium-large engines (>500 hp) that are intended for
stationary applications are occasionally sold without much of the auxiliary
equipment, such as radiators, turbochargers, aftercoolers, etc. Therefore,
5-5
-------
if the purchaser does not use a correctly-sized cooling system (radiator or
cooling tower), or an undersized aftercooler on a turbocharged unit, the
engine could run hotter than specified by the manufacturer. In either case,
the temperature of the mixture in the cylinder would be higher than specified
by the manufacturer, and hence the NOX emissions would also be higher. In
other words, the use of auxiliary equipment that causes peak cylinder or
inlet manifold air temperatures to be higher than specified by the
manufacturer could be considered a modification.
Facilities with many engines usually have a staff of engineers and
technicians that are skilled in the operation and maintenance of their
engines. These users may occasionally change portions of an engine to
eliminate problems with the engine or to improve its performance. Presumably
these alterations involve only peripheral equipment such as radiators,
lubricating oil feed systems, etc., and hence have, at most, a secondary
effect on emissions. One user, for example, was able to reduce visible
emissions from large, two-stroke natural gas-fueled engines by changing the
lubricating oil feed system(20). it is unlikely, however, that these users
would redesign such emission sensitive components as the fuel injector,
intake valve or port geometry, piston head shape, cylinder head shape, etc.
5.2.3 Operating Changes
As discussed in Sections 3.2.2.3 and 4.1 through 4.4, emissions from
an engine depend, in part, on its operating conditions. Some of these changes
are a consequence of the normal operation of an engine over the range of
conditions for which it was designed. For example, variations in torque,
speed, output power, or air-to-fuel ratio that are within the engine
specifications belong in this category of changes. In contrast, injection or
5-6
-------
Ignition retard or advance from the manufacturer's specification or
purposeful alteration of the air-to-fuel ratio do not follow from typical
usage patterns and, therefore, could be considered modifications.
The use of fuels not specified by the manufacturer for a particular
engine could also constitute a modification, Such a condition would arise,
for example, 1f an owner or operator began to fuel an existing engine, which
1s specified to burn only No. 2 dlesel oil, with No. 6 o1l.l/ In general,
though, manufacturers specify a range of fuels to be used 1n their engines.
In fact, precombustlon chamber engines are specifically designed to
accommodate a variety of light fuels and dual-fuel engines are designed to
operate In either of two modes: full diesel or dual-fuel operation.
Therefore, although fuel type and quality can effect emissions (Section
3.2.2.3.4), the use of alternative fuels would constitute a modification only
if the change were to a fuel not originally intended for use in the engine
when it was manufactured.
5.3 RECONSTRUCTION
A reconstructed engine, as discussed at the beginning of the chapter,
is essentially an engine which has undergone a major rebuilding when it would
otherwise have been scrapped or replaced with a new engine. It is difficult
to apply the definition of reconstruction ("...replacement of a substantial
majority of the existing facility's components...") to a stationary engine
because substantial portions of large- and medium-bore engines are replaced,
in practice, as a matter of routine maintenance. That 1s, stationary (and
industrial mobile) engines are currently given several major overhauls during
I/At the present time there is only one known domestic installation of
engines burning residual fuel.
5-7
-------
the normal life of the engine without any stimulus from emission regulations.
The problem, then, is to distinguish routine maintenance and overhaul from
the rebuilding of an engine to avoid a standard of performance.
Large engines receive several minor and major overhauls during the
lifetime of the engine(21-24). Minor overhauls are performed after 8,000 to
16,000 hours and include new rings, new or reground valves, and occasionally
new or rebuilt pistons. After 20,000 to 80,000 hours of continuous service
these large-bore engines are given a major overhaul.^/ At this time the
cylinder liners (which are removable from the engine block) are replaced in
addition to the items covered under a minor overhaul. Generally, the
cylinder head, the crankshaft, rods, main bearings and rod bearings will last
indefinitely; nevertheless, these parts do sometimes fail, usually due to
lubrication failures, which can occur when the oil becomes contaminated, or
to extensive operation at overload conditions. However, the basic engine
frequently remains in service indefinitely.
In reality, several factors discourage perpetuation of these engines.
First, it can become difficult to obtain manufacturer service and replacement
parts for older engines. Second, power requirements change with time and
frequently dictate a larger engine or an alternative power source (e.g., a gas
turbine). This is especially true with generator and compressor applications
which are the major uses of large-bore engines. Third, improvements in engine
design have led to lower fuel consumption in newer engines, which favors
replacement of older, less efficient engines given the ever-rising costs of
fuels.
2/
-'Twenty-thousand'hours for large-bore, medium-speed (1200-rpm) spark
ignition engines and 80,000 hours for large-bore, low-speed (<400-rpm)
compression ignition engines.
5-8
-------
Medium-large power (>500 hp but <100 hp/cyllnder) diesel-fueled
engines also receive minor and major overhauls during their life(25-28).
These overhauls, however, are different than overhauls performed on large-
bore engines because they are patterned after maintenance practices used by
the truck Industry. The minor overhauls are termed "kit" overhauls and are
performed on the engine where installed. They include replacement of the
pistons, rings, liners, valves, and injectors. Major overhauls, on the other
hand, require removal of the engine and include replacement or regrinding of
the crankshaft and replacement of the main bearings, gears, cylinder heads,
and rods. Although major overhauls usually cost less than one-third the
price of a new eng1ne(29), the problems of servicing older engines, changing
power requirements, and Increasing labor costs encourage eventual retirement
of an engine.3/ However, since it is current practice to replace
substantial portions of these engines, 1t would be difficult to discriminate
between a major overhaul that was performed to avoid the purchase of a new
controlled engine from one that was performed 1n accordance with a routine
maintenance program.
5.4 SUMMARY
The preceding sections of this chapter have detailed those engine
changes which an owner or operator might undertake during Its lifetime. It
is noted that both minor and major overhauls are performed routinely and
usually the old parts are rebuilt to original specifications or replaced with
Identical new ones. Thus, such overhauls should be exempt from the
regulatory consequences of being a modification or reconstruction, despite
cost of an overhaul would have to be 50 percent of the cost of a
new engine for an overhaul to be considered a "reconstruction" in the
sense of having to meet a standard of performance(30).
5-9
-------
the fact that substantial portions of the engine can be replaced.
Alterations are occasionally made to major components, such as pistons,
cylinder heads, or turbochargers and aftercoolers, or to the operating
parameters, such as timing, air-to-fuel ratio (beyond the manufacturer's
specifications or purposeful alteration), or fuel (when not among those
specified for use by the manufacturer), and these changes can cause an
Increase In emissions for which a standard of performance applies. In this
case, these changes constitute a modification and would subject the engine to
a standard of performance.
5-10
-------
REFERENCES FOR CHAPTER 5
(1) Federal Register, Volume 40, Number 242, "Standards of Performance for
New Stationary Sources: Modification, Notification, and Reconstruction,"
Subpart A, 40 CFR 60.14, Tuesday, December 16, 1975.
(2) Ibid.. Subpart A, 40 CFR 60.7.
(3) Ibid., Subpart A, 40 CFR 60.15.
(4) Ibid.. Reconstruction.
(5) Ibid.. Subpart A, 40 CFR 60.7.
(6) Ibid.. Subpart A, 40 CFR 60.14(e).
(7) Ibid.. Subpart A, 40 CFR 60.14(c).
(8) Youngblood, S. B. (Aerotherm/Acurex) and C. L. Newton (Colt Engines),
private communication, February 11 and March 18, 1975.
(9) Youngblood, S. B. (Aerotherm/Acurex) and A. L. Foltz, Jr. (Enterprise
Engines), private communication, February 10 and March 18, 1975.
(10) Offen, G. R. (Aerotherm/Acurex), private communication (Information
supplementing July 26, 1974 Trip Report), July 26, 1974.
(11) Youngblood, S. B. (Aerotherm/Acurex) and Mr. Dick Kendal, Peterson
Tractor (Caterpillar dealer), Oakland, California, private communication,
February 14, 1975.
(12) Hanley, 6. P. (General Motors Corporation), "Marketing and Technical
Data on Reciprocating Engines for Stationary Applications," Attachment VI,
January 24, 1975.
(13) Shaw, J. C. (White Motors) letter to J. S. Fletcher (Aerotherm/Acurex),
dated November 14, 1974.
(14) Hanley, G. P., Reference 12, Attachment IV.
(15) Hanley, G. P., Reference 12, Attachment VI, p. 2.
(16) Hanley, G. P., Reference 12, Attachment VI, p. 16.
(17) Hanley, G. P., Reference 12, Attachment VI, p. 14.
(18) Hanley, G. P., Reference 12, Attachment VI, p. 17.
(19) Hanley, G. P., Reference 12, Statement, p. 7.
5-11
-------
(20) Offen, G. R. (Aerotherm/Acurex) and J. D, Martin (Union Carbide,
Seadrift Plant, Port Lavaca, Texas), private communication, December 1.1,
I •?/ "t»
(21) Offen, G. R. (Aerotherm/Acurex) and C. L. Newton (Colt Engines),
private cormunication, March 3, 1975.
(22) Youngblood, S. B. (Aerotherm/Acurex) and C. L. Newton (Colt Engines),
Reference 8.
(23) Youngblood, S. B. (Aerotherm/Acurex) and A. L. Foltz (Enterprise Engines),
Reference 9.
(24) Offen, G. R. (Aerotherm/Acurex), Reference 10.
(25) Youngblood, S. B. (Aerotherm/Acurex) and C. Marone, Watson and Meehan
(Cummins dealer), San Francisco, California, private communication,
February 11 and March 18, 1975.
(26) Youngblood, S. B. (Aerotherm/Acurex) and spokesman, Clementina Rental,
San Francisco, California, private communication, March 18, 1975.
(27) Youngblood, S. B. (Aerotherm/Acurex) and Mr. Dick Kendal, Reference 11.
(28) Youngblood, S. B. (Aerotherm/Acurex) and L. Zankich, N. C. Marine
(Caterpillar dealer), Seattle, Washington, private communication, February 12,
i y / D •
(29) Offen, G. R. (Aerotherm/Acurex) and C. L. Newton (Colt Engines), private
communication, March 3, 1975.
(30) Federal Register. Reference 1, CFR 60.15.
5-12
-------
CHAPTER 6
ALTERNATIVE EMISSION CONTROL TECHNIQUES
This section presents alternative emissions control techniques that
can be used to reduce NO emissions. Section 6.1 is a brief review of
A
the control systems that were discussed in Section 4.4, indicating which
of the techniques are considered technically viable and effective. In
Section 6.2, the effectiveness of these techniques (percentage NO re-
A
duction) and their associated fuel consumption penalties are summarized.
Section 6.3 discusses the research and development effort that engine
manufacturers estimate is necessary to achieve NO reductions from their
A
products.
6.1 REVIEW OF CANDIDATE NOV CONTROL TECHNIQUES
A
The NO emission control techniques discussed in Section 4.4 are
A
summarized in Table 6-1. In general, these techniques require an adjust-
ment of the engine operating conditions or addition of hardware, or a com-
bination of both. Although these techniques are listed separately, they
can be used in combination (see Sections 6.2 and 4.4.10). The control
techniques described in Table 6-1 can be divided into three groups based
on the relative ease with which they can be implemented. This section
describes the techniques within each group and indicates which control
approaches are considered technically viable. These techniques will form
6-1
-------An error occurred while trying to OCR this image.
-------
the basis for establishing standards of performance for exhaust NO
A
emissions.
The first group of controls Includes retard, air-to-fuel changes,
decreased manifold air temperature, and derating. As discussed 1n Section
4.4, these techniques, or some combination of them, appear to be the most
promising NOX control approaches based on their demonstrated effective-
ness and relative ease of Implementation. They generally require only an
operational adjustment or resizing of some components (turbocharger,
aftercooler, etc.). The exception 1s derating, which would require addi-
tional units to compensate for decreased horsepower. In general, these
techniques will increase operating costs because of increases 1n fuel con-
sumption. Engine manufacturers have also indicated that changes in engine
operating conditions may lead to increased maintenance. For example, one
manufacturer established that 4 degrees of ignition retard on a dual-fuel
engine resulted in a 24 percent decrease in the service life of the ex-
haust valves (see Section 4.4.2). No other data have been presented, how-
ever, that show adverse effects on engine components for the promising
control techniques.
As discussed in Section 4.4, the degree of application of any of
these techniques will be limited by practical considerations. For large
amounts of ignition/injection retard, fuel penalties are large and exhaust
temperatures increase. This may cause reduced exhaust valve life or re-
quire valve material that can withstand the elevated temperatures
(>1200°F).
Similarly, the amount of air-to-fuel (A/F) change is limited by
misfiring or detonation. In addition, mutually aspirated engines may
6-3
-------
experience large Increases 1n fuel consumption particularly as A/F ratio
1s decreased from stolchiometrlc.
Manifold air cooling is limited by the ambient air temperature. At
maximum ambient conditions, manifold air temperatures with a reasonably
cost-effective heat exchanger can practically be reduced (after turbo-
charging) to 15 to 20°F above the ambient temperature.
The primary limitation of derating is the competitive disadvantage
it places on engines that must compete with non-derated models. For
example, applications consisting of several engines may require additional
units to satisfy total power requirements as a result of derating. Never-
theless, the data in Section 4.4.1 have shown that a small amount (40
percent) of derate can effect a large NOX reduction 1n two-stroke
engines. Derate, however, is generally less effective In four-stroke
engines.
As discussed earlier, manufacturers may apply these control tech-
niques alone, or in combination, to achieve a given emission level. The
particular combination and extent of control applied will largely depend
on the engine model. As was shown in Section 4.4.10 this approach 1s
effective and can achieve low NO emission levels.
A
The second group of controls Includes exhaust gas reclrculation
(E6R) and combustion chamber modifications. These controls have demon-
strated effective NOX reductions but will require additional development
and durability testing. In contrast to the techniques discussed in the
first group of controls, E6R will require additional hardware: plumbing
to recirculate the exhaust and a heat exchanger to cool the recirculated
gas. Although a limited number of EGR tests have shown effective NO
A
reductions (see Section 4.4.6), there is concern that the necessity for
6-4
-------
cooling the redrculated gas will lead to fouling the flow passages of the
cooling heat exchanger as well as the engine turbocharger and after-
cooler. The tests conducted to date on large-bore engines have been of
short duration ( < 100 hours). Thus, longer testing will be necessary to
establish the effects that cooling the recirculated gas will have on
maintenance practices. One manufacturer of heavy-duty diesel truck
engines has rejected E6R because of excessive fouling (< 200 hours), and
another manufacturer believes the application of EGR must be limited to
non-aftercooled turbocharged and naturally aspirated engines with full-load
/I p^
EGR cutoff to prevent excessive smoke (>10- percent opacity^ ' '.
(EGR has been successfully applied in combination with other techniques on
gasoline-fueled engines.) Therefore, large-bore engine manufacturers will
require additional testing of this technique to establish its effect on
maintenance practices. Nevertheless, this technique is considered a tech-
nically viable NO control.
A
Data from smaller-bore diesel engines indicate that combustion
chamber shape has a significant effect on NO emissions (see Section
rt
4.4.8). However, none of the domestic large-bore engine manufacturers has
any experience with new or modified combustion chamber geometries. Manu-
facturers have estimated that an extensive development program of 3 to 5
years would be required to establish the emissions benefits of such a
major engine redesign. Since there is little experience with this control
approach, it is not considered a feasible near-term alternative at this
time.
The third group of controls includes water induction, increased
speed, and catalytic reduction. These techniques have serious technical
or cost limitations, have not been shown effective, or have little
6-5
-------
experimental data to demonstrate their effectiveness. For example, water
Induction has shown significant NOX reductions (see Section 4.4.7), but
deposit buildup from untreated water, combustion products, and lube oil
fouling have led manufacturers to reject this NOX control approach.
With respect to speed changes, limited data Indicate that a speed Increase
of 10- to 15- percent (with a decrease 1n torque to maintain the same
power) results 1n NO reductions. This technique may be effective for
A
electric generation applications because these engines operate at a con-
stant speed. (New units would use different speed exchange gears than In-
stalled units). However, this approach could not be applied to gas
engine driven compressors because engine speeds are matched with compres-
sor operating speeds, varying as much as 10 percent in the course of normal
operation. Since the largest number of new large-bore engines are sold for
compressor applications in oil and gas production and transportation, this
technique is impracticable. Similarly, limited data exists for catalytic
reduction of NOV emissions from 1C engines. As discussed in Section
A
4.4.9, this approach would be difficult to apply to the oxygen-rich ex-
hausts of large-bore engines and probably would be very expensive (see
Section 8.2.3).
Therefore, retard, air-to-fuel changes, derating, manifold air tem-
perature reduction, exhaust gas recirculation, and combinations of these
techniques are considered technically viable NOX control approaches.
The following section summarizes the NOX reductions and fuel penalties
associated with these techniques.
6-6
-------
6.2 EMISSION REDUCTIONS AND FUEL PENALTIES FOR CANDIDATE NO CONTROL
A
TECHNIQUES
In general, application of the technically viable NOX controls is
accompanied by increased fuel consumption. The NO reductions and fuel
A
penalties for these techniques were discussed in Chapter 4.4 and are sum-
marized in Table 6-2. In practice, an engine user may apply more or less
of these controls, either alone or in combination in an effort to reduce
NO emissions with a minimum increase in fuel consumption. Conclusions
A
based on Table 6-2 are summarized below by fuel type.
Diesel
The data show that retard alone, or in combination with either in-
let air cooling or air-to-fuel changes, is an effective approach for
diesel engines achieving NO reductions of 28 to 65 percent. EGR also
A
demonstrates a significant reduction in NO (33 percent) accompanied by
A
only a 1-percent fuel consumption increase. With the exception of two
data points, these NO controls result in less than a 10-percent fuel
A
penalty for all diesel engines.
Dual-Fuel
A number of control techniques prove effective when applied to
dual-fuel engines. Retard and air-to-fuel changes register the largest
NO reductions (up to 73 percent). Fuel penalties are always less than
A
10 percent and in most cases less than 5 percent.
Gas
Similarly, several controls or combinations of controls were effec-
tive on gas engines. Maximum NO reductions ranged from 40 to 80 per-
A
cent accompanied by fuel penalties of 5 to 12 percent. As discussed in
Section 4.4.3, changes in air-to-fuel ratios are particularly effective in
6-7
-------An error occurred while trying to OCR this image.
-------
reducing NOX from gas engines. This is because NO emissions from
A
spark-ignition engines are very sensitive to air-to-fuel ratio. In fact,
investigators have shown that derating, retard, and manifold air cooling
achieve NOX reductions from gas engines by, in effect, changing the
air-to-fuel ratio^3^.
Ranking of Controls
The NOX reductions summarized in Table 6-2 are reorganized in
Table 6-3. The techniques have been grouped into three levels of NO
A
reduction: 0 to 20 percent; 20 to 40 percent; and greater than 40 per-
cent. This format readily identifies which control techniques have estab-
lished given amounts of N0x reduction. The table indicates that only
retard and retard plus air-to-fuel ratio changes achieve NO reductions
rt
greater than 40 percent for diesel and dual-fuel engines. Several tech-
niques, however, exhibit N0x reductions greater than 40 percent in
natural gas engines. These results are aggregated from a number of engine
tests, and not all engines within each fuel category have demonstrated the
N0x reductions shown at each level. Nevertheless, several different
techniques applied separately or in combination have been shown effective
in achieving significant NOX reductions (20 to 60 percent).
These techniques are summarized in Table 6-4 for specific levels
(20, 40, and 60 percent) of NOX reduction with the corresponding range
of increases in brake-specific fuel consumption. This table illustrates
that a number of techniques are capable of achieving a 20-percent NO
/\
reduction for engines of all three fuels, and fuel penalties are generally
less than 5 percent. Fewer controls have been demonstrated for a 40-
percent N0x reduction, and fuel penalties are larger, exceeding 10
6-9
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
percent 1n two cases. In general, combinations of retard, air-to-fuel
ratio changes, and manifold air cooling were required for engines of any
fuel category to achieve a 60-percent NOX reduction. Fuel penalties
are highest for diesel engines, exceeding 20 percent. Fuel penalties are
more scattered for dual-fuel and natural gas engines, ranging from 1 to 22
percent but, in general, exceeding 5 percent. The data indicate that re-
tard is the most effective N0y control for diesel and dual-fuel engines,
A
and that air-to-fuel changes are most effective for dual-fuel and natural
gas engines. The cost impact of these alternatives for NOX emission
reductions will be discussed further in Chapter 8.
6.3 ESTIMATED TIME TO IMPLEMENT N0v CONTROL TECHNIQUES
A
Because manufacturers of large-bore engines are committed to a
particular design approach, they conduct extensive research, development,
and prototype testing before releasing a new engine model for sale. Con-
sequently, these manufacturers will require some period of time to modify
or reoptimize and test engines to meet standards of performance. In
general, this requirement consists of time to accomplish the following
tasks:
1. Establish baseline (uncontrolled) emissions for all engine
models
2. Establish the effectiveness of various techniques on each model
and size
3. Perform durability tests on selected engines for certain
techniques
4. Engineer any necessary redesigns, including new patterns and
tooling
6-12
-------
As shown In Section 4.3, engine manufacturers have obtained uncontrolled
emissions for about 80 percent of all dlesel, dual-fuel, and natural gas
engine models. Furthermore, engine manufacturers have expended consider-
able effort to Investigate potential NOX emission controls. Therefore,
the time requirements for areas 1 and 2 will be primarily to investigate
the effectiveness of applying a given control to a particular engine model
to meet a specified emission level.
Manufacturers have also indicated that a minimum durability test of
2000 hours is required for any engine model whose operating or combustion
characteristics are modified. For major design changes, on the other hand
(e.g., combustion chamber redesign), they estimated a need for 8000 hours
(about 1 year) of testing.
Finally, control techniques that would require special hardware
and/or modification of existing engine hardware would involve some engi-
neering and pattern and tooling efforts as part of their development. EGR
and combustion chamber redesign are techniques that would require this
development.
Table 6-5 estimates the time required to accomplish the four tasks
for all of the candidate control techniques identified in Section 6.1.
The entries in this table are based on the following factors:
• An estimate of 9 months is required for all engines that need
controls. Since not all manufacturers own their own emission
measurement equipment (see Section 4.2.4) or have data on the
emissions of their current production engines, 6 months have
been allowed for the procurement of this equipment, staff
training, and the measurement of baseline emission data on a
variety of engine models. An additional 3 months are allotted
6-13
-------An error occurred while trying to OCR this image.
-------
to analyze the baseline results, consider the effectiveness of
the various controls available for each engine, and decide on a
control strategy.
If problems are encountered during the long-term durability
tests, the problems need to be corrected and some of the tests
rerun. Therefore, 12 months have been allowed for durability
testing of major changes. Since Internal E6R is not as signif-
icant a change as external EGR or a chamber redesign, only 8
months were allotted for durability testing of this technique.
Also, since there is no reason to expect maintenance or dura-
bility problems with an engine that is operated at less than
full capacity, no time is allocated to durability testing of a
derated engine.
Retard and air-to-fuel ratio adjustments require no new hard-
ware and could be implemented after durability testing. For
EGR (internal and external) the time to assemble new production
equipment was scaled down from the comparable time for a com-
bustion chamber redesign, again after considering the relative
complexities of the various techniques.
The time to develop a combustion chamber redesign for a large
engine was estimated for large-bore engine manufacturers at 3
to 3-1/2 years, exclusive of durability testing and time to
assemble the necessary production tooling(4'5). This esti-
mate, in one case, is for a redesign of the valves, injectors,
and piston head and is based on the use of twice the manufac-
turers 's current R&D resources for this effort^). Develop-
ment times for modified injectors, EGR, and water
6-15
-------
induction were estimated by considering the potential technical
difficulty with these systems relative to those of modified
chambers.
In summary, changes to engine operating conditions or the addition
of existing hardware could be implemented in 9 to 15 months, and a major
engine combustion system redesign would take 5 to 6 years.
6-16
-------
REFERENCES FOR CHAPTER 6
(1) S. B. Youngblood (Acurex) and Hanley, 6. P. (General Motors Corporation).
Private Communication. May 19, 1975.
(2) Henderson, R. D. (Caterpillar Tractor Co.). Commentary on Diesel
Exhaust Emission Control System. November 14, 1974.
(3) Dletzmann, H. E. and K. J. Springer. Exhaust Emissions from Piston
and Gas Turbine Engines Used 1n Natural Gas Transmission.
Southwest Research Institute. AR 923. May 19, 1975.
(4) Fletcher, J. S. (Acurex) and C. L. Newton (Colt Industries).
Private Communication. November 3, 1975.
(5) Offen, G. R. (Acurex) and F. S. Schaub (Cooper-Bessemer). Private
Communication. October 13, 1975.
6-17
-------
CHAPTER 7
ENVIRONMENTAL IMPACT
This chapter considers the environmental impacts, both beneficial
and adverse, associated with the control of exhaust emissions from stationary
reciprocating internal combustion (1C) engines. The candidate emission
control systems are discussed in Chapter 4 and are designed primarily
to reduce NOX emissions from large-bore stationary 1C engines. The major
emphasis of this chapter will be to identify the incremental impact as
compared to uncontrolled engines of these emission control systems on
air, water, solid waste, noise pollution, and energy consumption. In
addition, secondary effects resulting from these control systems (e.g.,
increased HC or CO emissions as a consequence of reducing NOX emissions)
will be identified. The incremental impact of these systems will be related
to uncontrolled engines since virtually no stationary 1C engines are currently
subject to emission regulations other than for smoke.
7.1 AIR POLLUTION IMPACT
The discussion in Section 9.1 indicates that currently installed
stationary 1C engines contributed approximately 8.4 percent of the total
U.S. emissions of NOX. Furthermore, large-bore engines (greater than
350 CID/cyl) contribute approximately three-fifths of the NOX emissions
from stationary 1C engines, or nearly 5 percent of the nationwide burden.
7-1
-------
The projected impact of a standard for large-bore stationary engines on
nationwide emissions will be addressed in Section 7.6. This section will
evaluate the incremental impact of the candidate NOX emission control
systems (identified in Chapter 6) on ambient air quality. To do this,
ambient air concentrations of NOX, CO, HC, SOX, and particulates resulting
from emissions from stationary engines will be estimated using an atmospheric
dispersion model. Calculations will be performed for both uncontrolled and
controlled engine emissions under adverse meteorological conditions. These
concentrations will then be compared with corresponding ambient air quality
standards. The following sections will briefly discuss model plant charac-
teristics, the atmospheric dispersion model used, and the results of the
calculations.
7.1.1 Model Plant Characteristics
Data presented in Section 4.3 indicate the following range of uncon-
trolled emissions for large-bore 1C engines:
Emissions, __g/hp-hr
NOx CO HCT
Diesel
Dual Fuel
Gas
5 - 18
4 - 13
8 - 29
0.5 - 8.5
1.2 - 4.5
0.2 - 29
0.1 - 1.0
0.8 - 5.7
0.3 - 5.8
In general, CO is quite low in comparison with controlled mobile
diesel and gasoline engine emissions (the current federal CO standard
for heavy-duty vehicle diesel engines is 40 g/hp-hr; the California 1977
standard is 25 g/hp-hr). Total hydrocarbon emissions for large-bore diesel
engines are about one-half those from currently controlled heavy-duty
7-2
-------
vehicle diesel engines. Also, these levels are about one-half the levels
that will be emitted when automotive (light-duty vehicle) engines are
controlled to comply with the ultimate statutory limit of 0.41 grams per
mile specified in the 1970 and 1977 Amendments to the Clean Air Act.
Four typical large-bore engine configurations were chosen for the
atmospheric dispersion calculations. In addition, a compressor station
was simulated by modeling the dispersion from a row of seven 1500-hp engines.
Table 7-1 presents the physical dimensions and emission source characteristics
of all these engines. The NOX emission rates are given for sales-weighted
uncontrolled levels (see Section 4.3.4) and for three levels of control cor-
responding approximately to 20, 40 and 60 percent NOX reductions from
sales-weighted uncontrolled levels. The CO and HCT levels correspond
to values that most engines can achieve when controlled for NOX. (See
Figures 4-52 and 4-53.)
An unfavorable feature of the designs of the five prototypes is
that the height of the exhaust discharge is not significantly greater
than surrounding structures. This causes aerodynamic complications which
can seriously interfere with the rise of the effluent plume, thereby
producing significantly higher ground-level concentrations.
7-1-2 The Dispersion Model and Meteorological Considerations
The dispersion model used to analyze these plants is the single
source model developed by the Meteorology Laboratory, EPA. The dispersion
analysis was performed by Walden Research Division of Abcor, Inc. under
contract to the Monitoring and Data Analysis Division of EPA(1).
In this model all pollutants are assumed to display the dispersion
behavior of nonreactive gases. The predicted pollutant concentrations
7-3
-------An error occurred while trying to OCR this image.
-------
are based on the application of state-of-the-art modeling techniques, which
Implies reliability of the estimates to within a factor of about two(2).
The following assumptions are applied 1n the analytical approach:
1. There are no significant seasonal or hourly variations 1n emission
rates for these plants
2, The plants are located 1n flat or gently rolling terrain. In
restrictive terrain, the dispersion of effluents could be more
Impaired, resulting 1n higher ambient concentration levels
3. The meteorological regime is unfavorable to the dispersion of
effluents
4. All engines are continuously in operation, The 24-hour and annual
ambient concentrations for engines which do not operate con-
tinuously will likely be lower
Preliminary analyses Indicated that, for the stationary internal
combustion engines, critical meteorological conditions (I.e., those giving
rise to maximum short-term impact) consist of a combination of stable
atmospheric conditions and moderate wind speed. If such conditions occur
frequently at a given location, especially if they are combined with a high
directional bias in the wind, then longer-term impact (e.g., 24 hours and
annual) will also tend to be high.
Stationary internal combustion engines also have a geographical
tendency which must be considered. A large portion of these engines are
located in the midwestern to southwestern United States. Considering
these factors, Oklahoma City was chosen as the "worst case" location.
The model is programmed to use a previously determined set of dis-
persion conditions derived from the basic meteorological data for each
hour of the given year. The calculations simulate the interaction between
7-5
-------
the plant characteristics and these dispersion conditions to produce a
dispersion pattern for each hour. The computations are performed for
each point in an array of 180 receptors for any number of hours. In the
case of stationary internal combustion engines, the averaging periods
of interest are 1 hour, 3 hours, 8 hours, 24 hours, and annual, depending
on the pollutant of interest.
The prototype stationary internal combustion engines are modeled
with aerodynamic effects taken into consideration -- i.e., with a com-
putational procedure that accounts for the flowfield interactions between
the plume wind, and buildings. Such effects were found to be critical
due to unfavorable design (i.e., the exhaust discharge location is not
significantly higher than the surrounding buildings).
7.1.3 Results and Discussion of Dispersion Calculations
The results of the dispersion calculations are presented in Table
7-2. Maximum short-term average (CO, SOX, HC, and particulates) and annual
average (NOX) concentrations (at a distance of 0.3 kilometers from the
engine exhaust stack) are presented for the levels of emissions listed
in Table 7-1. Levels for CO and HCy that were used in the dispersion
calculation are typical of controlled engines (see Figures 4-52 and 4-53).
The potential environmental gain from a regulation based on the lowest
achievable levels of CO and HCj are discussed in Subsection 7.1.5 below. The
last column in Table 7-2 indicates the percent of hourly concentrations that
are greater than 50 percent of the maximum concentration during a 1-year
period. National primary and secondary ambient air quality standards
for these pollutants are presented in Table 7-3(3) for comparison with the
results in Table 7-2.
7-6
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
TABLE 7-3. NATIONAL AMBIENT AIR QUALITY STANDARDS!3)
Pollutant
N02
NMHCa
CO
S02
Parti cul ate
Averaging Time
and Method
Annual arithmetic
mean
3 hr: 6 - 9 a.m.
8 hr
1 hr
Annual arithmetic
mean
24 hr
3 hr
Annual arithmetic
mean
24 hr
Standard, yg/m3
Primary
100
160
10 mg/m3
40 mg/m3
80
365
75
260
Secondary
100
160
10 mg/m3
40 mg/m3
60
260b
1300
60
150
aNonmethane hydrocarbons
^As a guide to assess implementation for the annual standard
cAs a guide to assess implementation for the 24-hr standard
7-10
-------
All maximum concentrations for the model stationary internal combus-
tion engine are noted at extremely close-in distances (0.3 km). This is
due to the aerodynamic effects on plume rise as well as to the relatively
low height of the exhaust discharge. These concentrations are estimated
to fall off with greater distance at the following rates:
Distance from source, km Percent of the 0.3-km level
0.6 33
1.0 15
2.0 5
5.0 1.5
The following subsections summarize the comparison of the computed
concentrations (Table 7-2) with the ambient air quality standards (Table
7-3).
7.1.4 NOy Concentrations
The 1200-hp dlesel and 1500-hp gas engines are within ambient air
quality standards at both uncontrolled and controlled emission levels
when used individually. The larger gas and dual-fuel engines (4300 and
4400 hp), however, can exceed the NOX standard at their uncontrolled and
controlled levels but satisfy this ambient standard when their uncontrolled
NOX emissions are reduced 60 percent. The group of seven 1500-hp engines
(typical of a pipeline compressor station) may exceed the standard at the
uncontrolled level and the controlled levels. Thus, these large engines
could cause the ambient air quality criteria for NOX to be violated, at a
distance where public exposure is probable, even if emissions were reduced
significantly by controls. These violations would occur during the adverse
meteorological conditions postulated for this analysis. At typical
7-11
-------
uncontrolled levels (15 g/hp-hr), the ambient air criteria would be met at
all receptors which are further than about 1.0 km from the compressor
station. The comparable distance would be less than 0.5 km 1f the seven
engines were all controlled to the lowest level Indicated.
7.1.5 00 and HC Concentrations
Ambient air concentrations presented 1n Table 7-2 for CO are based
on representative levels of CO that all engines for which data are available
could achieve when controlled for NOX emissions. These calculations Indicated
that none of the engines exceed ambient air quality criteria for CO.
This is not unexpected, considering the relatively low emissions of CO
from these large-bore engines, which are designed to run at optimum thermal
efficiency.
The HC input data to the model were based on representative levels of
total hydrocarbons, including methane, that nearly all engines were shown to
achieve when controlled for NOX. Measurements by researchers indicate that
no more than 10 percent of the total hydrocarbon emissions from gas engines
and 25 percent of the total hydrocarbons from dual-fuel engines are reactive
HC compounds(4»5). When the concentrations reported in Table 7-2 are
multiplied by these fractions, only the 4300-hp dual-fuel engine and the
compressor station caused the nonmethane hydrocarbon standard to be exceeded.
Within 0.6 km of the station, the ambient air would not exceed the nonmethane
standard if the station were the only source. Moreover, if these gas engines
at the compressor station are assumed to emit total hydrocarbons at a level
of 2.0 g/hp-hr, as achieved by some gas engines, instead of the upper limit
value of 5.0 g/hp-hr used for these calculations, then this station, by
itself, would not cause the nonmethane hydrocarbon standard to be exceeded.
7-12
-------
Although some NOX emission control technologies cause HC and/or CO
emissions to increase, the data in Figures 4-52 and 4-53 show that these
increases are generally not large. Thus, for dlesel engines the NOX
reductions were generally achieved with HC and CO Increased less than 0.2 and
2 g/hp-hr, respectively (from uncontrolled levels of 0.1 to 1.0 g/hp-hr for
HC and 1 to 8 g/hp-hr for CO). Similarly, for dual-fuel engines the
Increases for both HC and CO are usually less than 2 g/hp-hr (from
uncontrolled levels of 1 to 6 g/hp-hr), while for gas engines they tend to be
about 1 g/hp-hr (from similar uncontrolled levels). Carbureted or naturally
aspirated gas engines, however, experience greater increases in HC and CO,
since they operate closer to stoichiometrlc conditions.
7.1.6 SOv Concentrations
The SOX emission rate listed 1n Table 7-1 for the dlesel engine
corresponds to a fuel sulfur content of about 0.22 percent by weight. This
level is typical for distillate fuels. By comparison, 0.5 percent 1s the
maximum allowable fuel sulfur content in several states (e.g., Connecticut,
Massachusetts (outside of Boston), and Rhode Island). Other areas restrict
the fuel to a sulfur content no more than 0.3 percent (e.g., Philadelphia,
Boston, and New Jersey). Since the predicted ground level concentrations
shown 1n Table 7-2 are far below both the primary and secondary ambient
air standards, they would still be well below these standards even 1f
the fuel contained 0.5 percent sulfur. However, if one scales these computed
results directly on emission rate (I.e., 1f one assumes that doubling
either the fuel sulfur content or the engine size results In double the
ground level concentration), any engine larger than 5700 hp could cause
a violation in the secondary ambient air quality standard (the 3-hour
7-13
-------
standard 1s exceeded first in this case) 1f 1t burned fuel with 0.5 percent
sulfur and 1f 1t were the only source.
7.1.7 Participates
Partlculates are generally only a problem with dlesel engines.
As one can see by comparing Table 7-2 with Table 7-3, the model dlesel
engine does not, by Itself, cause the local air to exceed the ambient
particulate standard. Even the largest dlesel engine produced 1n the
U.S., a 13,500-hp unit, would just meet the primary annual and the secondary
24-hour standard if 1t were the only contributing source.
7.2 WATER POLLUTION IMPACT
The promulgation of standards of performance should have no significant
Impact on water pollution. Only two control techniques could result In
an additional discharge of water -- Increased manifold air cooling and
water Induction. However, since most newly Installed engines use a closed
cooling system(6), similar to an automobile radiator, the Increased cooling
would be obtained by means of a larger radiator, a higher speed fan, or
both. Even where an open cooling system 1s used -- e.g., a once-through
cooling tower -- no additional water would be discharged to streams, rivers,
lakes, or other groundwater systems. The major Impacts from a cooling
tower are airborne mist and the demand for makeup water.
If water Induction were chosen by an engine manufacturer or user
as a means of meeting a potential standard, a demand would be created
for water, and there would be an airborne discharge of additional water
vapor with the engine exhaust. More importantly, however, the inducted
water would require treatment to reduce total dissolved solids (TDS --
see Section 4.4.7). One researcher who investigated the effects of water
7-14
-------
Induction on emissions and engine components Indicated that the water
available to him, which contained approximately 400 ppm of calcium
bicarbonate, was unacceptable due to the formation of combustion chamber
deposits during dally operation^).
Although raw water sources and municipal water quality vary
considerably, typical values are 64 mg/£ IDS for municipal water, and they
should typically not exceed 500 mg/£ (-500 ppm) of TDS(8). Any water that is
to be withdrawn for treatment and distribution as a portable supply 1n the
Ohio River Valley Water Sanitation District, cannot exceed IDS 750 mg/£ (at
any time). The corresponding regulation for industrial water supplies
(industrial cooling and processing) is 100 mg/£ TD$(9). Therefore, depending
on the source of water to be inducted into the engine, the user may need to
treat water ranging from less than 100 mg/£ TDS to several hundred mg/£ TDS
before inducting the water into his engine to reduce NOX.
The wastes associated with this treatment would be required to meet
local regulations before discharge to the environment, According to one
source(10), the supplier of delonized water or a de1on1zat1on system assumes
the responsibility of waste treatment and disposal of solids associated with
the water "softening" process. This disposal problan is not new to the
engine industry: cooling water for large-bore engines 1s presently treated
to reduce TDS which would otherwise accumulate as scale and reduce cooling
efficiency. According to the above source, users presently neutralize
their waste steams and then drain the treated water observing any local
regulations that are in effect. Typical treatment and disposal costs
are included In the costs of a water treatment facility as discussed in
Section 8.2.3. If a 10,000-hp engine in continuous operation (8000 hours
per year) used water induction at 400 ppm TDS and a water-to-fuel ratio
7-15
-------
of 1.0, 1t would add 5.8 metric tons/yr of solids to the stream of a waste
disposal facility.
7.3 SOLID WASTE DISPOSAL IMPACT
The demonstrated control technologies discussed in Chapter 6 should
also have no Impact on solid waste disposal. Some control techniques
(e.g., EGR or water induction) may require more frequent oil changes as
a result of oil contamination. According to one operator of large-bore
engines^1), lubricating oil is changed every 40,000 to 60,000 hours,
usually during a major overhaul. In between, the oil is checked about
once a month for water and fuel content. 011 1s added periodically to
replace that lost through blowby and leakage. Water communication necessitates
an oil filter change (normally changed at approximately 500-hour intervals)
due to damage of the disposal paper element by the water. Contamination
of the lubricant by the more volatile fuel requires an oil change because
of an explosion hazard. There is no sludge, however, associated with
oil filtering, and the discarded oil can be sold to an oil reclaimer for
$0.OS/gallon or burned as a waste fuel in a boiler.
7.4 ENERGY IMPACT
The annual incremental impact on national energy consumption can
be estimated based on historical production data (see Section 9.2) and
fuel penalties for candidate NOX emission controls (see Section 6.3).
Table 7-4(12) presents the annual Incremental energy impact after 5 and
10 years of a standard of performance. A 10-percent fuel penalty was
assumed for all new sales during this period. This assumption is conservative
since many techniques discussed in Chapter 6 achieved large NOX reductions
with less than a 10-percent fuel penalty. Thus, this evaluation 1s an
7-16
-------An error occurred while trying to OCR this image.
-------
upper estimate based on an unusually high fuel penalty, and it is not
intended to apply to a specific emission level or engine and control combina-
tion. Despite this assumption, Table 7-4 shows that the application NOX
controls would cause less than a 0.07-percent increase in nationwide consump-
tion of oil and gas after 10 years.
Annual production shown in Table 7-4 is computed as an historical
5-year sales average based on information presented in Section 8.1. These
data show that sales of stationary engine horsepower that would be affected
by a standard of performance have been essentially constant over the past
5 years and are assumed to remain relatively constant in the future.
Applications of these affected engines include electric generation for
diesel and dual-fuel engines, and oil and gas production or transportation
for natural gas engines. All of these applications are high usage, estimated
as 8000 hours per year. Typical fuel consumption values are based on
the data from Figure 4-13.
These assumptions result 1n an annual increase of domestic oil and gas
consumption of 0.0067 percent relative to 1972 total consumption. This
penalty would increase to 0.034 percent after 5 years and 0.067 percent after
10 years, again relative to the 1972 total consumption. If, however, total
consumption were assumed to increase by 3.5 percent per year, the effect of
a 10-percent fuel penalty would cause the increase in domestic oil and gas
consumption to be only 0.029 percent of the nation's fuel consumption 5 years
hence, and 0.048 percent of the total domestic usage in 10 years.
7-18
-------
7.5 OTHER ENVIRONMENTAL IMPACTS
7.5.1 Noise Impact
Fan noise levels from large-bore stationary reciprocating engine
installations could increase as a result of increased cooling requirements
if decreased manifold air temperature or precombustion chambers (which
have higher heat rejection) are used to reduce emissions. However, in
typical installations such as municipal generator plants, pipeline com-
pressor stations, or industrial process plants, such increases in the
noise level are likely to be insignificant in comparison to other sources
of noise (e.g., generators, compressors, pumps, or process blowers).
Although stationary engines are not specifically regulated noise
sources, seme engine installations mentioned above may be subject to Occupa-
tional Safety and Health Act (OSHA) regulations or EPA guidelines regarding
noise exposure limits for employees. Allowable noise levels and employee
exposure times are specified in parts 55 and 56 of the August 7, 1974,
Federal Register, Volume 39, No. 153. In any new Installation, measures
should be taken as required to meet these regulations.
7.5.2 Thermal Pollution Impact
The application of emission controls can increase fuel consumption
with no corresponding increase in power. This additional heat input is
discharged to the environment as waste heat either by the cooling system
or with the exhaust gases. In an attempt to identify the increment of
thermal pollution caused by emission control techniques, two model engines
are considered below.
Figure 4-51 Illustrates that the range of fuel penalties associated
with emission control techniques for large-bore engines 1s about 1 to
7-19
-------
10 percent, The following data Indicates, as a worst case example, the
additional heat discharged to the atmosphere as a result of a 10-percent
fuel penalty for both a 500-hp and a 13,500-hp engine 1n continuous operation
(800 hr) for 1 year, If a fuel consumption rate of 7500 Btu/hp-hr 1s
used for the uncontrolled unit, the heat rejection 1s increased by about
15 percent.
Heat Exhausted to Atmosphere as Increase Due to
Waste Heat from Uncontrolled Control,
Engine
500 hp
13,500 hp
Engine, Btu/yr
20 x 109
535 x 109
Btu/yr
3 x 109
81 x 109
7.6 OTHER ENVIRONMENTAL CONCERNS
7.6.1 Irreversible and Irretrievable Commitment of Resources
No irreversible or irretrievable commitment of resources is anticipated
as a result of promulgating standards of performance for new stationary
reciprocating engines other than that already discussed resulting from
increased fuel consumption, and, possibly, use of catalysts.
7.6.2 EnvironmentalImpact of DelayedStandards
The environmental impact of delaying a standard that is based on
emission levels achieved by presently available control techniques (see
Chapter 6) is essentially the same as the impact of no standard, since
stationary reciprocating engines are presently unregulated for exhaust
emissions except for local regulations for smoke and the Los Angeles area
NOX regulation. The control techniques which require additional develop-
ment (i.e., E6R on turbocharged engines, water induction, and combustion
chamber modifications) can potentially achieve lower emission levels than
those achieved by currently available techniques. However, the imposition
,7-20
-------
of standards based on currently available technology does not preclude
or Interfere with the continued development of the emerging techniques.
Moreover, the promulgation of a standard which prescribes an emission
level that is based on current technology to take effect several years
hence would not lead to wasteful commitment of resources (i.e., does not
force engine manufacturers to expend large sums of money to develop one
kind of engine or control system now and a completely unrelated kind for
the future); the technology is at hand to enable manufacturers to meet
a near-term standard, so that most R&D funds could be devoted to preparing
for a lower, future standard. Therefore, no environmental gain is achieved
by delaying promulgation of standards.
The impact of no standard relative to a standard based only on
readily available control technology (i.e., a near-term standard) 1s dis-
cussed below. The additional Improvement to the ambient air that can
be obtained 1n the future through the application of more stringent standards
based on emerging control technology 1s also considered.
7,6,3 Environmental Impact of No Standard
The estimated impact presented here for no standard is based on
the same reasoning used for the estimated energy impact (see Section 7.4).
Table 7-5(13) shows the impact of a NOX standard for 1 year's production of
diesel, dual-fuel, and gas engines that would be affected by a standard
of performance. Three possible levels of control (20, 40 and 60 percent)
are considered based on the alternatives discussed in Chapter 6. This
table shows that the impact of the standard could range from 0.06 to 0.19
percent of the total U.S. NOX emissions, depending on the level of control
Imposed.
7-21
-------An error occurred while trying to OCR this image.
-------
These impacts are based on the assumption of constant production
of diesel, dual-fuel, and natural gas engines.
Nevertheless, several projections conclude that total U.S. NOX
emissions fron stationary sources will increase at annual compounded rates
of 3.9 to 6.4 percent(14,15). These projecte(j .^^ ^ ^ ^
economic indicators which suggest a substantial industrial growth during
the next 10 to 15 years. The growth in stationary source emissions is
partially offset by declining NOX emissions from automobiles due to the
application of controls. When both trends are considered, total nationwide
emissions (stationary plus mobile) are projected to increase by a factor
of 1.2 to 1.4 by 1985. Under these circumstances, and assuming only a
20-percent NOX reduction, the incremental NOX emissions impact in 1985
for all new controlled large-bore engines (all fuels beginning in 1979)
would amount to about 0.27 percent of the 1985 U.S. NOX emissions fron
all sources. If a 60-percent reduction were required starting in 1985,
the differential impact would rise to about 0.93 percent in 1990.
Industry spokesmen report that 3 to 4 percent of the present population
of large-bore engines is being replaced each year. Assuming production
remains constant and that a performance standard for emissions is in effect,
then all of the present population of engines will be replaced with controlled
engines in 30 years. This new population of controlled engines (assuming
controls achieve a 60-percent NOX reduction) would effect a reduction in
total U.S. NOX emissions of 5.6 percent based on present (1974) total
emissions of NOX, or a 2-percent reduction assuming NOX emissions increase
3.5 percent per year over this period.
7-23
-------
REFERENCES FOR CHAPTER 7
(1) Modeling Analysis of the Ambient A1r Impact of Stationary Internal Com-
bustion Engines. Report prepared by Walden Research Division of Abcor, Inc.,
for the Source Receptor Analysis Branch, Monitoring and Data Analysis Division,
OAQPS, EPA. October 1975.
(2) Ibid, p. 1.
(3) Title 40 - Protection of the Environment. National Primary and Secondary
Ambient A1r Quality Standards, Federal Register. 36 (84): 8187. April 30,
1971.
(4) Dletzmann, H. E. and K. J. Springer. Exhaust Emissions from Piston and
Gas Turbine Engines Used 1n Natural Gas Transmission. Southwest Research
Institute. AR-923. January 1974.
(5) Newton, C. L. (Colt Industries), and G. R. Offen (Acurex/Aerotherm).
Private Communication, November 20, 1974.
(6) Standard Practice for Low and Medium Speed Stationary Diesel and Gas
Engines. Diesel Engine Manufacturers Association. New York, New York.
p. 70. 1972.
(7) Shaw, J. C. (White Motor Corporation). Emission Reduction Study on a
Carbureted Natural Gas Fueled Industrial Engine. Draft ASME Paper. White
Superior Division. November 1974.
(8) Fair, G. M., J. C. Geyer, and D. A. Okun. Water and Waste Water Engi-
neering. John Wiley and Sons, Inc. pp. 19-20. 1962.
(9) Wastewater Engineering. Metcalf and Eddy, Inc. McGraw-Hill Book Co.
1972.
(10) Youngblood, S. B. (Acurex/Aerotherm). Interoffice Memorandum. Water
Treatment System Costs. February 13, 1975.
(11) Youngblood, S. B. (Acurex/Aerotherm). Interoffice Memorandum.
October 6-21, 1975.
(12) Minerals Yearbook 1972. U.S. Department of the Interior, Bureau of
Mines. U.S. Government Printing Office, p. 910. 1974.
(13) Monitoring and Data Analysis Division, OAQPS, U.S. Environmental
Protection Agency. Computer Printout of Nationwide Emissions Report,
National Emissions Data System. January 10, 1975.
(14) Hopper, T. G. and W. G. Marrone. Impact of New Source Performance^
Standards on 1985 National Emissions from Stationary Sources, Volume I final
draft report. Prepared by TRC for the Emission Standards and Engineering
Division, OAQPS, U.S. EPA under contract 68-02-1382, Task No. 3.
February 16, 1975.
7-24
-------
(15) Habegger, L. J., R. R. CIHllo and N. F. Sather. Priorities and
Procedures for Development of Standards of Performance for New Stationary
Nationai
7-25
-------
CHAPTER 8
ECONOMIC IMPACT
This chapter is divided into four sections. Section 8.1 presents a
general profile of the domestic stationary reciprocating engine industry.
Emphasis is given to large-bore stationary engine manufacturers and their
major end uses, since proposed New Source Performance Standards (see chap-
ter 9) will affect primarily this industry. Section 8.2 is a detailed
cost analysis of the alternative N0x control techniques summarized in
Chapter 6. Section 8.3 is a discussion of potential costs to stationary
engine manufacturers and users arising from other environmental regula-
tions. The chapter concludes with a detailed analysis, in Section 8.4, of
the economic impact that could arise from the use of the alternative emis-
sion control techniques to meet a performance standard.
8.1 INDUSTRY PROFILE
This section presents a business profile of the 1C engine indus-
try. It is divided into five subsections. In Section 8.1.1 manufacturers
of domestic stationary engines are presented, and general production
trends and applications are discussed. In Section 8.1.2 the discussion is
focused on the large-bore engine manufacturers. A detailed analysis of
large-bore engine markets is given in Section 8.1.3. Finally, a discus-
sion of engine imports and exports is presented in Section 8.1.4.
8-1
-------
8.1.1 Structure of the Industry
The stationary 1C engine Industry consists of 40 firms operating 1n
four separate submarkets: (1) high-power, low-to medium-speed engines;
(2) medium-power, medium to high-speed engines; (3) low-power, high-speed
engines; and (4) very small one-cylinder gasoline engines (<20 hp). These
manufacturers were listed by submarket 1n Table 3-1. Figures 3-1 and 3-2
depicted graphically the relationships between the first three groups on
the basis of engine power. All 40 firms are listed In alphabetical order,
with addresses, 1n Table 8-1. In addition, engine descriptions are given
1n Table 8-2* ' by manufacturer for all but some very small gasoline
units.
Table 8-3^ " ' shows total U.S. production of 1C engines by type
of fuel (gasoline, diesel, and natural gas) for the years 1964 through
1975. These data are also presented in graph form, with a separate plot
for each fuel, 1n Figures 8-1 to 8-3^5"6^. These data represent produc-
tion of all engines, irrespective of whether they were used in mobile or
stationary applications. Production data of engines sold for stationary
uses in 1973 have been estimated by direct contacts with manufacturers
(see Table 3-2 and Section 8.1.3). These data have not been compiled
previously for public dissemination by any government agency or trade
association, nor is there any reliable basis for extrapolating the 1973
estimates to other years. The reader should note that these data are for
shipments of engines produced by U.S. manufacturers and do not consider
imports or exports. For a discussion of imports and exports see Section
8.1.4.
Several broad trends are indicated by these production data: (1)
Gasoline engine production grew slowly up to 1974, with the bulk of the
8-2
-------An error occurred while trying to OCR this image.
-------
TABLE 8-1. Continued
General Electric Company
Diesel Power Products
Transportation Business Division
2901 East Lake Road
Erie, Pennsylvania 16531
Brand Name: General Electric
Homelite Division of Textron
70 Riverdale Avenue
Port Chester, New York 10573
Brand Name: Homelite
Ingersoll-Rand Company
Engine Process Compressor Division
100 West Chemung Street
Painted Post, New York 14870
Brand Name: Ingersoll-Rand
International Harvester Company
10400 West North Avenue
Melrose Park, Illinois 60160
Brand Name: International
Jacobsen Manufacturing Company
1721 Packard Avenue
Racine, Wisconsin 53403
Brand Name: Jacobsen
John Deere OEM Sales
Moline, Illinois
Brand Name: John Deere
Kohler Company
Kohler, Wisconsin 53044
Brand Name: Kohler
McCulloch Corporation
6101 Century Blvd.
Los Angeles, California 90045
Brand Name: McCulloch
Murphy Diesel Company
5317 W. Burnham Street
Milwaukee, Wisconsin 53219
Brand Name: Murphy Diesel
O&R Engines Incorporated
3340 Emery Street
Los Angeles, California
Brand Name: O&R
90023
Outboard Marine Corporation
Gale Products Division
100 Sea Horse Drive
Waukegan, Illinois 60085
Brand Name: Lawnboy
Sterling Engine Company, Inc.
3600 NW North River Drive
Miami, Florida 33142
Brand Name: Sterling
Stewart and Stevenson Services, Inc.
4516 Harrisburg Boulevard
Houston, Texas 77011
Brand Name: Stewart &
Stevenson
(Modified
Detroit-Diesel
Engines)
Tecumseh Products Company
Ottawa and Patterson Streets
Tecumseh, Michigan 49286
Brand Name: Tecumseh
Teledyne Continental Motors
Industrial Products Division
205 Market Street
Muskegon, Michigan 49443
Brand Name: Continental
Teledyne Wisconsin Motors
1910 South 53rd Street
Milwaukee, Wisconsin 53246
Brand Name: Wisconsin,
Hatz
Waukesha Engine Division
Dresser Industries Inc.
100 West St. Paul Street
Waukesha, Wisconsin 53186
Brand Name: Waukesha
White Construction
Equipment Division
White Motor Corporation
Hopkins, Minnesota 55343
Brand Name: Minneapolis
Moline
8-4
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
increased production occurring in marine and garden applications (small
engines, typically much less than 15 hp). The 1975 production, however,
declined coinciding with a general decline in the economy. (2) The diesel
fuel category experienced a much more rapid growth rate up to 1974 but
also declined in 1975. Note that the 1975 sales were nearly evenly dis-
tributed among construction, agriculture, generator sets, and general in-
dustrial applications. (3) Natural gas engine production has been declin-
ing rapidly since 1966. According to industry representatives this sharp
drop in sales of gas-fueled engines reflects the decline in usage of 1C
engines for irrigation pumps and pipeline compressors^7). (Ninety per-
cent of the installed reciprocating engine horsepower in pipeline applica-
tions is spark-ignited; slightly over two-thirds of the remainder is dual-
(8}
fueled)^ '. This trend is the result of the unavailability of natural
gas for irrigation and the end of the nation's expansion of interstate
natural gas pipelines.
B.1.2 Large-Bore Engine Manufacturers
The large-bore stationary engine industry consists of nine firms
manufacturing engines in the range of 400 to 13,500 hp and 300 to 1200
rpm. In general, the engines produced by this industry are designed for
heavy duty operation at high power outputs and for long lifetime (hence
the low rpm ratings). The primary applications of these engines are for
(1) oil and gas production, (2) oil and gas transmission, (3) electric
generation and (4) standby service (nuclear power plants, standby genera-
tors, and floodpumps). These applications are discussed in more detail in
Section 8.1.3.
As shown in Table 8-4 the corporations manufacturing high-power
stationary engines are all highly diversified, and many are among the 500
8-15
-------
TABLE 8-4. MANUFACTURERS OF HIGH-POWER ENGINES
Manufacturer
Alco Power Inc.
Caterpillar
Cooper-Bessemer
Superior
Ajax
Dresser-Clark Division
Waukesha Division
Enterprise Engine Division
of DeLaval Turbine Co.
ElectroMotive Division
Fairbanks Morse Engine
Division
General Electric
Ingersoll-Rand
Parent Corporation
G.E.C. Limited
Caterpillar
Tractor Co.
Cooper Energy
Services
Dresser Industries
TransAmerica
General Motors
Corporation
Colt Industries
General Electric
Ingersoll-Rand
Rank in
"Fortune's 500"a
NAb
36
340
101
NMC
2
178
9
117
*Based on 1976 sales data . J , „
bNA = Not applicable, Alco Power Inc., Division White Industrial Power
was purchased by G.E.C. Ltd of Great Britain, a conglomerate of
engine manufacturers and related firms.
CNM • Not meaningful since TransAmerica is primarily a financial
corporation, rather than an industrial firm.
8-16
-------
largest industrial firms in the U.S. Seven of the nine manufacturers of
large-bore engines either market gas turbines or belong to a parent com-
pany which markets turbines through another division. Alco, General
Electric, and ElectroMotive manufacture engines primarily for the railroad
industry. General Electric has only recently begun marketing engines for
stationary applications, and primarily exports for drilling and auxiliary
power applications. Some of the firms also manufacture large-bore engines
for marine propulsion and mobile applications (e.g., off-highway construc-
tion vehicles). As shown in Table 8-5<10), all Of these organizations
are part of diversified parent corporations for which sales of stationary
reciprocating engines account for less than 15 percent of total revenues.
Since 1973, three firms -- Chicago Pneumatic, Nordberg, and
Worthlngton -- have ceased production of stationary reciprocating engines.
Conversations with industry spokesmen indicate that the Industry has stab-
ilized, and no additional dropouts are anticipated.
Manufacturing plants employ from 600 to 10,000 people each. In the
larger plants that produce engines for other than stationary applications,
approximately 600 to 2000 employees per plant can be allocated to manu-
facturing stationary engines. Three of the nine plants which produce
large engines employ less than 1000 people.
Sales of high-power stationary engines are comparatively unconcen-
trated; the largest manufacturer has approximately five-times the dollar
volume of the smallest. Unit sales range from about 20 to 200 units per
year for stationary applications depending on the manufacturer.
8-17
-------
TABLE 8-5. PARENT CORPORATION DATA
Total Sales3 Rank in 1976
Parent Corporation ($ millions) Fortune 500
Dresser Industries $ 2,232 101
(Dresser-Clark,
Waukesha)
Ingersoll-Rand 1,922 117
Colt Industries 1,267 178
(Fairbanks Morse)
Cooper Industries 554 340
(Cooper, Superior,
Ajax)
General Motors Corp. 47,181 2
(ElectroMotive Div.)
Caterpillar Tractor Co. 5,042 36
TransAmerica NM^ NM^
(Delaval)
General Electric 15,697 9
All Firms $554-47,181 2-340
?Total sales are for 1976. (Reference 10)
NM = Not meaningful because TransAmerica is primarily a financial
corporation, instead of an industrial corporation.
8-18
-------
In general, firms which make smaller engines have a higher unit volume.
For example, the total engine sales of Caterpillar or Waukesha are many
times that of any other manufacturer. Furthermore, 90 percent of
ElectroMotlve's sales are for locomotive power, and their stationary sales
are comparable 1n number and dollar volume to their competitors.
Industry representatives report that purchased materials account
for approximately 60 percent of the cost of an engine. Most manufacturers
purchase large forglngs In finished or semifinished form, while some, for
example, DeLaval and Superior D1v1s1on/Cooper Energy Services, have their
own facilities and make their own large castings. Turbochargers, after-
coolers, radiators, fuel systems, governors, oil pumps and other auxiliary
systems are generally purchased from outside vendors.
Sales of high-power, low-speed engines are usually made through
competitive bids that Include a fuel consumption guarantee and the cost of
providing a complete engine Installation. This most commonly Includes
foundations, cooling systems, and other components such as generators and
compressors. Often the cost of the basic engine 1s as little as 40 per-
cent of the total'11). With the exception of standby service, most
applications of large-bore engines are high usage (>2000 hr/yr). Thus
fuel and maintenance costs, rather than initial price, are the most sig-
nificant cost Items for the purchaser of an engine. Nevertheless, manu-
facturing Industry spokesmen Indicated that engine prices have risen
rapidly due to Increases 1n raw material and labor costs.
It 1s anticipated that the financial impacts on the manufacturers
of engines which must meet a standard of performance will be primarily in
the form of research and development costs rather than costs to modify or
retool the manufacturing process (see Section 8.4). Unlike standards of
8-19
-------
performance for new sources that are applied to typical manufacturing pro-
cesses, the ones envisioned for 1C engines are directed at the product of
engine manufacturers rather than at the plants. Hence, controlling engine
emissions will primarily involve modifications within the product. Al-
though attempts to reduce engine emissions may require additions or alter-
ations of one or two machines on an assembly line, the costs of retooling
that could be required for EGR systems, new fuel pumps and injectors, new
head designs, etc., can be considerable. Manufacturers have advised us
that they usually do not proceed with the acquisition of the new tools or
machines until the engine modifications habe been thoroughly tested on
laboratory engines and their durability verified'12'13'.
Research and development costs are normally deducted from income
during the year incurred and not amortized over several years production
of a specific engine; however, a company's engines are usually priced to
recover these overhead expenses. Spokesmen for large-bore engine manufac-
turers indicate that the R&D programs typically span 3 to 5 years but may
require as long as 10 years, with about 1 year of endurance testing re-
quired before any final design can be incorporated in a production
model^14). The final development work is often carried out in coopera-
tion with a well-known customer who receives the engines on an economi-
cally attractive basis. The level of development work that can be under-
taken, however, will vary depending on the manufacturer and the applica-
tions the manufacturer serves.
8.1.3 Stationary Markets for Large-Bore Reciprocating Engines
The principal applications of large-bore stationary reciprocating
engines are listed below:
8-20
-------
Application Type of Engine
Oil and Gas Production Gas and Diesel
Oil and Gas Transmission Gas
Electric Generation (baseload) Dual-Fuel and Diesel
Standby Service Diesel
Oil and gas production applications consist of engine-driven compressors
for field gathering, pressurization, storage, and distribution, and oil
and gas transmission applications for engine-driven compressors installed
on natural gas pipelines. Dual-fuel and diesel engines are purchased
primarily by municipal electric utilities for baseload (continuous) elec-
tric generation. Standby service includes emergency power for buildings
and hospitals as well as nuclear generating stations and flood control
pumps. Miscellaneous applications of large-bore engines include auxiliary
engines on construction equipment (e.g., hoists and dredges) material
handling, and pumping and electric generation at sewage treatment plants.
Sales of large-bore engines have increased slightly from 1- to
1-1/4-mniion horsepower per year over the past 5 years (1972 to 1976).
Approximately 20- to 30-percent of this horsepower 1s exported, primarily
as electrical generator units and engine/compressors for oil and gas pro-
duction. Exports have increased during the past 5 years, and account for
more than 50 percent of total sales for one domestic manufacturer. In
general, spokesmen for manufacturers have Indicated that rapid increases
in sales for export and standby applications have offset depressed sales
for natural gas pipelines and municipal electric utilities.
8-21
-------
The following discussion of engine application 1s subdivided by
fuel type since this classification provides a clear distinction of
end uses. Aggregated sales Information are presented below for large-bore
engines by fuel type. These data were compiled from responses of engine
manufacturers to a Section 114 Request for Information mailed 1n June of
1976 by EPA.(15) EPA has decided not to release data for Individual
manufacturers since manufacturers consider this Information proprietary.
8.1.3.1 Diesel Engine Markets
The primary high usage (large emissions Impact), domestic appli-
cation of large-bore dlesel engines during the past 5 years has been for
oil and gas exploration and production. These and other applications are
Illustrated 1n Figure 8-4^16'21), As this figure shows, the market for
prime (continuous) electric generation and other Industrial applications
all but disappeared after the 1973 oil embargo, but was quickly replaced
by sales of standby electric units for building services, utilities, and
nuclear power stations. The rapid growth 1n the oil and gas production
market occurred because dlesel units are being used on oil drilling rigs
of various sizes. Sales of engines to export applications have also grown
steadily since 1972, and are now a major segment of the entire sales
market.
Some degree of overlap between large-bore diesels exists for
petroleum exploration applications. Smaller (250 to 1000 hp) medium-bore
designs (e.g., Detroit Diesel, Cummins, and Caterpillar) are used on port-
able drilling rigs to drill or service 2500- to 5000-feet wells. These
rigs are trailer-mounted or helicopter-transported; therefore, small
lightweight (approximately 4000-lb) engines are favored. In addition,
8-22
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
utilization are shown for these two power sources, plus fossil steam and
nuclear steam, in Figure 8-5(23). Note that the figure is based on 1968
information; however, the relative comparison of these prime movers is
still valid. These results do not include the cost Impact of standards of
performance for gas turbines. This impact is discussed in Section 8.4.1.3
As of January 1, 1974, six new 1C engine generators were scheduled
to be added to commercial power supplies -- five (total 32,840 kW) 1n 1974
and one (4415 kW) 1n 1975<24). This compares with six engines (30,930
kW) added in 1973. The Federal Power Commission reports that as of April
1, 1977, 565,000 kW (39 units) of diesel and dual fuel generating capacity
were scheduled to be Installed 1n the period 1977 to 1986^25^ Since
most of the 1C engines used for electric power generation are owned by
municipal utilities, which are generally smaller than the investor-owned
utilities, 1t is possible that uncertainties in fuel availability and cur-
rent high Interest rates are preventing these smaller, municipally owned
systems from raising the capital necessary to expand their systems. A
spokesman for one manufacturer, however, stated that sales have picked up
as the demand for additional power has reached critical level. In addi-
tion, another source believes that an increasing number of engines will be
used for onsite power generation by municipalities and large industrial
electricity users^).
Large dlesels are also used 1n nuclear powerplants, since these
facilities are required to have emergency power available to flood the re-
actor core with water 1n the event of a reactor failure. Industry repre-
sentatives (both manufacturers and users) have indicated that the high-
power diesel engines have no effective competition for this market^27).
Due to the quick startup requirements for nuclear power (10 seconds and
8-25
-------An error occurred while trying to OCR this image.
-------
load capability 1n 30 seconds), this service 1s almost exclusively met by
large dlesel engines. Since safety regulations require that there be at
least two engines for each reactor, the best Indicator of future engine
needs by this market 1s a record of scheduled construction of nuclear
power reactors, Table 8-6 shows the number of reactors scheduled for com-
pletion-'. This Indicates a market for 338 to 382 high-power engines 1n
the next 10 to 15 years. Although recent difficulties 1n raising capital
and 1n proving the safety of reactors and spent fuel disposal have caused
utilities to delay over 40 percent of the units under construction or on
order and to cancel 5 to 10 percent, the Nuclear Regulatory Commission
(formerly the Atomic Energy Commission) continues to project that 102,000
nuclear megawatts will be constructed by 1980 and 250,000 by 1985(3°).
It should be recognized that the engines for a particular reactor may be
purchased up to a year or two before the reactor becomes operational.
TABLE 8-6. PLANNED CONSTRUCTION OF NUCLEAR REACTORS
Completed Scheduled as of January 1. 1974
During 1973 1974 1975 1976 1977 1978 1979 & later
Reactors 7 27 10 7 12 11 102
Radar power stations are also served by reciprocating engines which
maintain precise power characteristics over sizable load variations.
i/The source of this table^8* lists 169 units completed or scheduled
for completion after 1973. Another source mentions that 191 units
are currently under construction or on order.
8-27
-------
Engines have also been in demand by flood control districts for pumping
applications along the Mississippi Delta.
8.1.3.2 Markets for Dual-Fuel Engines
The concept of dual fuel operation was developed to take advantage
of both compression Ignition performance and inexpensive natural gas.
These engines have been used almost exclusively for prime electric genera-
(31-34^
tion. Figure 8-6v ' illustrates, however, that shortages of natural
gas and the 1973 oil embargo have combined to significantly reduce the
sales of these engines in recent years. For example, sales of dual-fuel
engines in 1971, 1972, 1973, 1974, and 1975 were 95, 74, 53, 17, and 35
units, respectively^ 5> ). As discussed above, spokesmen for engine
manufacturers stated that sales have recovered somewhat as demands for
power have become critical and firm commitments for fuel are established.
8.1.3.3 Markets for Natural Gas Engines
The primary application of large gas engines during the past 5
years has been for oil and gas production. Figure 8-7^37" 2', based on
manufacturer's data from responses to the June 16, 1976 Section 114 Re-
quest for Information, Illustrates that 75 to 80 percent of all gas engine
horsepower sold during the past 5 years was used for this application.
The primary uses are to power gas compressors for recovery, gathering, and
distribution.
During this time, sales to pipeline transmission applications
declined. Combined with standby power, electric generation, and other
services (industrial and sewage pumping), these applications accounted for
the remaining 20 to 25 percent of horsepower sales. The growth of oil and
gas production applications during this period corresponds to the
8-28
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
Increasing efforts to find new, or recover marginal, gas reserves, and
distribute them to our existing pipeline transmission network, and store
In covered, underground reservoirs near cities for peak winter demands.
Figure 8-8 illustrates the number of gas engines sold for five size
groups during the past 5 years. The large number of smaller than 500-hp
engines that were sold during this period are primarily one or twocylinder
engines used on oil well beam pumps and for natural gas well recovery and
gathering. Most of the other, larger gas engines that were sold during
this period ranged from 500- to 2000-hp. In 1976, approximately 400
engines in this size range were sold, primarily for oil and gas production
(see Figure 8-7). Most of these gas engines were manufactured by
Caterpillar, Cooper, Waukesha, and Superior Division of Cooper.
Historical sales data for pipeline transmission and field compres-
sor stations (see Figure 8-9)<43) dearly indicate the recent market
position for 1C engines and gas turbines. Total sales of 1C engines have
been relatively constant since 1970, while total gas turbine sales have
decreased dramatically with the recent slowdown of new pipeline construc-
tion. A breakdown of sales for transmission and field applications in
1975 (year ending June 15) is given
Prime Mover Turbine
Engine
Compressor
Horsepower
New
Additions
Total
Transmission
Stations
21,933
23,300
45,233
'••
Field
Stations
2,000
3,500
5,500
Transmission
Stations
4,080
78,800
82,880
———————
Field
Stations
29,400
30,450
59.850
8-31
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
These data Indicate that new transmission pipeline projects are
almost exclusively equipped with turbine-powered centrifugal compressors,
while new field stations (gathering, recompresslon, storage) are powered
by engine-driven reciprocating compressors. Widely varying loads are en-
countered 1n these latter applications, and therefore, engine-driven re-
ciprocating compressors are better adapted to this service than are
turbine-driven rotating compressors. In 1975 nearly 80 percent of the
total sales of reciprocating horsepower was for additions to, or replace-
ment of, existing compressor stations, while turbine sales were evenly
divided between new and existing compressor stations. In some cases the
gas turbine sales to existing stations have displaced reciprocating
engines. This has occurred in growing compressor stations or in facil-
ities where all the old engines reach retirement age at the same time and
can be replaced more cheaply by one large turbine than by several recipro-
cating engines.
Engines used 1n pipelines are concentrated 1n the major gaspro-
dudng areas, such as the Gulf Coast, and along the major natural gas
pipelines. Pipeline construction has dropped in the last several years,
but applications to the FPC for pipeline construction increased during
1973 and the first half of 1974 as plans were made to exploit the natural
gas discovered in Alaska'45'. If the percentage of compressor stations
utilizing reciprocating engines remains the same as in the past, this will
bring an increase 1n engine sales over the next several years as pipeline
companies purchase compressors to move the gas. The exact impact is un-
certain, however, as firm orders for equipment await final approval by the
FPC.
8-34
-------
According to one major western on company, 1n the past 10 years
there has been a movement away from reciprocating engines used 1n refinery
operations to electric motors, steam turbines for high-load requirements,
and, occasionally, fossil-fueled turbines'46). Moreover, a plant
manager from one of the major chemical processing firms, who utilizes
large reciprocating engines to compress gases, believes that large power
users will purchase gas turbines In the future for fuel conservation
reasons. Most of these users can utilize the waste heat from the prime
mover, and more of this energy can be recovered economically from one very
large turbine than from several large engines'47).
Municipalities use high-power, spark-Ignited engines to generate
electricity from digester gas in sewage treatment plants and to pump
water. Competition for these applications comes from gas turbines and
electric motors. As with electric power generation, 1C engines have an
advantage over turbines for applications when fuel costs are a significant
portion of annual costs. Engines are preferred over electric motors in
areas where electricity is relatively more expensive than liquid or
gaseous fuels and 1n applications such as sewage processing where a by-
product (such as sewage gas) can be burned to supplement other fuels.
8.1.4 Balance-of-Trade
The U.S. Bureau of the Census (Department of Commerce) does not
classify imported and exported 1C engines Into stationary and nonstatlon-
ary applications. Furthermore, there are no priority reasons for assign-
ing a breakdown by application to either imported or exported engines.
Therefore, information on the balance-of-trade for the stationary engine
market is limited to the following categories, for which the Department of
Commerce does report data:
8-35
-------
Imports: Engines for agricultural machines; compression
ignition engines; aircraft engines; bus, auto, truck
engines; outboard motors; and engines not elsewhere
classified (NEC)
Exports: Diesel engines (automotive, marine, and NEC); gas engines
other than turbines; outboard engines; gasoline engines
(marine, automotive, and NEC); and 1C engines NEC
Imported stationary engines would be included in the "Compression
Ignition Engines," "Agricultural Machines," and "Not Elsewhere Classified--
categories. These categories also include engines used for marine and
construction applications. For this reason it is not possible to de-
termine the exact number of imported stationary engines or the exact im-
pact stationary source regulations would have on imports.
The classifications are less of a problem for exports because the
categories are more narrowly defined. Furthermore, emission regulations
on future domestic engines would only affect exported engines if the U.S.
manufacturers added control devices to all their engines produced in the
U.S. rather than maintaining two lines of engines -- one for regulated
engines and one for unregulated engines. Table 8-7<48>49) gives import
and export data for the appropriate categories of 1C engines during the
fiscal years 1969 to 1973.
Except for 1972, the trade balance for 1C engines has been positive
at about $60 million to $75 million per year and is improving. It is in-
teresting to note that the average value of exported diesel engines is
about six-times the average value of imported diesel engines, while the
average value of exported gasoline engines is about one-third the average
value of imported gasoline engines. Using 1971 Commerce Department price
8-36
-------An error occurred while trying to OCR this image.
-------
data for the value of engines produced 1n the U.S. as a function of their
rated power, the following Information can be derived for 1973 Imports and
exports
Average value
Corresponding
per engine, $
average hp
Diesel Engines
Imports Exports
889 4996
60 300
Gasoline Engines
Imports Exports
148 49
11 6
Thus, future Imports of dlesel engines should correlate with U.S. demand
for small dlesel engines. Most of the dlesel-powered portable refrigera-
tion units and underground mining machinery 1n the U.S. use Imported
dlesel engines (51»52),
Since International markets for all capital equipment are highly
competitive, trade balances of engines may be affected more by monetary
exchange rates and tariff restrictions than by price changes due to emis-
sion control systems. Moreover, based on the average horsepower shown
above, large-bore engines play an Insignificant role 1n the Import or
export market of stationary reciprocating 1C engines.
8-38
-------
8.2 COST ANALYSIS FOR CONTROL OF NO EMISSSIONS
A
This section presents a discussion of the cost impact to the engine
user and manufacturer of implementing the viable NOX control options
designated in Chapter 6. The costs to the engine user of purchasing and
operating engines equipped with selected N0x controls are discussed in
Section 8,2.1. The costing was done on the basis of information supplied
by the manufacturers and users, and is applied to a group of ''model-
engines that are typical of those used in a particular application.
In Section 8.2.2 the costs to the manufacturers for the implementa-
tion of the alternative N0x controls (presented in Chapter 6) are dis-
cussed. These cost considerations include additional manufacturing asso-
ciated with adaptation of controls to existing designs and the costs of
engineering, tooling, and verifying the effectiveness of a particular
control approach.
Section 8.2.3 presents those costs associated with emerging control
techniques. !n Sections 8.2.4, 8.2.5 and 8.2.6 costs associated with fuel
pretreatment, modified facilities, and reconstructed facilities, respec-
tively, are identified.
8.2.1 New Engines
The application of N0x controls will affect costs to the engine
manufacturer and the engine user. The degree of the effect will depend
upon both the amount of reduction applied and the type of control
applied. As was shown in Section 6.3, various control approaches affect
initial costs, fuel consumption, and maintenance differently. Further-
more, manufacturers of stationary engines may incur different costs to
achieve a given N0x reduction depending on a number of factors includ-
ing: (1) their degree of advancement in emissions testing, (2) the
8-39
-------
uncontrolled emission rates of their engines, and (3) the necessary R&D
required to produce engines which can meet proposed standards of perfor-
mance. Therefore, the discussion of incremental costs to employ NOX
controls for manufacturers and engine users will be treated separately.
This section will be restricted to the discussion of incremental costs
incurred by engine users, and Section 8.2.2 will discuss NOX costs rela-
ted to engine manufacturers.
To illustrate the effect of NO controls on costs, "model"
/\
engines will be selected to represent major end users of diesel, dual-
fuel, and natural gas engines. Baseline costs, comprised of investment
and operating expenses, will be established for each model. Computations
for these model units will then be used to illustrate the range of incre-
mental costs to the user resulting from the application of the NOX con-
trol systems described in Section 6.2. These incremental costs will be
illustrated for several control systems that achieve any one of three
levels of NO reduction (20 percent, 40 percent, and 60 percent). This
A
approach is not intended to be a comprehensive cost analysis of all pos-
sible NO control systems; rather it is intended to illustrate a range
A
of costs that an engine user would incur to achieve a given level of N0x
reduction. The discussions are subdivided by major end uses, since engine
types and costs are unique to each end use.
Section 8.2.1.1 briefly describes the models selected to represent
major engine applications. The cost analysis methodology is then discus-
sed in Section 8.2.1.2, and the results of the cost analysis are presented
in Section 8.2.1.3.
8-40
-------
8.2.1.1 Model Engines
Four model engines have been selected to represent the major appli-
cations of diesel, dual-fuel, and natural gas engines. (The applications
were described in Section 8.1.3) The following paragraphs briefly des-
cribe these models.
Diesel Engine Model: Electrical Generation
As described in Section 9.3.1, affected diesel engines are large-
bore, exceeding a displacement of 560 cubic inches per cylinder. Typi-
cally these engines are used as prime movers for electrical generators in
municipal utilities. These engines operate from 6000 to 8000 hrs/year
(baseload) and consume approximately 7000 Btu/hp-hr of operation. Manu-
facturers of these engines include Alco, Colt, Cooper and Superior Divis-
ion (of Cooper), DeLaval, and ElectroMotive division of General Motors.
Dual-Fuel Engine Model: Electrical Generation
These engines are nearly identical to the diesel engines except
that they burn predominantly natural gas (typically 95 percent of the
total fuel heating value). In general, these engines operate more effi-
ciently than their diesel counterparts, consuming 6500 Bth/hp-hr of opera-
tion. Manufacturers of affected dual-fuel engines include Colt, Cooper
and Superior Division (of Cooper), and DeLaval.
Gas Engine Model: Oil and Gas Transportaion
These engines are installed on pipeline compressors for long-range
transportation of natural gas. They generally exceed 1000 hp, averaging
3000 to 4000 hp. Typical annual usage is 8000 hrs, and representative
fuel consumption is 7000 Btu/hp-hr. Manufacturers of gas engines for this
application include Colt, Cooper and Superior Division (of Cooper),
DeLaval, and Ingersoll-Rand.
8-41
-------
Gas Engine Model; 011 and Gas Production
These engines are Installed on compressors that gather, store, pro-
cess, or distribute gas from gas production fields. (These engines gener-
ally burn gas that has been treated to reduce the sulfur content.) As
described 1n Section 9.3.3, engines that would be affected by proposed
standards of performance range from 300 to 2000 hp. An average size
engine 1s about 1000 hp and consumes about 8000 Btu/hp-hr. These engines
are manufactured primarily by Caterpillar, Cooper and Superior, Waukesha,
and Ingersoll-Rand.
8.2.1.2 Costing Model for Users of Stationary 1C Engines
The objective of the cost analysis is to estimate how an engine
user's life cycle costs will change with the application of NOX controls
for the model diesel, dual-fuel, and gas engines described in Section
8.2.1.1. The following paragraphs describe the cost analysis approach and
the basic assumptions that are used to estimate the incremental costs cre-
ated by NO controls. Section 8.2.1.3 will then present the results of
A
the cost analysis.
Methodology
The costs of owning and operating a large-bore engine can be repre-
sented as follows:
TAG = AIC + M + F
where TAC = total annual cost of ownership and operation of engine
AIC = annualized initial cost = initial engine cost x capital
recover factor (CRF)
M = maintenance costs
F = fuel and lubrication costs
8-42
-------
The annuallzed Initial cost Includes capital recovery of the Initial in-
vestment (assuming 100 percent debt financing), depreciation, property
taxes, and Insurance. Capital recovery rates typically range from 15 to
25 percent. Conversations with Industry spokesmen Indicate that a rate of
20 percent 1s appropriate!/ for estimating Installed engine costs for
electrical generation, gas production, and gas transportation appllca-
The procedure for computing the Incremental costs of various NO
control techniques 1s as follows:
• Estimate the Increase 1n the costs of AIC, M, and F due to
NOX controls
• Compute the Increase 1n the total annual cost, TAC
• Present the results as (TACC - TACU)/TACU x 100 = percent
Increase 1n TAC where c - controlled and u » uncontrolled
Basic Costs and Parameters for Cost Analysis
Table 8-9 summarizes the basic Inputs for computing the total an-
nual costs of uncontrolled engines. As this table Indicates, the costs
will be presented 1n a brake specific format, that 1s, 1n $/hp-hr. The
Initial costs are normalized by the output power and usage rate to obtain
$/hp-hr. This format makes 1t possible to compare ownership costs for a
number of differently sized engines that are used 1n the same applica-
tions. This format also permits a direct comparison of the Incremental
NOX control costs among engines using different fuels.
Ph^ical life* 20^ear accounting
? 10-Percent ^terest on debt, and
and fixed
8-43
-------
As shown in Table 8-8, typical initial costs for diesel and dual-
fuel electrical generation and oil and gas transmission engines are
$150^55^. (This cost is for the engine only, F.O.B..) Costs for gas
production engines are estimated at $50/hp and are representative of
engines sold by Caterpillar and Waukesha' '. Total capital investments
for installed electrical generation stations are approximately $300/hp
(1976)'57'. Current investments for installed gas transmission compres-
sor installations range from $318 to $575/hp, and range up to $584/hp for gas
field stations^ '. Thus, the cost increases computed in this section
will be considerably smaller when expressed as a percentage of the total
application investment.
Maintenance costs for these engines have been estimated based on
information supplied by engine manufacturers. These costs are typical of
engines that operate continuously at rated load.
Fuel costs assumed for the electrical generation applications are
representative of costs of distillate oil and natural gas transported both
intra- and interstate. (The current (1978) regulated price of interstate
gas is $1.48 per Mcf.) Lower gas costs have been assumed for oil and gas
production and transportation applications since gas companies own these
facilities and pay less for the gas. In addition, average gas costs for
these companines are a composite of contracted supplies of gas that span
several years' '. Fuel consumption estimates are average values based
on the data presented in Section 4.3.1. Note that this analysis assumes
baseload or continuous annual operation («8000 hr/yr). Figure 8-9 illu-
strates the relative proportion of each of these items relative to total
uncontrolled costs. Fuel and lubrication costs are the largest fraction
8-44
-------
TABLE 8-8. BASELINE COST PARAMETERS FOR LARGE-BORE ENGINES
COST
PARAMETERS
Initial Costa>b, $/hp
Capital Recovery Factor
Annual Usage, hr/yr
Maintenance*
Parts, $/hp-hr
Labor, $/hp-hr
Total, $/hp-hr
Fuel Costc, $/106 Btu
Fuel Consumption, Btu/hp-hr
Lubrication8, % Fuel Cost
ENGINE M
DIESEL
(Electrical
Generation)
150
0.2
8000
0.0012
0.0018
0.0030
2.50
7000
5
DUAL- FUEL
(Electrical
Generation)
150
0.2
8000
0.0012
0.0018
0.0030
3.00
6500
10
3DELS
NATURAL GAS
(011 & Gas (011 & Gas
Transport) Production)
150 50
0.2 0.2
8000 8000
0.0012 0.0012
0.0010 0.0018
0.0030 0.0030
2.00 2.00
7000 8000
10 10
'Aggregated from confidential communications with engine manufacturers.
Initial cost divided by annual usage (8000 hr) and multiplied by the capital recovery factor
(0.2) gives annuallzed cost 1n $/hp-hr.
cFuel cost calculations for dual-fuel units assume fuel consumption 1s lOOiS gas. All fuel
costs based on 1977 Information.
8-45
-------An error occurred while trying to OCR this image.
-------
of total costs, accounting for 70 to 80 percent of the total. The varia-
tion 1s a result of the different engine and fuel costs assumed for each
of the model applications.
The capital and maintenance cost Increases resulting from the
application of N0x controls are estimated in Table 8-9. These estimates
were aggregated from information supplied by engine manufacturers or ven-
dors of auxiliary equipment. Increases in fuel consumption will be esti-
mated for various control techniques based on the information given in
Table 6-4 of Section 6.2.
8.2.1.3 Results of Cost Analysis
A cost analysis based on the preceding discussion was performed for
each of the four model engines described in Section 8.2.1.1. Total annual
cost penalties (as a percentage of total uncontrolled costs) were computed
for each model engine and alternative level of NO reduction (i.e., 20-,
40-, and 60-percent reduction) for the control techniques that were dis-
cussed in Section 6.2. Table 6-4 is repeated here as Table 8-10 to Illu-
strate those techniques and their fuel penalties for each level of control
alternative and fuel type.
Fuel penalties are the major factor affecting cost increases for
high usage engines. Table 8-10 shows that fuel penalties increase with
increasing level of control. They also vary with control type. For ex-
ample, derating results in substantial penalties (>10 percent) for NO
reductions greater than 20 percent. Retard, manifold air cooling, and
air-to-fuel controls, however, generally achieve NO reductions at a
/\
penalty less than 10 percent. It should be noted that the decrease in
data at the 60 percent NOX reduction level is the result of both: (1)
manufacturer's inexperience with the application of controls to the extent
8-47
-------
TABLE 8-9. COST INCREASES TO THE ENGINE USER RESULTING FROM NO CONTROL3
/v
CONTROL
Retard
Alr-to-Fuel
Derate
Manifold
A1r Temperature
Reduction
External
Exhaust Gas
Rec1rculat1on
CAPITAL
None
None
Increase by ratio of rated
power to derated power (to
compensate for power loss)
Increase engine costs 1.5%
to achelve 100CF Inlet
air. Larger heat exchangers
cost assumes engine equipped
with Intercooler
Increase engine cost 5% for
plumbing, heat exchanger
and controls for 10-123S
redrculatlon.
MAINTENANCE
33% Increase 1n base cost due to 25%
reduced service life of exhaust valves for
dual-fuel engines
Increase of 0.0001/hp-hr for Increased
cleaning of turbochargers
Increase by ratio of rated power to derated
power (more units or cylinders to service)
D ) Increase $0.0005/hp/hr for cooling
DF) water chemical treatment (cooling
towers)15
G ) Increase $0.0001/hp-hr for Increased
) service of radiator and aftercoo1erc
Double parts for dlesel engines'1
Triple parts for dual-fuel and gas engines'1
"Aggregated from confidential communications with engine manufacturers.
Cooling water from cooling towers for dlesel and dual-fuel engine must be treated to prevent
sludge and scale buildup due to water "hardness".
Closer tolerances to achieve lower manifold air temperature will require more frequent cleaning
and servicing of radiators and Intercoolers (or aftercoolers) of gas engines.
dD1esel unit has fixed rate of EGR, dual-fuel, and gas units have a variable rate of EGR
Charge for parts Includes periodic replacement of the EGR system and Its controls.
8-48
-------An error occurred while trying to OCR this image.
-------
necessary to achieve that level, and (2) 1n some cases the Inability of a
particular control approach to achieve reductions at this level.
The differential control costs for the techniques shown on Table
8-10 are tabulated In Tables 8-11 to 8-14 for the four end use applica-
tions described above. Table 8-15 1s a summary of these cost penalties.
In general, retard, manifold air cooling, air-to-fuel change, or some com-
bination of these achieved 60-percent NO reductions for less than a 10-
rt
percent cost penalty for each application except diesel/electrlc genera-
tion. The cost penalty for the diesel/electric generation, 60-percent
NOX reduction category, however, is based on data from tests of only one
engine model; therefore this result may not be representative of costs for
other engine models.
The data in Table 8-15 indicate a wide variation in cost penalty at
any level of NOX reduction. Moreover, average cost penalties are less
than 6 percent (with the exception of diesel engines) for a 60-percent
NOX reduction. Nevertheless, average cost penalties increase as the
level of NO reduction increases.
A
Since average uncontrolled N0v emission rates from engines of
X
different manufacturers vary, cost penalties to achieve a given alterna-
tive performance standard will also vary among manufacturers. These dif-
ferential costs are important to identify so that potential economic
impacts in various end use markets can be identified (see Section 8.4.1).
Table 8-16 illustrates the cost penalties for each manufacturer and
fuel type for each of the three alternative levels of performance stan-
dards. In general, the maximum cost penalty for any fuel is less than 10
percent with the exception of the 40- and 60-percent reduction levels for
diesel engines. The data for gas engines do not indicate differential
8-50
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
cost penalty ranges among manufacturers at the 40- and 60-percent reduc-
tion levels; however, the variation within a cost penalty range is large
enough (e.g., 2 to 7 percent) for cost differentials to exist among manu-
facturers. That is, two manufacturers could be in the 2- to 7-percent
cost penalty range and one could incur a 2-percent penalty while the other
incurred a 7-percent penalty. (The potential impacts of these differen-
tials are discussed in Section 8.4.1.)
This observation also holds for the dual-fuel and diesel cate-
gories. The results for the diesel category indicate that Superior Divis-
ion of Cooper has a cost advantage. However, Superior diesels are
smaller, and in general, serve smaller power applications than Colt,
DeLaval, or larger Cooper engines. Futhermore, the data are based on
results from only one diesel engine model; therefore, the magnitude of
this penalty may not be representative of penalties for all of these
engines at this level of reduction. The data from Table 8-16 is analyzed
in detail in Section 8.4.
8.2.2 Engine Manufacturers
Manufacturers of stationary reciprocating 1C engines will incur ad-
ditional costs due to the proposed standards of performance. As discussed
in Section 6.3, these costs are a result of one or more of the following
activities that may be needed to manufacture engines which meet standards
of performance:
• Extended testing to verify the effectiveness of a particular
control approach
• Development and application of NOX controls to existing
engine designs
8-57
-------
• Engineering, tooling, and pattern costs for the redesign of an
engine family
Costs related to these actions have been estimated for the control techni-
ques summarized 1n Section 6.3 (technically viable approaches to meet pro-
posed standards of performance) and are shown on Table 8-17^60»61^.
These estimates are representative of the costs that manufacturers would
Incur to adapt each of the NOX control systems to an engine family
(I.e., group of engines with same air and fuel charging system and combus-
tion chamber geometry). These figures include costs to test engines for
durability and to retool their production facilities where necessary, but
do not Include costs to purchase or manufacture components placed on the
engine. These latter costs are included in the user-oriented cost analy-
ses of Section 8.2.1, which considered primarily additional hardware costs
in the initial price. The table also gives estimated times to implement
the various control technologies. It should be noted that these costs
will double if a manufacturer is required to meet emission standards for
two types of fuel (e.g., diesel and dual-fuel).
In general, manufacturers believe their present overhead budgets
are sufficient for the development of the controls shown on Table 8-17,
with the possible exception of EGR and combustion chamber modifications
which will require considerably more development over a longer time. As
shown in Section 4.3.1, all of the manufacturers have established baseline
emissions for most of their engines. They believe that controls such as
retard, air-to-fuel, manifold air cooling, derating, and combinations of
these approaches would be relatively simple to implement, although some
development time would be required to optimize a particular approach and
8-58
-------
TABLE 8-17. ESTIMATED COSTS FOR ENGINE MANUFACTURERS TO DEVELOP
NOX CONTROLS, (BASED ON REFERENCES 60, 61)
CONTROL
Retard, R
A1r-To-Fuel Change, A
Manifold A1r Temperature
Reduction, M
Derate, D
R + M
R + M + A
External Exhaust*
Gas Redrculatlon, EGR
Combustion Chamber
Redesign6
DEVELOPMENT TESTS*
TIME, HR COST, $
200 25,000
200 25,000
400 50,000
200 25,000
300 40,000
400 50,000
2,500 300,000
10,500 1,260,000
EXTENDED TESTSb
(DURABILITY)
TIME, HR COST, $
2,000 100,000
2,000 100,000
2,000 100,000
0 0
2,000 100,000
2,000 100,000
12,000 450,000
16,000 560,000
ESTIMATED
TOTAL COSTC, $
125,000
125,000
150,000
25,000
140,000
150,000
750,000
1,820,000
TOTAL
DEVELOPMENT
TIME3, MONTHS
•••••II ii 1 1 n 1 1 j
15
15
15
9
15
15
35
69
°f exp1oratory and development time to establish operation and
Jh1s estimate assumes that one technician 1s 1n attendance full-time during a test of a 2000-hp engine
e^lmlSlt;!;6"^! e^lnriodeTSJmll"8' """"« ***' "< ^^ COStS' Tota
dFrom Table 6-5.
"Estimates for development time and costs Include engineering and redesign of engine components.
T-774
8-59
-------
establish its durability. Since several manufacturers have less experi-
ence in emission control research than others, the costs and times shown
in Table 8-17 for these techniques have been estimated conservatively,
taking into account the variation in emission control experience among
manufacturers.
As discussed in Section 4.4, modifications of engine operating con-
ditions, such as retard or air-to-fuel changes, may necessitate modifica-
tions of materials used for exhaust values and turbochargers, if exhaust
temperatures greater than 1200°F (present material limits) are experi-
enced. However, temperature data submitted with the emission data of
Appendix C do not indicate excessive temperature will be encountered for
the levels of retard and air-to-fuel changes reported.
Research and development of NOX emission control techniques that
require more extensive development and/or redesign (e.g., EGR or combus-
tion chamber changes) are difficult to quantify. They will be determined
to a large extent by a particular manufacturer's experience with emission
controls, the uncontrolled emission levels of his engines, and the res-
ponse of a particular engine to the control technique. Manufacturers have
indicated that the time required to incorporate major engine design
changes range typically from 3 to 5 years, but may extend to 10 years in
some cases' •'. This time includes initial design, endurance testing
(including 1 to 2 years operation in a real application, but under close
monitoring by the manufacturer), and tooling-up (9 to 12 months). It is
unclear to what extent the costs for these activities would be added to
present R&D expenditures or absorbed into the existing budgets, thereby
displacing R&D that would have been undertaken in the absence of emission
standards.
8-60
-------
These issues will be addressed further in Section 8.4 which discus-
ses the economic impact of both users and manufacturers of 1C engines sub-
ject to standards of performance.
8.2.3 Emerging Controls
In this subsection, the costs to control N0x emissions from
large-bore engines using exhaust gas treatment, combustion chamber modifi-
cations, or water induction are estimated. None of these systems has been
used on a large-bore engine for any length of time; hence they cannot be
considered for immediate application. At the same time they are not
dependent upon any technological breakthroughs and should be available for
use within 6 years, given appropriate priorities in the R&D budgets of the
engine manufacturers.
Exhaust Gas Treatment: Ammonia/Catalyst syctQm
Even though no engine manufacturer has reported on the application
of N0x reduction catalysts to large-bore engine exhausts, at least one
source considers the reduction of N0x by ammonia injection over a
precious metal (e.g., platinum) catalyst as a promising control technique
of the future (see Section 4.4.9)<">. An estimate of the cost of this
technique is presented in order to provide a comparison with the costs
presented in Section 8.2.1 for the other NO controls.
A
The above-mentioned source reports that approximately 2 cubic feet
of honeycomb catalyst (platinun,-based) would be required for a 1000-hp
engine to ensure proper operation of the system. The cost of the catalyst
was estimated at $1500/cubic foot (in 1973). Assuming that the engine
costs $150/hp and that the cost of the cata,yst accounts for about onehalf
the cost of the whole system (container, substrate, and catalyst), the
8-61
-------An error occurred while trying to OCR this image.
-------
less than 5-percent oxygen^65). Large-bore reciprocating engine exhaust
contains considerably more oxygen (e.g., about 15 percent for gas
engines); hence NOX reduction may be less.
It is also Important to note that the consumption of ammonia can be
expressed as a quantity of fuel since natural gas is generally used to
produce ammonia. Assuming a conservative NOV emission rate of 20
A
g/hp-hr, and engine heat rate of 7500 Btu/hp-hr, a heating value of 21,800
Btu/lb for natural gas, and a requirement for approximately 900 Ibs of gas
per ton of ammonia produced^, then the ammonia necessary for the
catalytic reduction has the same effect on the supply of natural gas as a
2-percent increase in fuel consumption. Additional fuel is required to
operate the plant which produces the ammonia.
Combustion Chamber Modifications
Section 4.4 described several chamber designs that have been shown
to produce relatively low NOX emissions from truck-size engines. There
seems to be no technological reason why these designs cannot be adapted to
larger engines (careful engineering analysis and design would be required
to make the transition). Moreover, it was noted in Section 4.4 that a
variable throat precombustion chamber design had been adapted to a large-
bore engine in a laboratory. Therefore, an estimate of the costs associ-
ated with such a change is presented.
It is difficult to estimate the cost of designing a new, major
change to an engine. One large-bore manufacturer estimates that a combus-
tion chamber redesign would require 4 to 5-1/2 years to complete depending
on the engine design (2- or 4-stroke cycle I,, this .ase). This redesign
could affect pistons, cylinder heads and liners, injection components, and
valves and would require an additional 12-month endurance testing before
8-63
-------
the final design could be released for production. This manufacturer
would not need to expand his facilities for such a redesign program;
nevertheless, he estimates the R&D expenditure would be in addition to
normal development work and on the order of 1-1/2 times their typical
development costs.
The estimate to be presented here comes from private contacts with
manufacturers of truck-size engines who have considered the advisability
of converting their direct-injection engines to precombustion chambers.
They estimate an approximate development cost of $0.5 million, and note
that this figure would be higher if their staff had not already developed
familiarity with precombustion chambers through previous experiences. The
estimate is based also on the experience of Teledyne Continental who
reported a $3-million development and retooling cost for a gasoline engine
with a new chamber and ignition system design to replace an existing pro-
duction moder . This engine is rated at less than 100 horsepower,
and several tens of thousands are produced each year (14,700 alone for
stationary applications in 1974). Teledyne Continental's expenditure
probably represents an upper bound for the costs that a large-bore engine
manufacturer might face; his R&D costs might be more, but retooling
expenses should be significantly less.
Based on this information, we estimate that it would cost a manu-
facturer of large-bore engines no more than $2 million to convert his
units to precombustion chambers or squish lips. If one assumes that the
average manufacturer in this group sells 50 engines a year, that the aver-
age rating of these engines is 1200 hp, that the cost of $150/hp applies,
and that the manufacturer should be able to recover such an investment in
8-64
-------
5 years, then he would have to increase the sales price of his engines by
nearly 5 percent.
Presumably fuel consumption in these redesigned engines could also
be 5 to 8 percent higher than it is for current models, just as in some
precombustion chamber truck-size engines relative to their closest open-
chamber equivalents. In that case, the economic impact from a standard
that could be met only by such a redesign would be about 5 to 8 percent of
the uncontrolled total annual costs for engines used in continuous ser-
vice, depending on fuel penalty.
Water Induction
Research has Indicated that the induction of water into the engine
in quantities equivalent to the mass of fuel consumed can be effective in
reducing N0x emissions'68-"). Reductjons of approxfmate]y 6Q ^^
in N0x have been demonstrated on several gaseous fueled engines, using
»ater to fuel mass flow ratios (,b water/lb fuel) of nearly one. In gen-
eral, no increase in fuel consumption occurred (see Chapter 4 and Appendix
0. However, this research indicated that serious maintenance and dur-
ability problems are associated with water induction, and therefore, this
control, technique has been omitted from the cost analysis of currently
available techniques (Section 8.2.1).
Since this technology, however, could become available in the
future after further testing with water treatment, different allowances in
the water injectors, exhaust valves, etc., costs are estimated below for
this potential technology.
Testing has shown that the water must be deionized prior to induc-
tion to remove minerals which would otherwise deposit in the engine (e.g.,
on the intake and exhaust valves) and adversely affect performance'70).
8-65
-------
For a 1000-hp engine, approximately 1 gallon per rain- ute would be
required to obtain a water/fuel ratio of 1. Deionized water could be
supplied in bulk with an appropriately sized storage facility at a price
of $15/1000 gallons, excluding the initial price of the storage
facility^. Based on the information supplied in Table 8-8, such a
system would cause the total annual cost (including water treatment and
disposal, storage, and delivery system) for a 1000-hp gas or dual-fuel
engine used in continuous service (8000 hr/yr) to increase about 8 per-
cent, assuming no additional maintenance. Total annual costs, however,
would increase approximately 25 percent if one made the reasonable assump-
tion that engine maintenance requirements would be doubled (cleaning flow
passages, dewatering or replacing lubricant, etc.). Corresponding
increases for a similarly sized diesel-fueled engine would be approxi-
mately one-half as much, since diesel fuel costs about 2.5 times more than
natural gas.
If the water is deionized at the engine location, the water cost
will be approximately $50/1000 gallons for a system that supplies 1 gallon
per minute(72). Total annual costs for a 1000-hp engine in continuous
service fueled with natural gas would increase approximately 15 percent
for no engine maintenance increase and 85 percent assuming engine mainten-
ance costs doubled.
A larger engine, e.g., 6000-hp, would require a proportionately
greater water rate to maintain a water/fuel ratio of one. A reverse
osmosis water treatment system would be better suited for this higher vol-
ume application (,5 gallons per minute). Based on a $50,000 investment
for this system and a raw water cost of $0.5/1000 gallons, the total
nnual costs for a 6000-hp gas-fueled engine in continuous duty
8-66
-------
(8000 hr/yr) would Increase approximately 10 percent assuming no Increase
1n engine maintenance, and 35 percent assuming maintenance charges doubled.
Summary of Emerging Control Technique
The costs associated with an exhaust gas treatment system (ammonia)
and with combustion chamber modifications are comparable to those for
derating and EGR (see Table 8-15). Water Induction control costs appear
to be generally higher than for most control techniques summarized in
Table 8-15, especially 1f delonlzed water 1s not readily available 1n
bulk, and 1f maintenance Is significantly Increased.
8-2.4 Fuel Pretreatmervfr
Desulfur1zat1on
Sulfur oxides arise from the nearly quantitative combustion of sul-
fur in the fuel. The fuels traditionally burned 1n reciprocating engines
(I.e., gas and distillate oil) are low In sulfur, nitrogen, and ash.
According to a recent survey, over 50 percent of all distillate contains
less than 0.3 percent sulfur, and the average nitrogen content 1s about
0.03 percent(73).
As explained in Section 4.4.13 of this report, large-bore dlesel
engines are occasionally equipped to burn crude or residual fuel oils.
These generally contain higher levels of sulfur than the more commonly
used distillate fuels.*/ Therefore, engines fired with the heavier
high-sulfur fuels may Incur Increased control costs above those outlined
in Section 8.2.1 for NOX control alone. The purpose of this section of
the report 1s to determine the cost of possible standards for the user who
burns residual oil.
8-67
-------An error occurred while trying to OCR this image.
-------
a 3.5-percent increase in total fuel costs. Comparable increases for 0.5-
and 0.8-percent sulfur (over 1 percent) are 3.2 and 1.9 percent,
respectively.
Denitrification
While S02 in the reciprocating engine exhaust originates exclu-
sively from the fuel, N0x arises from thermal fixation of nitrogen in
the air and from partial oxidation of nitrogen in the fuel. The control
methods evaluated in Section 8.2.1 were devised only for thermal N0x and
have been tested with the traditional clean fuels (natural gas and distil-
late oil), where nearly all the N0x generated is thermal. The possible
use of heavier fuels suggests that fuel nitrogen may someday become a
problem.
Thus, the currently available technology for control of NOX from
fuel nitrogen is, like that for control of S0x, modification of th'e
fuel. But whereas the desulfurization of residual oil is practiced delib-
erately, denitrification is practiced almost inadvertently. Some nitrogen
is removed as a byproduct of desulfurization. The fraction removed is
consistently less than that of sulfur but is rarely monitored. Neverthe-
less, nitrogen removal occurs at no cost above that for sulfur removal.
The fuel is not sold under a nitrogen specification, and indeed the refin-
ers have resisted (successfully, in times of fuel shortages) such a speci-
fication. Moreover, desulfurization competes with other refining proces-
ses for the Umited supply of hydrogen. New developments in desulfuriza-
tion catalysts have the incentive of reduced hydrogen consumption, but the
newer catalysts remove less nitrogen. For the time being, then, the owner
of an affected facility cannot buy low-nitrogen fuel by specification as
he buys low-sulfur fuel. Low-nitrogen fuel can be used to help reduce
8-69
-------
NOV emissions but cannot be used 1n lieu of applying control techniques
n
to the engines themselves to meet the standard because the standard(s)
win be based on test data from Installations burning the traditional
clean fuels.
A standard of performance for NO which failed to recognize the
A
special problem of fuel nitrogen would create a bias against the burning
of residual fuels. One manufacturer Indicated crude or residual oil 1s
used 1n place of distillate where a long-term commitment for the heavy oil
1s at a price lower than distillate. This spokesman pointed out that
since dlesel fuel 1s regulated but not crude, the price of crude 1s not
necessarily lower than distillate. In fact, crude oil must be signifi-
cantly lower than No. 2 oil to justify the increased capital and mainten-
ance costs for heavy fuel handling^76'. Furthermore, the lack of infor-
mation regarding fuel-bound nitrogen formation and removal precludes a
more comprehensive discussion. Therefore, since the proposed standards
for dlesel and dual-fuel engines will be based on emissions data obtained
when using No. 2 oil, compliance with the standards may also require oper-
ation with No. 2 oil.
8.2.5 Modified Facilities
As discussed in Chapter 5, a user usually does not make physical or
operational changes to an existing engine installation which would
increase Its N0tf emission rate. However, 1f he did, he would be
A
required to conform to a standard of performance for new sources. There-
fore, we will briefly consider the potential cost impacts of such a change.
8-70
-------
Essentially the user contemplating a modification has two
alternatives:
1. Make the modification, Including use of the most practical
available control technology (see Chapter 5) that prevents the
emissions from Increasing as a result of the modifications
2. Buy a new engine, or alternative power source, which satisfies
the new requirement for which the modification was Intended.
If a new engine were purchased, the appropriate standard of
performance would apply.
Normally a user would contemplate a modification for one of the
following two reasons:
1. To increase the power output of his engine
2. To install newly designed parts, such as injectors or pistons,
in place of the old ones during an overhaul
For example, a user might wish to install a new injector that the manufac-
turer designed to reduce smoke or fuel consumption. Other changes, as
discussed in Chapter 5, could include alterations of the cylinder head,
Piston, valve or porting configuration, and manifolds.
If these changes resulted in greater N0x emissions, the user
would probably rely on the control techniques described in Chapter 6 to
bring the modified engine within standards of performance. The cost
impact would depend on the initial emission level and the N0x emission
level after modification, if the level after a modifications in the
same range as those from new, uncontrolled units, the cost increases would
also be similar to those presented in Section 8.2.1. (Derating is exclu-
ded as a possible control strategy since power requirements at existing
facilities are fixed.)
8-71
-------
At present there is insufficient data to judge the response of
older engines to NO control techniques. Some older engines are likely
A
to have emissions lower than a standard of performance for new sources,
and emissions after modification could also be less. In other cases, this
might not be so. Therefore, it is not possible to state quantitatively
the economic impact for users who make a modification to their engine and
are then required to meet a standard of performance.
When faced with the need to meet a standard as a result of an
intended modification, an owner or operator would normally weigh the cost
impact of this approach against that of replacing his existing facility
with a new one. This new engine would, of course, meet the standard as
well as satisfy his new need. The cost of a major overhaul is probably a
reasonable estimate of the maximum price of a modification-plus-alteration
to bring the engine into compliance with a standard of performance for new
sources. For a large-bore, low- or medium-speed reciprocating engine,
this cost is typically less than one-third (see Chapter 5) of the purchase
price of the engine.-/ Therefore, to be competitive, an alternative
engine or power source would need to approximate this cost, taking into
account both initial and operating costs. Thus, it is unlikely that the
user would substitute a different engine (to comply with performance stan-
dards) rather than apply control technology to his existing engine.
-/A spokesman for one engine manufacturer indicated that practically all
the development of design changes for new engines occurred in-house with
the manufacturer incurring about 90 percent of the cost for endurance
testing of the design change. Thus, end user-evaluation of design
changes are rare, and hence, those costs are not included in this
estimate.
8-72
-------
8.2.6 Reconstructed Facilities
In large-bore, lower medium-speed engines the main housing, or
structure, is never replaced during the engine's life, but at some time or
other almost every component may be renewed. In general, the various
parts are replaced at different times, as they wear out or break, but not
at the same time (except for parts replaced during routine maintenance and
overhaul). As noted in Chapter 5, overhauls are performed routinely
throughout an engine's life and, therefore, should not be considered a
reconstruction even though substantial portions of the engine are replaced
(generally with parts identical to those originally installed). There-
fore, since reconstruction within the meaning of 40 CFR 60.15 is not
expected to occur, there is no need for estimating the cost impact of
standards on reconstructions.
Industry spokesmen have expressed concern that a standard of per-
formance might deter the replacement of older engines with newer, more ef-
(78)
ficient onesv '. It does not seem likely, however, that the potential
cost increases due to a standard would affect that decision. In fact, if
one accepts an industry estimate that new engines consume about 75 percent
as much fuel as older ones^79', then one can show that it is more cost
effective to purchase such a new, efficient engine (i.e., one with a fuel
consumption rate of 7500 Btu/hp-hr) for continuous duty than to simply
maintain the old one. This is true even if the new engine is burdened
with a 5-percent initial and 80-percent maintenance cost increase (e.g.,
with E6R) to represent the maximum cost penalty expected from controls, if
it is assumed that the old engine is completely amortized, and if the same
maintenance costs are assigned to the old engine as to an uncontrolled new
unit. In fact, the difference in favor of the purchase is about 9 percent
on an annualized basis.
8-73
-------
8.3 OTHER COST CONSIDERATIONS
This section identifies any costs incurred by large-bore recipro-
cating engine users as a result of environmental regulations other than
standards of performance for air pollution. Such regulatory requirements
might concern solid waste disposal, water pollution control, or noise
control. The purpose of this section is to identify Incremental costs
imposed by these regulatory requirements that may in some way limit the
ability of the user to bear the cost of control techniques presented in
Section 8.2.
Stationary reciprocating engines do not cause solid waste disposal
or water pollution problems. Used lubricating oil is not considered a
solid waste because it is either sold to an oil recycler or burned as fuel
in a boiler. The only conceivable source of solid wastes or contaminated
water from the operation of these engines might arise from water purifica-
tion and demineralization for a water induction control system. These
wastes, however, are currently hypothetical since such systems do not now
exist; they may be used in the future if water induction is developed as
an emission control technique. Treatment costs to prepare water for use
in engine jacket and aftercooler cooling systems are presently accounted
for in maintenance charges for these engines, and hence, have been in-
cluded in the analysis presented in Section 8.2.1.
Similarly, no noise regulations are presently in effect specifi-
cally for large-bore reciprocating engines. These engines are typically
installed in remote locations (e.g., gas pipelines) or separate buildings
(e.g., electric utilities and standby service) where remote control of the
engine or process reduces noise exposure. In addition, mufflers are used
to reduce noise. However, one manufacturer of large-bore engines has
8-74
-------
reported that complex noise control systems could be cost-prohibitive to
an Industry whose manufacturers only sell a few hundred units per year and
who typically have limited staff and facilities for noise control research
. This source reports, nevertheless, that manufacturers are begin-
ning to Incorporate noise control Into their research and design programs.
End-users, on the other hand, have been faced with Occupational
Safety and Health Act (OSHA) regulations limiting worker exposure to noise
levels and EPA regulations for the protection of communities from annoy-
ance. One source reports that noise control for engines 1n the 800- to
2000-hp range can cost as much as $10,000 to $30,000. <81) The level of
effort to meet OSHA requirements 1s often unclear. For example, simple,
relatively cheap devices such as earmuffs and earplugs can be used as a
last resort when other methods of noise control prove technically unfeas-
ible. However, the definition of technical feasibility 1s uncertain, and
depends on the circumstances of each particular application.
In conclusion, there are no other regulatory requirements that, at
present, will limit stationary reciprocating engine user's ability to
absorb incremental costs as a result of standards of performance for air
pollution.
8-75
-------
8.4 ECONOMIC IMPACT
This section analyzes the economic impacts of alternative standards
of performance for NO emissions from large stationary reciprocating internal
/\
combustion engines on engine manufacturers (Section 8.4.1), gas and elec-
tricity prices (Section 8.4.2), employment (Section 8.4.3), and foreign
trade (Section 8.4.4). These impacts have been evaluated for alternative
standards of 20, 40, and 60 percent below sales-weighted average emissions
of uncontrolled gas, dual-fuel, and diesel engines. The extent of the
impact for each alternative, assuming retard, air-to-fuel change, or
manifold temperature reduction techniques are used to achieve compliance,
is summarized in Section 8.4.5 and in Table 8-18.-/ The following is a brief
description of the results of the economic impact analysis.
The capital budget requirements for testing engine models are an
estimated $5 million for a 60 percent alternative. These expenditures will
be made over a two-year period and could be financed internally by engine
manufacturers from profits on internal combustion engine sales. No firm is
expected to lose more than seven percent of its sales to competitors. Gas
turbine sales will not make additional inroads into sales of reciprocating
engines with the possible exception of diesel engines used for electricity
generation. The total U.S. electric bill would increase by 0.3 percent
when controls are applied to all engines. This level would not be reached
until all engines are replaced (full phase-in would take about 30 years).
Localities using internal combustion engines exclusively to generate
-/Other techniques-derate, combustion chamber modification, and exhaust gas
recirculation -- are treated separately in this section. This is done
since there is little likelihood they would be employed to meet these alter-
native standards. In addition, their wide range of possible penalties and
applications preclude meaningful analysis.
8-76
-------An error occurred while trying to OCR this image.
-------
electric power, however, could face a maximum Increase of nine percent.
Delivered gas prices will Increase by 0.4 percent when controls are fully
Implemented. No loss in jobs will take place nationwide. Local changes
will be minor because sales shifts among manufacturers will not be large.
U.S. oil Imports will Increase by 0.6 percent when controls are fully phased-
1n. Total Imports of goods and services will Increase by slightly more than
0.1 percent.
40 Percent Alternative
The Impacts would be similar to those above, but somewhat less 1n
the cases of capital budget requirements, product price Increases, and im-
port Increases. Capital budget requirements would be $4.5 million because
of the need to test fewer models; this is $0.4 million less than for a 60
percent standard. ElectroMotive diesel engines would still be vulnerable
to gas turbine competition. The increase in the total U.S. electric bill
would be 0.1 percent. The maximum increase for localities would be three
percent. Gas prices would increase 0.3 percent. Oil imports would rise by
0.4 percent and total imports by slightly less than 0.1 percent.
20 Percent Alternative
Capital budget requirements would be $4.1 million. The possibility
of any sales losses to turbine manufacturers would be remote. The total U.S.
electric bill would increase by 0.1 percent, and the maximum increase for
localities would be three percent. Gas prices would increase 0.1 percent.
Oil imports would increase by 0.2 percent and total imports by less than
0.1 percent.
Comparing the impacts among the various alternatives -- 60, 40, and
20 percent — there is no evidence that any of the alternatives would cause
an extraordinary impact. The following sections present a detailed dis-
cussion of the economic impacts that were considered.
8-78
-------
8.4.1 Impact on Manufacturers
The most direct economic effect of alternative standards of performance
1s on the manufacturers of large stationary reciprocating Internal combustion
engines. This Impact Involves three areas - capital budget requirements,
1ntra-1ndustry competition, and competition with gas turbines.
8.4.1.1 Capital Budget Requirements
To Implement NOX reductions for their engines, manufacturers will
require capital outlays to develop and test engine control techniques and to
maintain production of existing model engines under emissions regulations.
The size of these outlays will depend primarily on the number of models each
firm would need to test, the extent of further testing required, the fuel
prices paid during testing, and whether or not adequate laboratory facilities
were 1n place. The ability to finance the outlays will depend upon the
profitability of the engine Hne and the ease with which the Initial costs
could be absorbed by the companies' current capital resources.
A precise estimate of the outlays cannot be determined without a
detailed evaluation of specific control levels by each company. Certain
models have uncontrolled emissions that already meet the 20 and 40 percent
alternatives, although not the 60 percent alternative. These are shown
1n Table 8-19. Models that already meet the alternative levels without
controls would not require testing. Other models with high uncontrolled
emissions may be such a minor part of a company's business that the company
would drop them rather than test them with controls. The data as collected,
however, do not reveal the importance of individual models to the companies.
Furthermore, it is also possible that certain models would not have to be
tested because of the test results gained from other models. The amount of
8-79
-------
TABLE 8-19. THE CUMULATIVE NUMBER OF MODELS CURRENTLY ACHIEVING
VARIOUS ALTERNATIVE STANDARDS
Alternative Met
(Reduction from average
uncontrolled emissions)
60%
40%
20%
Uncontrolled
Diesel
0
3
6
11
Dual Fuel
0
1
1
5
Gas
0
1
2
23
Total
0
5
9
39
Notes: Totals include only engine models for which data are avail•
able. This includes 39 out of a total of 49 models.
previous testing and the stringency of the standard may also determine the
outlay required.
As discussed in Section 8.2.2, two phases of testing are required by
the manufacturer to establish that an engine can meet an emission standard:
1) development and 2) extended durability tests. The tests used as a data
base for this study will have met the development needs of manufacturers in
many cases. However, where models must be controlled more than 40 percent,
additional development tests may be needed. Such tests would cost about
$25,000 per model tested. Ingersoll-Rand and Alco may need to establish
in-house testing laboratories at an additional cost of $50,000 for test
instrumentation for each firm.
8-80
-------
After development testing, engine models must undergo extended
tests to prove the durability of emissions reductions and operations.
These tests would be about three times more costly than the development
tests. As noted in Section 8.2.2, the entire testing process can take 15-18
months. Several models can be tested concurrently, though several develop-
ment tests might be needed before a manufacturer can select the best control
technique. Overall, development and extended durability testing would cost
about $100,000 per engine model. This estimate is based upon confidential
correspondence from two manufacturers; it could overstate the actual costs
for the 20- and 40-percent alternative standards (which would need fewer
development tests).
The total number of internal combustion engine models produced by each
firm and the estimated capital budget requirement for testing to satisfy
standards of performance are shown in Table 8-20. The total industry bill,
including the cost of establishing new laboratories at two firms, would be
approximately $5,000,000 for a 60 percent alternative (requiring testing of
up to 49 models). The bill would be about $500,000 less ($4,5 million) for
the 40 percent alternative, and about $900,000 less ($4.1 million) for the
20 percent alternative, since fewer models would need testing (see Table 8-19).
The capital test requirements would be regarded as an added expense
for the manufacturers. The expense would be measured against the profitability
of each engine line. The larger the profits, the smaller the burden of the
expense. Manufacturers would either absorb the added expense by reducing
profits, pass it on to customers in the form of higher prices, or drop the
engine line as an uneconomic part of the business.
8-81
-------
TABLE 8-20. ESTIMATED CAPITAL BUDGET REQUIREMENTS
TO MEET NOX STANDARDS OF PERFORMANCE
Manufacturer
Colt
DeLaval
Caterpillar
Waukesha
El ectroMotl ve
Cooper Industries
(Cooper & Superior)
Ingersoll-Rand
Alco
Number of Models8
7
5
2
4
2
24
4
_!
Capital Budget Requirements
for NOX Standards of Performance'3
$ 700,000
500,000
200,000
400,000
200,000
2,400,000
450,000
150,000
All Firms
49
$5,000,000
aAn engine model 1s defined by a set of fuel, air charging, number of
strokes, and displacement per cylinder (bore and stroke) parameters.
bBased on an average test cost of $100,000 per engine model. Ingersoll-
Rand and Alco will have additional expenses of $50,000 each to establish
test laboratories.
A major question is whether the internal combustion engine manufac-
turers will have the financial resources from which to fund the initial
capital requirements. They can be funded either externally by increasing
debt or internally by using current capital budgets or allocating funds from
the capital budgets of other divisions In the parent company. Price increases
could be used in addition to these financing techniques to recover the
expenses over a number of years. If manufacturers were to seek to recover
the annual 1 zed cost of test outlays over a five-year period, on average they
8-82
-------
would have to ratse engine prices only one percent (ranging among manufac-
turers from 0.4 to 2.1 percent).
Although any combination of the above financing techniques could be
used, engine divisions are likely to operate from their own internal resources.
Unlike investments in new products or plants, the capital test outlays would
represent investments in established lines of business and thus, would entail
much lower risk. Moreover, these divisions are small parts of their parent
companies, as Table 8-5 in Section 8.1.2 shows. Therefore, raising debt or
obtaining funds from other capital budgets would not be difficult for such
an investment.
Although the precise capital budgets cannot be determined without a
detailed evaluation of a specific control level by each company, as previously
mentioned, the prospects for internal financing can be put in perspective by
comparing capital budget requirements with the sales and profits of internal
combustion engine operations. This can be done in a rough manner by compar-
ing parent companies' profits as a percentage of sales (see Table 8-21) to
their internal combustion engine divisions' additional capital budget re-
quirements as a percentage of sales. The internal combustion engine divisions'
test requirements as a percentage of sales cannot be shown in order to pre-
serve the confidentiality of sales data disclosed by manufacturers. However,
in no case did the percentage exceed five percent of sales or the ratio of the
parent company's after-tax profits to sales. This is true even in the case
of Colt or ElectroMotive where parent company profits fell below five percent
of sales during the 1975-1976 period.
8-83
-------
TABLE 8-21. FINANCIAL RESOURCES
Consolidated Parent Firm of Profits as
Engine Manufacturer 1976
Ingersoll-Rand
General Motors (ElectroMotive)
Colt Industries
Caterpillar
Cooper Industries (Cooper & Superior)
Dresser (Waukesha)
TransAmerica (DeLaval)
General Electric
5.6
6.2
4.9
7.6
7.5
7.0
a
5.9
% of Sales
1975
7.0
3.5
5.1
8.0
6.5
6.2
a
4.3
All Firmsb 6.4% 5.8%
aTransAmerica is primarily a financial corporation whose sales
are not easily compared with sales of products of manufactur-
ing corporations.
bAyerage profit margin computed as a simple average for all
firms for which data were available.
SOURCE: Securities and Exchange Commission Corporate 10-K
reports and annual reports to stockholders.
Providing that the internal combustion engine part of the business
has approximately the same profit margins based on sales as the parent com-
panies, this indicates that in the case of each manufacturer, the additional
capital budget could be financed with funds generated internally from domes-
tic sales of large stationary reciprocating internal combustion engines. In
addition, testing expenses would not be subject to taxes which would otherwise
be applied if the firm earned profits of that amount. For example, if
8-84
-------
$100,000 of expenses were Incurred instead of realizing $100,000 of gross
profit subject to the marginal federal corporate tax rate of 48 percent, the
corporation would have $52,000 less cash flow and the U.S. Government would
receive $48,000 less taxes. In effect, the expenditure would cost the corpora-
tion $52,000 rather than the nominal $100,000 if it were absorbed from profits
and not passed on through price increases.
In the event standards were set that required exhaust gas recirculation
or combustion chamber modification, the capital budget requirements would
change. Based on confidential correspondence with manufacturers, it is esti-
mated that test requirements for the industry would double -- to ten million
dollars -- and the testing time required would double or triple — to three to
five years. Hence, although cost doubles, the time involved doubles or
triples. This means that the annual requirement stays the same or decreases.
This capital requirement would compete more for a company's overall resources
because it is larger and more extended. If only one manufacturer had to incur
this amount, it might mean less resources for normal product development or
investment and thus some competitive disadvantage.
8.4.1.2 Intra-Industry Competition
Manufacturers would not have significant differential impacts for
alternative standards that require spark retard, air-to-fuel changes, or mani-
fold temperature reduction. If derate, exhaust gas recirculation, or combus-
tion chamber modification were required, cost penalties among engines would be
disparate and might cause competitive shifts in the sales shares of manufac-
turers.
To identify whether or not significant changes in sales would take
place, the engine penalty data were analyzed in conjunction with confidential
8-85
-------
sales data for each manufacturer. Worst-case impacts were determined by
looking at the maximum variation possible among engine penalties in a partic-
ular market. Where the range of penalties shown in Table 8-16 of Section
8.2.1.3 was small, such as In gas engines, one company would not have an
advantage over another company. For instance, at the 40 percent alternative
for gas engines It 1s possible that Waukesha could Incur a seven percent
penalty and Caterpillar a two percent penalty or vice versa, since both of
them would Incur cost penalties of from two to seven percent.^ The maximum
differential cost penalty for these two competitors in the gas production
market is then five percent (seven percent less two percent).
Cross-price elasticities were then considered using the above results.
One spokesman estimated that, in one of their highly competitive markets, he
would expect a 10 percent increase in price to lead to a 20 percent decrease
in sales.2/ Another spokesman, referring to the internal combustion engine
market as a whole, estimated that a five percent increase in price would not
have any noticeable effect on their sales, but that a 10 percent increase,
even industry-wide, would lead to a 10 percent decrease in sales. At markups
beyond 10 percent, we have no estimates of the price elasticities, but have
assumed that every one percent increase in price would result in a two percent
decrease in sales.
i/It should be noted that just because the data show that CateH" l
have to reduce its emissions by an average of 46 percent versus 33 percent
for Waukesha, it does not mean that the Waukesha penalty could not fall
iSto the high end of the range and the Caterpillar penalty into the low
end of the range.
i/This assumes sales are in terms of dollars rather than units. If sales
werl based on units, the impact on dollar sales would be less.
8-86
-------
A third factor considered was the Importance of each market to the
companies' total Internal combustion sales. If standby and export markets
(which make up almost half of all large stationary reciprocating Internal
combustion engine sales) are exempt from standards of performance, this
would leave manufacturers with a substantial part of their business unaffected.
In addition, as described 1n Chapter 9.3, smaller bore engines will be
exempt from proposed standards of performance. Therefore, only a portion of
a manufacturers' engine sales will be affected by standards. Thus differential
cost penalties arising from standards would lead to only limited Impacts on
sales shares.
In combining these factors 1n a hypothetical example, a manufacturer
Incurring a possible six percent penalty over his competitors would lose 12
percent of his sales 1n that market (assuming the worst-case cross-price
elasticity), which might be 25 percent of his total sales to all markets —
hence, he would lose only three percent of his total sales. (It should be
noted that these are large stationary reciprocating Internal combustion
engine sales, not total engine sales or parent corporation sales which would
make this percentage much smaller.)
In addition, one manufacturer Indicated that parts and services
accounted for over 25 percent of his annual sales.(82) Since standards of
performance would not affect the outstanding population of engines, parts,
and services, revenues would provide a stabilizing factor for all manufac-
turers in the short run, though this would lessen over time.
In the following analysis of intra-industry competition for each of
the major submarkets, sales losses of more than 10 percent in any market were
used to identify significant effects. For most manufacturers, the potential
8-87
-------
sales loss was considerably less than 10 percent in most of the submarkets
they participated. Moreover, these potential losses were constructed
from an admittedly extreme set of assumptions concerning the penalty dif-
ferential, cross-price elasticities, and cost pass-through. The exact
percentage of potential sales loss that could occur under these conditions
was withheld to protect the confidentiality of the data.
8.4.1.2.1 Electricity Generation
There are two markets to discuss in the area of electricity genera-
tion. The first is the dual fuel market supplied by Colt, DeLaval and Cooper/
Superior. Based on uncontrolled emission data, DeLaval has a distinct advan-
tage over Colt and Cooper/Superior at the 20, 40 and probably 60 percent
alternative levels. The maximum differential impact would be six percent,
but no manufacturer would lose 10 percent of its internal combustion engine
sales at any of the three control levels.
The second market is the diesel fuel market supplied primarily by
Colt, DeLaval, Cooper/Superior, and ElectroMotive. Here, a maximum differ-
ential impact of 18 percent is possible at the 40 percent level of control
with ElectroMotive suffering the disadvantage. Nevertheless, ElectroMotive
has substantial mobile engine sales for locomotives which would be unaffected
by proposed standards. It is important to note that the 14-18 percent pen-
alties shown in Table 8-16 are based on data from one engine model. Such
limited test data do not necessarily reflect the range of penalties for strin-
gent alternatives since not all engines were tested. Moreover, it is possible
that some models might be able to attain low levels of emissions only through
i
techniques like derate or combustion chamber modification at much higher
penalties. These would be outside the range shown in the table. It is the
-------
likelihood of such additional penalties, rather than the data shown in the
test sample, that lead to the possibility of significant differential impacts
at stringent levels of performance standards. For these reasons, no firm
conclusions can be drawn except that even with an 18 percent differential no
manufacturer would lose 10 percent of its internal combustion engine sales.
8.4.1.2.2 Gas Production
Ingersoll-Rand, Cooper/Superior, Waukesha, Caterpillar, DeLaval and
Colt sell internal combustion engines to the gas production market. As pre-
viously mentioned, the average cost penalties for manufacturers would range
from one to seven percent -- a maximum differential of six percent. At
worst, this would mean that the most any manufacturer would lose would be
12 percent of its sales in this market, but no manufacturer would lose 10
percent of its internal combustion engine sales for all markets at any of
the three alternative levels.
8.4.1.2.3 Gas Transmission
Cooper, DeLaval, Ingersoll-Rand and Waukesha sell internal combustion
engines to the gas transmission market. Because the engines are gas-fueled,
cost penalties are similar to those in the gas production market, and no
manufacturer would lose 10 percent of its internal combustion engine sales
at any of the control levels.
8.4.1.2.4 Other Markets
All manufacturers have sales to other markets. These include gas-
fueled, dual-fueled, and diesel engines. However, the applications involved
are diverse and comprise what are actually many different segmented markets.
It would not be accurate to characterize differential penalties across such
diverse applications. As noted in subsection 8.4.1.2.3, gas-fueled engines
8-89
-------
have a low differential penalty and therefore, would not Involve much change.
Dual-fuel sales to other markets are too small to have a significant impact.
Only Colt and ElectroMotive sell diesel engines to other markets. Again, it
can be expected that the other markets for diesel engines include many
diverse applications. For a 20 percent control alternative, the maximum
differential penalty would be five percent. For 40 percent, the maximum
average penalty differential increases to 15 percent. This still would not
result in a 10 percent loss 1n internal combustion engines sales for either
firm, even if their applications overlapped completely. For the 60 percent
alternative, the maximum differential is only four percent for the two firms
(but is based on limited data).
8.4.1.2.5 Standby. Export, and Small Engines
Standby and export sales of large stationary reciprocating Internal
•combustion engines accounted for 44 percent of all sales of large stationary
reciprocating internal combustion engines (by horsepower) in the years 1972-
1976 for all the manufacturers. In addition, data on sales of stationary
reciprocating Internal combustion engines below the regulated size limits
(see Chapter 9.3) were not available for this analysis. Assuming these
applications are exempt from proposed standards of performance, possible
percentage loss of sales for each manufacturer is reduced further. By
focusing only on a small section of the NOX emitting stationary reciprocating
engine population (which nonetheless emits the bulk of NOX from installed
sources), the proposed standards of performance affect less than half of the
total stationary reciprocating internal combustion engine sales of manufac-
turers.
8-90
-------
8.4.1.2.6 Aggregate Impact on Manufacturers
Because of the broad range of most control penalties, it 1s not
always clear whether the engine manufacturers will gain, lose, or stay even
In a given market. This 1s especially true where combustion chamber modifi-
cation, derate, or exhaust gas redrculation is involved. It is reasonable
to assume that the disadvantages some manufacturers may face in one market
may be at least partly offset by advantages they gain in another market.
This is highly dependent on the manufacturers' product mix and the cross-
price elasticities of each market. It has also been shown 1n the previous
subsections that the Industry as a whole is buffered by substantial sales
on nonregulated Internal combustion engines. The market segmentation within
major markets and the importance of other application factors would also
soften any impacts.
Of the seven manufacturers studied for intra-industry impacts
(Caterpillar, Colt, Cooper/Superior, DeLaval, Electromotive Division of
General Motors, Ingersoll-Rand, and Waukesha), only Colt and ElectroMotive
appeared to have a clear disadvantage in more than one market at certain
control levels. Partly because Colt's sales are concentrated in the highly
competitive electricity generation equipment market, and partly because their
nonregulated sales are not nearly as significant as ElectroMotive's, Colt
could potentially suffer the most significant intra-industry impact.
Assuming that Colt, the most vulnerable manufacturer, were to suffer
the most extreme differential in each of the markets in which they partici-
pate, and assuming the worst possible cross-price elasticities, Colt would
suffer a loss in sales of about six percent.^/
Across-the-board standards of 20, 40, and 60 percent could result in six
six, and five percent sales losses, respectively, by Colt. '
8-91
-------
The conclusion that intra-industry impacts would be sustainable and
not cause any major dislocations within the indsutry holds for the degree of
control for which the test data represent comprehensive ranges. At stringent
levels for the standards (e.g., 60% reduction from uncontrolled data), the
possibility arises that some models might require expensive control tech-
niques that would widen the ranges and differentials. Only further testing
can ascertain how significant this would be. It is also possible that even
at the most stringent levels of control, the differential impact might be
insignificant.
8.4.1.3 Competition From Gas Turbines
To assess the possible inroads turbines might make in the reciprocat-
ing engine market as a result of performance standards on new 1C engines,
market structure, cost, and other factors must be considered. The three
major markets for turbines and reciprocating engines (electricity genera-
tion, oil and gas production, and oil and gas transport) are segmented into
several submarkets in which factors such as size, weight, durability, relia-
bility, vibration, and ability to handle load variations often dictate the
choice of engine. Previous experience with the vendor, reputation, service,
and familiarity with existing equipment are usually important considerations
in the replacement market.
Turbines do not compete with reciprocating engines based on annualized
costs alone, due to their higher operating (fuel) costs, at least not in the
normal operating range of reciprocating engines which is 6,000 to 8,000
horus per year. As Table 8-12 shows, reciprocating engines controlled to the
8-92
-------
60 percent level are less expensive to own than even uncontrolled turbines,
with one possible exception in the electricity generation market U/
Table 8-22 also shows that for all the other markets, even the maxi-
mum penalty which could be imposed on internal combustion engines would not
bring the cost of owning internal combustion engines up to that of turbines.
Proposed NOX NSPS standards for gas turbines used in electricity
generation are estimated to increase their cost by about two percent, and
for oil and gas transportation and production applications by one to four
percent. Consideration of the proposed NOX NSPS will not affect the con-
clusions which can be drawn from this table.
TABLE 8-22. TURBINE VERSUS RECIPROCATING INTERNAL
COMBUSTION ENGINES BREAKEVEN ANALYSIS
Application
Electricity
Generation
Oil and Gas
Transporta-
tion
Oil and Gas
Production
Fuel
dual fuel
diesel
gas
gas
Maximum Reciprocating
Engine Cost Penalty
(as % of total annual ized costs)
NOX Reduction Alternative
20%
6%
8%
7%
7%
40%
6%
18%
7% •
7%
60%
6%
18%
7%
7%
Breakeven3
for 6000 to
8000 hr/yr
35-39%
12-14%
13-14%
25-30%
aTh1s represents the total annualized cost penalty which would have to be
eeoXuPaT?h?? "? T 1prOCft1,"? ;|"t«™l coJbSstionyenginesTfore they would
equal the cost of uncontrolled turbines.
— New Source Performance Standards were proposed for stationary gas turbines
in the 3 October 1977 Federal Register, Volume 42, Number 191.
8-93
-------
8,4,1,3.1 Electricity Generation
Data shown 1n Table 8-22 of the cost analysis section Indicate that
cost penalties of 14-18 percent could be Incurred by some dlesel engine
manufacturers due to controls, One manufacturer, Electromotive, would Incur
that penalty even at a 40 percent level of NOV reduction, For all other
A
manufacturers, turbines would still cost more,
Table 8-22 shows that a 12-14 percent cost penalty would bring dlesel
reciprocating engines used 1n electricity generation up to the point where
they would have no cost advantage over uncontrolled turbines. Since turbines
would only Incur about a two percent penalty for the proposed NOX new source
performance standard controls/ ' it appears that they would become competi-
tive with diesel reciprocating engines. This conclusion, however, is based
on only one data point used to calculate the 14-18 percent diesel engine
penalty and, as such, should not be used as a decision criteria, considering
that some uncertainty exists at the greater control alternatives (60 and
40 percent) for which there are test data. Furthermore, it is unlikely that
turbines would replace diesel engines in plants using banks of smaller recip-
rocating engines, unless the entire bank were replaced with one turbine.
At the 20 percent alternative, turbines do not compete with the in-
ternal combustion engines on a cost basis. As load factors decrease, how-
ever, turbines become increasingly competitive due to their lower capital
costs as shown in Figure 8-11. The formulas and methodology used to pro-
duce this and other graphs illustrating the breakeven points in Table 8-22
are shown separately on Table 8-23.
8.4.1.3.2 Oil and Gas Transportation
Although the reciprocating gas engines in gas transportation lose a
large portion of their cost advantage over turbines in the high end of the
8-94
-------An error occurred while trying to OCR this image.
-------
TABLE 8-23. METHODOLOGY FOR CALCULATING THE BREAKEVEN CURVES FOR FIGURES
8-10, 8-11, AND 8-12
Cost (in mills/kWh) = - + OM + P(E)
The breakeven factor for the cost penalty to reciprocating engines (X)
occurs where:
OMt + P(Et) - XT + OMr
Solving for the breakeven factor:
P(Et)
X =
(CRF)K + OM +
Hr/Yr r u r
The percentage breakeven cost penalty = lOO(X-l), where:
X = breakeven factor
CRF = capital recovery factor (i.e., .2 = 20%)
K = installed cost of engine or turbine ($/kW)
OM = operating and maintenance cost ($/kWh)
P = price of fuel (i.e., distillate $2. 50/MMBtu; gas $2. 00/MMBtu)
E = heat rate efficiency in MMBtu/kWh
t = turbines
r = large reciprocating internal combustion engines
^Assuming a 20-year accounting life, 10 percent interest rate, and four >
percent fixed capital charge (includes property taxes, insurance, administra-
tion and overhead).
8-96
-------
penalty range presented in Table 8-22, they are generally used in different
applications, and substitution would not be indent, even at the breakeven
point of total annualized costs.
Gas turbines in gas compression uses serve primarily mainline transrnis-
sion roles and provide power 1n new installations. Gas turbines are larger,
so one turbine would provide the same power as several internal combustion '
engines. When existing reciprocating engine compression installations are
expanded, however, reciprocating engines are purchased unless the old engines
are scrapped. Moreover, turbines power centrifugal compressor equipment,
while internal combustion engines power reciprocating compressor equipment
Hence, existing stations would not mix or change motive foms unless they
also changed their compressor equipment.
Furthermore, turbines are not as well suited for gathering, storage
or prescription where the flow is highly variable and the discreate power
requirements are in the range that internal combustion engines offer Thus
engines are typically located at the distribution end of pipelines, whereas'
turbines are located on the main trunk lines.
In addition, high load factors (characteristic of compressor applica-
tions) favor the more efficient internal combustion engines, especially as
9as prices rise. Figure 8-12 shows that internal combustion engines maintain
this advantage, even at load factors as low as 1,000 hours per year.
8.4.1.3.3 Oil and Gas Production
Turbines are clearly uncompetitive on purely a cost basis in this
market. They are less efficient and much more expensive than internal combus-
tion engines and are primarily used on offshore rigs where lighter and more
portable equipment is a necessity. Uncontrolled turbines used for extraction
8-97
-------An error occurred while trying to OCR this image.
-------
purposes are at least 25 percent more expensive to own than Internal combus-
tion engines, as Figure 8-13 shows,
8-4.2 Impact on Product Prices
NOX standards will affect product prices primarily for electricity and
natural gas. However, cost Increases will be insignificant, even if emissions
standards are set at 60 percent average reductions. The analysis 1s explained
below in terms of 20, 40, and 60 percent NOX reduction alternatives, Because
there are three kinds of engines (diesel, dual fuel, and gas) and three
alternative levels (20, 40, and 60 percent) for emissions standards, the
number of possible options for standards would be too large to cover in a
readily comprehended manner. To simplify the presentation, this analysis
considers across-the-board standards for all the fuels together.
8.4.2.1 Cost Pass-Through
When the demand for a manufacturer's goods is inelastic (i.e., In-
sensitive to price changes), added costs are likely to be transferred to the
consumer. This cost pass-through occurs where the manufacturer's product
1s viewed as a necessity by the consumer with few, if any, substitutes avail-
able. The extent of pass-through, then, depends on many factors in addition
to the unique qualities of the product - the price elasticity of demand for
the industry as a whole (determined in part by other demands which are in-
directly related) and the cross-price elasticities of the products within
the same industry. It is assumed that since manufacturers produce recipro-
cating engines for essential applications in specialized market, cost in-
creases would be passed through to the consumer.
Section 8.4.1.3 demonstrated that the internal combustion engine
industry does not now face competition from gas turbines based on cost factors
8-99
-------An error occurred while trying to OCR this image.
-------
X
alone for either of the two main applications-electricity generation and
gas compression during production and pipeline transmission--in which NOV
standards will cause cost penalties. In electric generation, it may be
recalled, turbines are uneconomic for baseload use and for additions to
plants already using banks of reciprocating engines. In gas compression,
gas turbines are not suited to gathering applications (because of variable
gas flow) and are not easily added to stations already using reciprocating
engines.
Furthermore, the industry is segmented into many specialized markets
and submarkets, as discussed in Section 8.1.3. Most internal combustion
engines have carved out a niche in which they have few, if any, substitutes
in the applications for which they are used.
At the same time, demand for engines would not likely fall significantly
as a result of higher prices for engines in the range that the test data in-
dicate would take place. The increase in some localities for electric rates
would be, at its maximum, somewhat higher—nine percent—but is much less
than recent increases from other causes, while past rate increases have not
led to much decrease in demand for electricity (see subsection 8.4.2.2).
Municipalities have the option of purchasing power from larger utilities,
but tend to produce the bulk of their own needs and purchase only small amounts
of power. (Frequently, these utilities sell power.) Furthermore, consumers
would not consume appreciably less gas because of a 0.4 percent increase in
delivered prices (see subsection 8.4.2.3).
With little decrease likely for overall engine demand and little compe-
tition from substitutes, manufacturers will likely pass through cost increases
to consumers.
8-101
-------
8,4.2.2 Electricity Generation
Reciprocating Internal combustion engines are used to generate elec-
tricity on a continuous (baseload) basis for small municipalities. They
account for only a small fraction of total electricity generation In the United
States. There were 936 Internal combustion plants (excluding gas turbines) at
year-end 1975,(84) These had a total generating capacity of 5,021 megawatts-
equal to one percent of the total U.S. electric generating capacity of 505,772
megawatts.(85) Therefore, the Impact on electricity prices is best measured
1n two dimensions-local and weighted national impacts.
Although localities deriving all of their electricity from internal
combustion are rare, and the chances of all of those engines being subject
to NSPS are remote at least in the near future, an evaluation of this case
was done to provide the maximum possible increase in electricity prices
which could be experienced by any consumer. Based on a sample of ten utili-
ties using internal combustion engines to generate at least 90 percent of
their electricity sales, engine costs (engine price, maintenance, and fuel
expenses) typically account for half the costs of delivered electricity to
consumers.(86) Electricity distribution and general overhead costs account
for the remainder and would be unaffected by NOX standards. Therefore, an
engine penalty from NOX standards would be halved when applied to the price
that consumers pay for electricity in these localities. Table 8-24 shows
the inflationary impact on local electricity prices for various emissions
standards.
The maximum impact is nine percent and takes place in the case of
diesel engines at a 60 percent alternative (based on data from only one
model). Recent sales of diesel and dual fuel engines to the electric
8-102
-------
TABLE 8-24. MAXIMUM INFLATIONARY IMPACT ON LOCAL ELECTRICITY PRICES
Engine type
Diesel
Dual Fuel
Alternative NOX
Reductions
20% 40* 60*
3* 3* 9*
2% 3% 3%
generation market Indicated that new sales were running approximately 50
percent dlesel and 50 percent dual fuel.
In comparison, electric rates have Increased far more than other
factors. Since 1970 residential, commercial, and Industrial electric rates
have risen by five percent to 30 percent annually. This 1s shown 1n Table
8-25.
TABLE 8-25. ^STORICAL^PERSPECTIVE ON ANNUAL INCREASES IN
Period
•MMMMMMM
1970 to 1971
1971 to 1972
1972 to 1973
1973 to 1974
1974 to 1975
1975 to 1976
Range
Residential
500 kWh
H 6.2*
'2 7.6*
'3 4.6*
'4 12.4*
5 27.3*
6 7.2*
Commercial —30, kW,
6,000 kWh
5.5*
7.3*
4.9*
11.1*
24.8*
6.3%
Industr1al«300 kW,
60,000 kWh
•MMBMMM
7.6*
8.5*
5.7*
15.6*
30.6*
6.5*
4.6*-27.3*
4.9*-24.8*
5.7*-30.6*
SOURCE.: Federal Power Commission, Typical Electric Bills. 1976.
8-103
-------
Because internal combustion engines account for just a small fraction
of all electricity generated nationwide (with the national electric bill from
private utilities alone valued at $44.4 billion in 1976),(87) the NOX emis-
sions standards on new source reciprocating internal combustion engines would
raise the national electric bill by just a fraction of a percent. If the
standards had been fully implemented in 1976, the inflationary inpact on the
national electric bill would have been just 0.1 percent in the case of 20 or
40 percent alternatives and 0.3 percent in the case of a 60 percent alterna-
tive. This is shown in Table 8-26.
TABLE 8-26. MAXIMUM INFLATIONARY IMPACT ON THE NATIONAL
ELECTRIC BILL WITH FULL PHASE-IN
Impact
Increase
Alternative NOX
Reductions
20% 40%
0.1
60%
% 0.1% 0.3%
In actuality, new source performance standards are phased-in only
gradually. Penalties are not incurred until new controlled engines are
purchased, while old uncontrolled engines are retired from service after a
30-year lifetime. Assuming new sales equal to three percent of the existing
engine population and retirements also equal to three percent (in effect, a
steady population as indicated by recent sales data), after five years,
approximately 15 percent of all engines will be controlled and will incur
penalties. The inflationary impact of the standards at that time is shown
in Table 8-27. It ranges from 0.02 percent for a 20 percent standard to
0.04 percent for a 60 percent standard.
8-104
-------
TABLE 8-27. INFLATIONARY IMPACT ON THE NATIONAL ELECTRIC
BILL, AFTER FIVE YEARS
Impact
Increase
Al
20%
0.02%
ternative NOX
Reductions
40% 60%
0.02% 0.04%
8-4.2.3 Gas Production and Transmission
Reciprocating internal combustion engines are used to transport most
of the gas consumed in the United States. Internal combustion engine costs
account for only six percent of the delivered price of gas ($1.60 per Mcf in
(88)
1976), though the percentage is somewhat higher for areas distant from
producing states. The average cost penalties for gas engines (which account
for nearly all recent sales to the reciprocating pipeline engine market)
ranged from two percent at a 20 percent alternative to six percent at a
60 percent alternative. Table 8-28 shows the impact of these penalties on
delivered gas prices.
TABLE 8-28. INFLATIONARY IMPACT ON DELIVERED GAS PRICES
AFTER FULL PHASE-IN
Impact
Increase
Al
20%
0.
1 V
\ 7o
ternative NO
Reductions
40%
0.3%
X
60%
0.
4%
The largest inflationary impact, at a 60 percent alternative, would
involve a price increase of just 0.4 percent. (This was calculated as if
8-105
-------
all engines 1n use 1n 1976 had been controlled and Incurred penalties.) By
comparison, the national average delivered price of gas has risen by 103 per-
(89)
cent 1n recent years-from $.79 per Mcf 1n 1973 to $1.60 per Mcf 1n 1976--
from other causes,
The Increase 1n gas prices would not reach the above levels until all
engines were controlled. With sales and retirement rates equal to those 1n
the electric generation market (I.e., three percent annually), 1n five years
controls would cover 15 percent of all engines, The Inflationary Impact
at that time would be a 0.06 percent Increase 1n the case of the most stringent
standard (60 percent). This 1s shown 1n Table 8-29.
TABLE 8-29. INFLATIONARY IMPACT ON DELIVERED GAS PRICES,
AFTER FIVE YEARS
Impact
Increase
Alternative NOX
Reductions
20% 40% 60%
0.02% 0,04% 0.06%
8.4.2.4 Impacts Over Five Years
Users of Internal combustion engines will have to lay out additional
capital expenditures to purchase more expensive engines (the engine purchase
price component of the cost penalty from NOX controls). In the case of In-
ternal combustion engines, however, the capital cost penalty 1s small. Most
of the penalty comes from higher fuel or maintenance costs. A two percent
engine price penalty can be expected on average for all alternative
8-106
-------
standards.^ with annual Industry domestic non-standby Internal combustion
engine sales of $96 million (projected to remain roughly constant), the
additional capital cost for users would equal $1.9 million per year -- a
total of $9.6 million on a cumulative basis over the first five years.
Total costs 1n the fifth year (Including amortized capital costs,
maintenance costs, and fuel costs) would range from $16 million for a 20 per-
cent alternative to $26 million for 40 percent and $45 million for 60 percent.
In dollar terms, the Impact of the standards 1s shown 1n Table 8-30
for all markets 1n the fifth year after the standards are Implemented.
TABLE 8-30. COSTS OF VARIOUS ALTERNATIVE STANDARDS IN
THE FIFTH YEAR (IN MILLION DOLLARS)
. ,. . Alternative NOX Reductions
Application 20% 401 60%
Gas Production & Trans-
mission 5.9 13.6 17.8
Electric Generation 7.9 8.9 19.9
Other Applications 2.2 3.4 7.1
All Applications 16.0 25.9 44.8
12/
-^Two percent reflects a one percent increase from pass-through of test costs
and an average of one percent increase for use of manifold temperature
reduction in some cases. (Manifold temperatures reduction would incur a
price increase of two percent where used, but would probably be used less
frequently than other techniques.)
8-107
-------
8.4,3 Impact on Employment
Since NO standards will not cause significant changes in manufac-
A
turers' sales if retard, manifold temperature reduction, and air-to-
fuel changes only are used, they will not cause significant impacts on
employment.
Nationally, sales shifts among manufacturers will tend to balance out
with no decrease in aggregate sales or employment in the industry. As noted
in Section 8.4.4 (below), exports and imports of internal combustion engines
will not likely experience any changes because of NOX standards on engines
used in the U.S. Therefore, few jobs would be lost to foreign firms. More-
over, because sales changes of greater than 10 percent are not expected
(and most would be much less), the extent of local shifts will also be minor.
8.4.4 Impact on Foreign Trade
The foreign trade balance will not be significantly affected by NOX
standards. The amount of imported engines is not expected to change. Most
internal combustion engine imports have been engines that are smaller than
the minimum size to be controlled by the proposed standards. Department of
Commerce data showed that the average value of imported diesel engines in 1973
was $889 per unit.(90) This is far below the likely cost for a diesel engine
of the size that would be controlled by the proposed new source performance
standards since a typical 1,000 horsepower diesel engine would cost about
$150,000. In addition, imported engines would have to meet NOX emissions
standards. Because foreign firms would have smaller U.S. sales volumes over
which to spread capital test requirements for NOX reductions, the NOX stand-
ards would actually tend to create a barrier to imports.
Proposed NOX standards do not apply to engine exports. Control tech-
niques like retard, air-to-fuel changes, manifold temperature reduction, and
8-108
-------
derate do not Involve changes in mass production items. There would be no
loss in scale economies to manufacturers if they had to produce controlled
engines for domestic sales and uncontrolled engines for export sales. More-
over, even techniques like combustion chamber modification and exhaust gas
recirculation would have little impact on exports, because of the specialty
nature of the large engine industry, where even without controls, each
engine is typically tailored to a specific customer's needs rather than Just
mass produced.
Fuel imports would be increased marginally by the fuel penalties
involved in meeting NOX standards. In the fifth year after standards take
effect, 15 percent of all engines will be controlled. The additional fuel
requirement (to be met by additional imports of oil) in that year would be
1.0 million barrels of oil for a 20 percent alternative, 1.5 million barrels
for 40 percent, or 2.4 million barrels for 60 percent. This 1s shown in
Table 8-31.
TABU' 8-31. ADDITIONAL FUEL NEEDS, IN THE FIFTH YEAR
(MILLION BARRELS)
Engine
Type
Gas
Diesel
Dual Fuel
Alternative
20%
0.6
0.3
0.1
NOX Reductions
40% 60%
1.0 1.0
0.4 1.4
0.2 0.1
All Engines 1.0 1.5 2.4
8-109
-------
(91)
Measured against 1976 oil Imports of 2,850 million barrels/" the
Increase 1n oil Imports would be 0.04 percent for a 20 percent alternative,
0.05 percent for 40 percent, and 0,09 percent for 60 percent. Measured
against 1976 Imports of all goods and services of $160 billion and valued at
an average of $12.10 per barrel/92' the Increase 1n U.S. Imports would be
0.008 percent for a 20 percent alternative 0.011 percent for 40 percent, and
0.018 percent for 60 percent.
8.4.5 Summary of Economic Impact Analysis
With minor exceptions, there appears to be little difference among
the Impacts of a 60 percent, 40 percent, or 20 percent alternative. This
1s a qualified judgment based on the assumption that derating, exhaust gas
redrculatlon, or combustion chamber modification will not be necessary 1n
order to meet these alternatives.
Although aggregate price Impacts on consumers vary directly with the
level of control, even a 60 percent alternative would raise gas and elec-
tricity prices less than half a percent. On the local level, however, 1t
1s theoretically possible for a locality entirely dependent on new reciprocat-
ing Internal combustion engines for generating Its electricity to experience
a nine percent Increase 1n electricity prices at the 60 percent alternative.
This 1s three times more than the Increases for the 40 or 20 percent alter-
native. Such localities, however, would represent an extremely small part
of the overall population.
Manufacturers would face only limited Impacts. Capital test require-
ments would be within their ability to finance Internally from profits,
while the costs could be recovered through a one percent average price
Increase over a five-year period. Despite variations 1n cost penalties, no
8-110
-------
ffrm 1s Hkely to lose more than five to six percent of Us sales (In all of
their markets). Gas turbines may make Inroads Into certain manufacturers'
sales for dlesel engines only at a 40 or 60 percent alternative, but their
cost advantage in the normal operating range of 6,000 to 8,000 hours per
year would be slight. Employment, like sales, would experience little
change. Imports and exports of internal combustion engines would also face
little change. Oil imports would increase by only a fraction of a percent.
Several additional points can be noted. In summary, the full impact
of the standards would not be markedly different for the 20, 40 and 60
percent alternatives. Second, based on conservative analytic techniques,
the only possible impact of notable magnitude would be the rise in electric
prices for isolated localities using all internal combustion engines. Third,
full attainment of the impacts on users will not be realized until all engines
are replaced by controlled engines incurring penalties -- a process that will
take 30 years to complete.
8-111
-------
REFERENCES FOR CHAPTER 8
(1) Diesel and Gas Turbine Worldwide Catalog. Diesel and Gas Turbine
Progress. Volume 39. 1974.
(2) Current Industrial Reports, Internal Combustion Engines. U.S.
Department of Commerce, Bureau of the Census, Washington, D.C. MA-35L.
1968, 1970, 1972, 1973, 1974, 1975.
(3) Shipments of Internal Combustion Engines, 1958-1967. U.S. Department
of Commerce, Business and Defense Services Administration. Washington,
D.C. 1969.
(4) Hare, C. T. and K. J. Springer. Exhaust Emissions from Uncontrolled
and Related Equipment Using Internal Combustion Engines, Part 5 -
Heavy-Duty Farm, Construction, and Industrial Engines. Southwest Research
Institute, San Antonio, Texas. AR-898. October 1973.
(5) Op. Cit., Reference 2.
(6) Op. Cit., Reference 3.
(7) 16th Annual Pipeline Installation and Equipment Costs. Oil and Gas
Journal, p. 81. August 13, 1973.
(8) McGowin, C. R. Stationary Internal Combustion Engines in the United
States. Environmental Protection Agency. EPA-R2-73-210. April 1973.
(9) The Fortune Listing of the 500 Largest U.S. Industrial Corporations,
Fortune, May 1977.
(10) Ibid.
(11) Norton, G. S. (Acurex) and A. L. Foltz, Jr. (Enterprise Engine
Division, DeLaval Turbine). Private Communication. July 29, 1974.
(12) Offen, G. R., (Acurex) and R. D. Henderson, et al. (Caterpillar
Tractor Company). Private Communication (Memorandum to supplement Trip
Report dated August 30, 1974). March 21, 1975.
(13) Offen, G. R. (Acurex) and C. T. Ahlers, et al. (Cummins Engine
Co.) Private Communication (Memorandum to supplement Trip Report dated
August 30, 1974). August 30, 1974.
(14) Offen, G. R. (Acurex) and F. Schaub (Cooper-Bessemer Co.). Private
Communication. October 13, 1975.
8-112
-------
(15) Goodwin, D. R. (EPA/Emission Standards and Engineering Division)
Request for Information. June 16, 1976.
(16) Snyder, W. E. (Waukesha) and D. R. Goodwin (EPA). Private
Communication. July 20, 1976.
(17) Thompson, M. P. (Superior) and D. R. Goodwin (EPA). Private
Communication. August 3, 1976.
(18) Hanley, G. P. (CMC) and D. R. Goodwin (EPA). Private
Communication. September 7, 1976.
(19) Greiner, S. D. (Cooper) and D. R. Goodwin (EPA). Private
Communication. August 4, 1976.
(20) Newton, C. L. (Colt) and D. R. Goodwin (EPA). Private
Communication. August 2, 1976.
(21) Fleischer, A. R. (DeLaval) and D. R. Goodwin (EPA). Private
Communication. July 30, 1976.
(22) 1974 Report on Diesel and Gas Engines Power Costs. The American
Society of Mechanical Engineers. March 25, 1974.
i«3)«S!?nn9ei' F: ?;MB19 D1esels for Power Plants. Power Engineering.
pp. Jo-40. March 1968.
(24) 1973 Year End Summary of the Electric Power Situation in the United
December II 1973 SUrVey Committee' Ed1son Electric Institute.
(25) Federal Power Commission Bureau of Power Staff Report. Proposed
'• 1977- oocket R-362-
9^
(27) Youngblood, S. B. (Acurex) and J. Blasingame (Bechtel). Private
Communication. December 6, 1974.
(28) Op. Clt., Reference 17.
(29) Power, p. S-3. November 1974.
(30) Ibid.
(31) Op. Clt., Reference 20.
(32) Op. Clt., Reference 19.
(33) Op. Clt., Reference 21.
8-113
-------
(34) Op. C1t., Reference 17.
(35) Op. Cit., Reference 2.
(36) Op. C1t., Reference 3.
(37) Op. Cit., Reference 20.
(38) Op. C1t., Reference 19.
(39) Op. Cit., Reference 21.
(40) Sheppard, R. W. (Ingersoll-Rand) and D. R. Goodwin (EPA). Private
Communication. July 28, 1976.
(41) Op. Cit., Reference 16.
(42) Op. Cit., Reference 17.
(43) Annual Pipeline Installation and Equipment Costs. Oil and Gas
Journal. Pipeline Economics Issue. 1960-1977.
(44) 18th Annual Pipeline Installation and Equipment Costs. Oil and Gas
Journal, p. 70. August 18, 1975.
(45) FPC Pipeline Backlog up Over $2 Billion. Oil and Gas Journal, p.
115. September 23, 1974.
(46) Youngblood, S. B. (Acurex) and R. Dubner (Standard Oil, Richmond,
California). December 6, 1974.
(47) Offen, G. R. (Acurex) and J. D. Martin (Union Carbide Corporation,
Port Lavaca, Texas). Private Communication. December 11, 1974.
(48) U.S. Bureau of the Census, U.S. Exports - Schedule B - Commodity by
Country. Report FT410. December 1969-1973.
(49) U.S. Bureau of the Census, U.S. Imports for Consumption and General
Imports, TSUSA Commodity and Country. Report FT246. 1969-1973.
(50) Op. Cit., Reference 4.
(51) Norton, G. (Acurex) and Thermo-King Sales and Services Company. San
Leandro, California. Private Communication. October 16, 1974.
(52) Reyl, G., Deutz Diesel Engines Operating 1n Underground Mines.
Klockener-Humboldt-Dentz AG. Cologne, Germany. Diesel and Gas Turbine
Worldwide Catalog, Vol. 41, 1976.
(53) Youngblood, S. B. (Acurex) and J. Garrity (General Electric).
Private Communication. March 1975.
8-114
-------
(56) Henderson, R. D. (Caterpillar) and R. 0. Selffert
Private Caimun1cat1on. December 19, 1974? M1"«rt
(57) Op. Clt., Reference 55.
St't1on Costs' «' "x1 Sas Journal, p. 89.
o!!'i/ou!l5b'<)0dl Sl B> (teurix) and Mr. Schjffersman (Natural
1976?' 8>
' Se"iUb (Co^er-B»»~^ «••). Private
^^^
["I E?S)?1rSov,J;b^r {9T!'l9^? Co"tt™"1 *tori>. Letter to D. Goodwin
8-115
-------
(54) Youngblood, S. B. (Acurex) and A. Underman (Northwestern Public
Service, Huron, South Dakota). March 24, 1975.
(55) Youngblood, S. B. (Acurex) and A. L. Foltz, Jr. (Enterprise
Engines). Private Communication. January 15, 1975.
(56) Henderson, R. D. (Caterpillar) and R. D. Selffert (EPA/ISB).
Private Communlcation. December 19, 1974.
(57) Op. C1t., Reference 55.
(58) Completed Compressor Station Costs. 011 and Gas Journal, p. 89.
August 23, 1976.
(59) Youngblood, S. B. (Acurex) and Mr. Schaffersman (Natural Gas
Pipeline Company of America, Chicago, Illinois). Private Communication.
October 4, 1977.
(60) Schaub, F. S. (Cooper) and S. B. Youngblood (Acurex). Private
Communication. February 23, 1976.
(61) Newton, C. L. (Colt) and S. B. Youngblood (Acurex). Private
Communication. February 18, 1976.
(62) Fletcher, J. S. (Acurex) and C. Newton (Colt Industries). Private
Communication. November 3, 1975.
(63) Offen, G. R. (Acurex) and F. Schaub (Cooper-Bessemer Co.). Private
Communication. October 13, 1975.
(64) McGowIn, C. R. Stationary Internal Combustion Engines 1n the United
States. Environmental Protection Agency. EPA-R2-73-210. pp. 73-77.
April 1973.
(65) Anderson, H. C., et al. A New Family of Catalysts for Nitric Add
Tall Gases. Engelhard Industries Technical Bulletin. Vol. VII. No.
3:100-105. December 1966.
(66) Shreve, R. N. Nitrogen Industries. In: Chemical Process
Industries. McGraw-Hill, pp. 302-313. 1967.
(67) Requelro, J. F. (Teledyne Continental Motors). Letter to D. Goodwin
(U.S. EPA). November 19, 1974.
(68) Shaw, J. C. Emission Reduction Study on a Carbureted Natural
Gas-Fueled Industrial Engine. Draft ASME Paper. White Superior
Division. White Motor Corporation. November 1974.
(69) Storment, J. 0. and K. J. Springer. Assessment of Control
Techniques for Reducing Emissions from Locomotive Engines. Southwest
Research Institute. AR-884. April 1973.
8-115
-------
(89) Ibid.
(90) U.S. Bureau of Census, U.S. Imports for Consumption and General
Imports, TSUSA Commodity and Country. Report FT246. 1973.
4.uUAS* JfiPartment °f Commerce, U.S. International Transactions:
Fourth Quarter and Year 1976, Survey of Current Business, p. 41. March
Ly// .
(92) Ibid.
8-117
-------
9. RATIONALE
9.1 SELECTION OF SOURCE FOR CONTROL
Previous investigators have concluded that stationary internal
combustion (1C) engines are major contributors to nationwide emis-
(1>2,3)
sions. ' ' In particular, stattonary 1C engines are sources of NO ,
/\
hydrocarbons (HC), particulates, sulfur dioxide (SOJ, and carbon
^
monoxide (CO) emissions. NOX emissions from 1C engines, however, are
of more concern than emissions of these other pollutants for two rea-
sons. First, NOX is the primary pollutant emitted by stationary engines
Second, EPA has assigned a high priority to the development of standards
of performance limiting NOX emissions. Assuming existing levels of
emission controls, national NOX emissions from stationary sources are
projected to increase by more than 40 percent in the 1975-to-1990
period. Applying best technology to all sources would reduce this
increase but would not prevent it from occurring. This unavoidable
increase in NOX emissions is attributable largely to the fact that
current NOX emission control techniques are based on combustion rede-
sign. In addition, few NOX emission control techniques can achieve
large (i.e., in the range of 90 percent) reductions in N0¥ emissions.
A
Consequently, EPA has assigned a high priority to the development of
standards of performance for major NOX emission sources wherever signifi-
cant reductions in NOX can be achieved. Studies have shown that 1C
engines are significant contributors to total U.S. NOX emissions from
stationary sources. Figure 9-1(4) shows that internal combustion
engines account for 16.4 percent of all stationary source NOV emissions,
/\
-------
Industrial Process Combustion 1.6%
Noncombustion 1.7%
Warm Air Furnaces 2.8%
Gas Turbines 3.0%
Fugitive 4.4%
Incineration 0.4%
Utility Boilers
49.0%
Packaged Boilers
20.7%
Figure 9-1. Distribution of stationary NOX emissions for the year 1974
(Reference 4).
9-2
-------
exceeded only by utility and packaged boilers.
An inventory of emissions from installed stationary engines was
computed based on the information presented in summary form in Table 9-
1. As a group, stationary 1C engines (based on 1975 data) currently
account for 3 to 9 percent of the NOX> carbon monoxide (CO), and hydro-
carbons (HC) emitted from all sources, and 9 to 14 percent of those
emitted from stationary sources. This table also shows the percentage
contribution to nationwide totals from installed engines as a function
of their size and the type of fuel they consume. Table 9-2(6) shows
the emission factors used to generate Table 9-1. Annual production
rates are estimated in Table 9-1 to indicate the potential number of
sources that could be affected by New Source Performance Standards
(NSPS).
Table 9-3(7> presents a clearer picture of the relationship between
the number of potentially controllable sources and their contributions
to the nationwide inventory from currently installed units.* This
table also shows that numerous small engines (nearly 13 million units
of 1- to 100-hp) are the most significant contributors of HC and CO
emissions from 1C engines. (Note that nearly 80 percent of the HC
emissions from engines smaller than 350 CID/cyl are methane, a noncri-
teria pollutant.) Therefore, it can be concluded that NOX emissions
constitute the most significant pollutant emitted by stationary engines
since three-quarters of these emissions are emitted by large-bore
(greater than 350 CID/cyl) engines.
Ln^n^MM 'fTl! r?w of 6m1ss1on estimates for engines
than 350-cub1c-1nch displacement per cylinder (CID/cyl). As is
w--wAwa^
9-3
-------
TABLE 9-1. NATIONWIDE EMISSIONS FROM INSTALLED 1C ENGINES
(Percent of Total Emitted 1n U.S. Each Year)
Fuel
Diesel
Natural Gas
Dual-Fuel
Gasoline
hp Range
20 - 100
101 - 500
>500
Subtotal
<500
>500
Subtotal
<15
15 - 99
>100
Subtotal
Total
Annual
Production,
Unitsb
39,000
14,000
3,400
56,400
5,400
600
6,000
Included 1n Diesel
12,600,000C
85,000
10,000
12,600,000
+ 95,000
12,600,000
+ 157,400
Percent All Sources
Percent Stationary Sources
In Mass Units (10s metric tons/yr)
NOX
0.36
0.55
0.48
1.39
1.93
4.16
6.10
0.28
0.16
0.31
0.11
0.58
8.4
13.7
2.0
CO
0.029
0.45
0.016
0.09
0.107
0.229
0.336
0.02
1.84
0.81
0.31
2.96
3.4
11.0
3.6
HCT
0.062
0.095
0.033
0.19
0.81
1.73
2.54
0.11
0.56
0.29
0.10
0.95
3.8
8.8
0.9
aTotal U.S. emission from EPA Nationwide A1r Pollutant Inventory for
bBased on estimates of average hp of engines used 1n each application
^Includes all engines 1n this size category (mobile and stationary).
Listed separately 1n the totals because of the unique nature of
this group
9-4
-------
TABLE 9-2. EMISSIONS FACTORS FOR INVENTORY ON TABLE 1 ,a g/hp-hr
(Reference 6)
Fuel
Gasoline >15 hp
<15 hp
Diesel >500 hpb
<500 hpc
Natural gas
Dual-Fuel
NOX
8,85
5,63
12,9
12,4
11,5
8,2
CO
102
295
1,8
4.47
2,81
2.0
HCT
8.38
20.5
0.43
2.12
4.86
3.1
Emission factors for gasoline and diesel engines are modal
averages; those for natural gas and dual-fuel are for
rated conditions.
bBased on an average of rated condition levels from engines
considered
°We1ghted average of two- and four-stroke engines, Weighting
factors • 2/3 for four-stroke and 1/3 for two-stroke
9-5
-------An error occurred while trying to OCR this image.
-------
Other studies have investigated the emissions of various sta-
tionary sources to aid in establishing a priority for setting standards
of performance. For example, the Research Corporation of New England
determined the effect that standards of performance would have on
nationwide emissions of particulates, NOX, SOX, HC, and CO from sta-
tionary sources.(8) As per EPA-450/3-78-019, "Priorities for New
Source Performance Standards under the Clean Air Act Amendments of
1977," sources were ranked according to the impact, in tons per year,
that standards promulgated in 1980 would have on emissions in 1990.
This ranking placed spark ignition 1C engines second and compression
ignition 1C engines third on a list of 32 stationary N0¥ emission
/\
sources. Consequently, stationary 1C engines have been selected for
development of standards of performance.
In a subsequent study, Argonne National Laboratory used the re-
sults of the TRC study to develop a priority listing for setting NSPS.(9)
In developing this list, source screening factors were used to aid in
establishing these control priorities. Factors considered were:
• Type, cost, and availability of control technology
• Emission measurement methods and applicability
• Enforceability of regulations
• Source location and typical source size
• Energy impact
• Impact on scarce resources
• Other environmental media constraints
The study found that even with the application of maximum NSPS control
efforts, a significant increase of more than 40 percent in NOX emissions
x
9-7
-------
occurs in the 1975-to-1990 period, Furthermore, the study concluded
that the control of Internal combustion engines emissions is a matter
of high priority.
Other factors favoring the control of 1C engines are summarized
briefly below:
t Control techniques for NOX emissions have been shown to be
effective and applicable to installed 1C engines, These
techniques can reduce NOX emissions from 40 to 60 percent on
the average (see Section 4.0).
• No federal, state or local NOX standards exist (with the
exception of Los Angeles and Chicago). Therefore, since
engines are manufactured for a variety of dispersed appli-
cations, a single national standard is preferable.
• 1C engines compete with gas turbines in certain applications.
Since NSPS are currently being developed for gas turbines,
the absence of standards for 1C engines may result in a shift
away from gas turbines to 1C engines. This could cause
greater NOX emissions from both sources than if no standard
were applied to gas turbines, since 1C engines emit NOX at
greater rates than gas turbines.
Furthermore, as shown in Section 3.0, sales of large-bore engines,
primarily for oil and gas exploration, have been substantial during the
past five years, and are anticipated to continue and possibly increase.
Stationary 1C engines, therefore, are significant contributors to total
nationwide emissions of NOX. Consequently, based on all these factors,
stationary 1C engines have been selected for development of standards
of performance.
9-8
-------
9.2 SELECTION OF POLLUTANTS
Oxides Of Nitrogen
Stationary engines emit the following pollutants: NOX, CO, HC,
partlculates, and SOX< As Table 9-3 Indicates, the primary pollutant
emitted by stationary engines 1s NOX, accounting for over six percent
(or 16 percent of all stationary sources) of the total U.S. Inventory
of NOX emissions. This table also Illustrates that large-bore engines
emitted three-fourths of these NOX emissions. It will be shown in
Section 4.0 that the control technology exists to effectively reduce
NOX emissions from large-bore engines. Furthermore, N0¥ emissions are
rt
projected to Increase despite promulgation of all possible New Source
Performance Standards. Therefore, NOX emissions from stationary en-
gines have been selected for control by means of NSPS.
Hydrocarbons and Carbon Monoxide
Table 9-3 also shows that stationary 1C engines emit substantial
quantities of CO and HC as well. Numerous small (1-100 hp) spark Ignition
engines, which are similar to automotive engines, account for about 20
percent of the uncontrolled HC emissions and about 80 percent of the
uncontrolled CO emissions. However, as mentioned in Section 9.3, the
large annual production of these small spark Ignition engines (approxi-
mately 12.7 million) makes enforcement of a new source performance
standard for this group difficult.
An additional factor 1n considering CO and HC control 1s that
Inherent engine characteristics result in a trade-off oetween NO
control and control of CO and HC. A detailed discussion of the trade-
off can be found 1n Section 9.4. In some cases, particularly naturally
9-9
-------
aspirated gas engines, the application of NOX emission control tech-
niques could cause increases in CO and HC emissions. This increase in
CO and HC emissions is strictly a function of the engine operating
position relative to stoichiometric conditions, not the NOX control
technique. Any increase in CO and HC emissions, however, represents an
increase in unburned fuel and hence a loss in efficiency. Since 1C
engine manufacturers compete with one another on the basis of engine
operating costs, which is primarily a function of engine operating
efficiency, the marketplace will effectively ensure that CO and HC
emissions are as low as possible following the application of NOX
control techniques. In addition, promulgation of CO and HC emissions
standards of performance could, in effect, preclude significant NOX
control. CO emissions, which are primarily a function of oxygen avail-
ability and only secondarily of temperature, show a pronounced rise as
the mixture becomes richer than stoichiometric, but little variation as
it becomes leaner. Carbureted engines, however, which are beset by
large variations in cylinder-to-cylinder air-to-fuel ratios, must
operate near the stoichiometric ratio to ensure that no individual
cylinder receives a charge which is too lean to ignite (i.e., exceeds
the lean misfire limit). Consequently, increasing the air-to-fuel
ratio to near stoichiometric to reduce the CO emissions increase has
the effect of also limiting the NOX emissions reduction.
These and other factors discussed in Section 9.3 led to the recom-
mendation of a NOX NSPS for large-bore engines but not for HC and CO
A
emissions since:
9-10
-------
• The 1C engines which emit significant quantities of NO are,
with some exceptions, low emitters of HC and CO
• Many of the NOX reduction techniques discussed In Section 4.4
cause little or no Increase In the already low HC and CO
emissions rates from most large-bore engines
• Individual engines can cause violation of the National Am-
bient Air Quality Standards for HC only under worst-case
atmospheric conditions, and then only very close to the
source (less than 0.3 km)
• No controls for HC used in conjunction with NO, controls have
been demonstrated which reduce the already low nonmethane HC
emissions from large-bore engines
Particulate
No standards of performance are recommended for either particulate
emissions or visible emissions (plume opacity). This recommendation
stems from the following considerations:
• Virtually no data are available on particulate emission rates
from stationary engines because it is so difficult, expensive,
and time-consuming to measure particulates, especially when
done in strict compliance with EPA Method 5 sampling tech-
niques
• It would be very expensive to enforce a standard on required
measurements for particulates in compliance testing which
would be in accordance with EPA Method 5
• It is believed that particulate emission from stationary
engines are relatively unimportant because the plumes from
most of these engines are not now visible
9-11
-------
Sulfur oxides (SO.) emissions are strictly dependent upon the
A
sulfur contained In the fuel, Thus, annual sulfur oxide emissions from
an engine depend on the percent sulfur 1n the fuel and the fuel consump-
tion of the engine during that year, Most engines burn low-sulfur
fuels and will continue to do so since crude and residual fuels must be
treated to remove the salts from the fuels, and Inhibitors must be added
to prevent the vanadium 1n the fuel from corroding 1C engine components.
Treatment facilities exist, but their function 1s minimal. The primary
reason for the shift away from the treatment of crude and residual
fuels 1s one of economics. In today's market, It simply costs more to
buy and treat the crude and residual oils than to purchase and burn the
distillate oils.
The cost of flue gas desulfurlzatlon for 1C engines does not appear
to be reasonable from an economic viewpoint. Therefore, the only viable
means of controlling SO, emissions would be combustion of low-sulfur
fuels. If users 1n urban or S0x-sens1t1ve areas decide to buy new en-
gines and to use crude or residual oil as a fuel with these engines, then
the local air pollution authorities could Impose fuel restrictions on
these engines. Such fuel restrictions would be entirely Independent of
the standards of performance from both a technological and enforcement
viewpoint. That 1s, the absence of federal emission limits on SO, would
not prevent a local air pollution control district from setting such a
standard since the engine would not have to be changed In order to meet
the local standard. Thus, standards of performance are not recommended
for SO emissions.
A
9-12
-------
9.3 SELECTION OF AFFECTED FACILITIES
In sections 1.0 and 2.0 It was shown that NOX emissions constitute
the most significant pollutant emitted by stationary 1C engines, and that
large-bore (greater than 350 ciD/cyl) engines account for over 75 percent
of all NOX emission from stationary engines. This section will establish
criteria that define which large dlesel, dual-fuel, and natural gas
engines (referred to as "affected facilities") are to be affected by the
proposed standards of performance. The objective here 1s to apply stan-
dards of performance to significant sourer of NOX emissions.
Thus, the following sections will present and explain the criteria
that define affected facilities after considering the applications
served by stationary engines, the number of units produced annually, and
the Incremental NOX contributed by the annual production. The following
discussions are subdivided by the three operational fuel types: dlesel.
dual-fuel, and natural gas. As will be discussed in the following para-
graphs, this classification separates large-bore engines Into three
relatively distinct categories of engine applications. Initially, large-
bore engines will be defined as those exceeding 350 CID/cyl. Then, if
necessary, other criteria will be presented and explained to define
affected dlesel, dual-fuel, and natural gas engines.
The following discussion summarized an extensive study of the
applications of large-bore engines. Many of the conclusions presented
here are based on information concerning engine sales and applications
during the past five years. This information was voluntarily submitted
by engine manufacturers in response to Section 114 requests for informa-
9-13
-------
This information has not been cited for particular manufac-
turers since it is considered proprietary by the manufacturers.
Affected Diesel Engines
The primary high usage (large emissions impact), domestic appli-
cation of large-bore (i.e., greater than 350 CID/cyl) diesel engines
during the past five years has been for oil and gas exploration and
production. These and other applications are illustrated in Figure 9-
2>(12-17) As this figure shows, the market for prime (continuous)
electric generation and other industrial applications all but disappeared
after the 1973 oil embargo, but was quickly replaced by sales of standby
electric units for building services, utilities, and nuclear power sta-
tions. The rapid growth in the oil and gas production market occurred
because diesel units are being used on oil drilling rigs of various
sizes. Sales of engines to export applications have also grown steadily
since 1972, and are not a major segment of the entire sales market.
Medium-bore (from 35 to 350 CID/cyl) as well as large-bore engines
are sold to oil and gas exploration, standby service, and other indus-
trial applications. Furthermore, manufacturers of medium- and large-bore
engines often compete for the same applications, although, in general,
medium-bore engines have a cost advantage (lower $/hp). This is because
the higher initial costs for a large-bore, heavy-duty, continuous-service
engine more than offset their lower maintenance and fuel costs. This
overlap in sizes is best illustrated in Figure 9-3 which shows a consi-
derable number of medium- and large-bore engines in the 500- to 2000-
horsepower range. Figure 9-4 shows the displacement per cylinder that
9-14
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
corresponds to the ranges of horsepower offered by the manufacturers
shown 1n Figure 9-3. Table 9-4 shows the overlap for particular engine
models.
The application with the greatest degree of overlap for medium and
large-bore dlesels 1s petroleum exploration. Smaller (250-to 1000-hp)
medium-bore designs (e.g., Detroit Diesel, Cummins, and Caterpillar) are
used on portable drilling rigs to drill or service 2500- to 5000-foot
wells. These rigs are trailer-mounted or helicopter-transported; there-
fore, small, lightweight (approximately 4000-lb) engines are favored. In
addition, multiple units are preferred to Insure some backup power 1n the
event one engine 1s down, ruling out a single unit of comparable total
horsepower.
Larger horsepower engines are used 1n groups of three to five to
provide 800 to 3000 hp for wells ranging 1n depth from 5000 to 25,000
feet. On most of these rigs, engines supply mechanical power to operate
the drilling (rotary table), mud pumps, and hoisting equipment. In the
larger units several engines from one manufacturer's engines operate
pumps or generator sets for auxiliary power.(18) A relatively new
approach is to generate AC power, rectify some of it for drilling power
(variable load DC motors), and use the rest to drive AC auxiliaries.
This approach 1s used primarily on offshore platforms, although there is
interest 1n applying it to land-based sites despite its higher cost.
In conclusion, then, larger land-based drilling sites are the major
areas of overlap of service provided by both large-bore and medium-bore
manufacturers. These applications and baseload electric generation (to a
lesser extent, since horsepower sales are small) have the most signifi-
9-18
-------An error occurred while trying to OCR this image.
-------
cant NO emissions impact because they are high usage (approximately 6000
A
hr/yr). However, a greater-than-350-CID/cyl definition of affected
facilities would result in some manufacturers (e.g., Waukesha) being
subject to control technology development costs, while medium-bore en-
gines (of same power, but more cylinders) serving Identical applications
would not incur these costs. This 1s clearly undesirable since this
definition would unfairly place some large-bore engines in a less com-
petitive position than similarly sized (by horsepower), smaller-bore
designs.
On the other hand, applying the standards of performance to medium-
bore engines serving the same applications as large-bore designs would
Increase the number of affected facilities from about 200 to about 2000
units per year (based on 1976 sales information) but consequently further
reduce NO emissions. Medium-bore sales accounted for significant NOX
A
emissions in 1976 (approximately 12,500 megagrams). It 1s estimated that
approximately 25 percent, or about 500 of these units, in high usage
applications accounted for most of the medium-bore NOX emissions, since
most of the remainder of these units were sold as standby generator sets.
Though the potential achievable NOX reduction 1s significant, considering
this large number, and the remoteness and mobility of petroleum applica-
tions, this alternative would create serious enforcement difficulties.
Additionally, this alternative causes the standard to apply to lower
power engine models with fewer number of cylinders competing in different
stationary markets with other unregulated engines. Consequently, a
definition is required that distinguishes between large-bore engines that
compete with medium-bore high power engines used for baseload electrical
9-20
-------
generation from large-bore engines that compete solely with other large-
bore engines.
One approach would be to define dlesel engines covered by standards
of performance as those exceeding 560 CID/cyl. This alternative would
exclude engines presently manufactured by Waukesha as well as those
produced by Caterpillar, Detroit Diesel, and Cummins. This definition,
however, shifts the area of overlap 1n horsepower between regulated and
unregulated engines to other large-bore dlesel manufacturers. This
situation 1s depicted 1n Figure 9-5, which Illustrates the relationship
between displacement per cylinder and rated (continuous) horsepower. All
Waukesha engines are excluded above the 560 CID/cyl limit. However,
Superior's dlesel engines ranging In size from 596- to 825-CIO/cyl would
be subject to standards. These engines compete 1n very few cases with
«au*esha dlesel engines. Raising the Hm1t to 700-CID/cyl would exclude
Superior engines 1n the 500- to 100-hp range, but It would also exclude
EMD and Alco models, which compete with Colt (700-CID/cyl, hence regula-
ted) 1n the 1000- to 3000-hp range. Establishing a 560-CID/cyl defini-
tion, therefore, appears to be a viable method of excluding engines which
compete with medium-bore designs without Introducing a significant over-
lap problem at a different power level.
After considering the sizes and displacements offered by each
dlesel manufacturer and the applications served by dlesel engines, a 560-
CID/cyl definition was selected as a reasonable approach for separating
large-bore engines that compete with medium-bore engines from large-bore
engines that compete solely with each other. This cylinder displacement
size was chosen because engines below this size are generally used for
9-21
-------An error occurred while trying to OCR this image.
-------
different applications than those above 1t. Therefore, 1t 1s recommended
that dlesel engines greater than 560 CID/cyl be affected by standards of
performance.
Affected Dual-Fuel Engines
The concept of dual-fuel operation was developed to take advantage
of both compression Ignition performance and Inexpensive natural gas,
These engines have been used almost exclusively for prime electric power
generation, Figure 9-6^19"22) Illustrates, however, that shortages of
natural gas and the 1973 oil embargo have combined to significantly
reduce the sales of these engines In recent years, The few large-bore
units that were sold (11 1n 1976) were all greater than 350 CID/cyl. In
fact, with the exception of Superior D1v1s1on/Cooper and Stewart-Steven-
son (modified Detroit Diesel engine) products, all were greater than 500
horsepower and 1000 CID/cyl as shown 1n Figures 9-7 and 9-8. Moreover,
nearly all of the dual-fuel engines sold since 1972 have been larger than
1000 hp, Only Stewart-Stevenson manufactures dual-fuel engines less than
560 CID/cyl, Sales of these units are less than 100 units per year and
about 70 percent of these are exported,^3)
Although a greater-than-350-CID/cyl limit would subject nearly all
new dual-fuel sources to standards of performance (only engines manu-
factured by Stewart-Stevenson would be excluded), 1t 1s recommended that
the definition chosen to define affected dlesel engines (560 CID/cyl)
also be applied to dual-fuel engines. The reason 1s that supplies of
natural gas are likely to become even more scarce, possibly causing
recently installed or future dual-fuel units to convert to diesel fuel
operation. Any additional diesel engines that would be created by con-
9-23
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
version from dual-fuel operation should be subject to the same regula-
tions applicable to other large diesel engines.
Affected Gas Engines
The primary application of large (greater than 350 CID/cyl) gas
engines during the past five years has been for oil and gas production.
The primary uses are to power gas compressors for recovery, gathering,
and distribution. Figure 9-9(24~29\ based on manufacturer's data from
response to the June 16, 1976 Section 114 requests for information,
illustrates that 75 to 80 percent of all gas engine horsepower sold
during the past five years was used for these applications.
During this time sales to pipeline transmission applications de-
clined. Pipeline applications combined with standby power, electric
generation, and other services (industrial and sewage pumping). These
other applications accounted for the remaining 20 to 25 percent of
horsepower sales. The growth of oil and gas production applications
during this period corresponds to the increasing efforts to find new, or
recover marginal, gas reserves and distribute them to the existing
pipeline transmission network.
Figure 9-10 illustrates the number of gas engines sold for five size
groups during the past five years. The large number of smaller-than-500-
hp engines that were sold during this period are one or two cylinder
engines used on oil well beam pumps and for natural gas well recovery and
gathering. Most of the other larger gas engines that were sold during
this period ranged from 500 to 2000 hp. In 1976, approximately 400
engines in this size range were sold, primarily for oil and gas production
(see Figure 9-9). Most of these gas engines were manufactured by Cater-
9-27
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
pillar, Cooper, Waukesha, and White Superior.
With the exception of standby service, all the applications of
Figure 9-9 are high usage (approximately 6000 hr/yr), and therefore,
contribute significant NOX emissions. It 1s estimated that the 400,000
horsepower of large-bore gas-engine capacity sold for oil and gas pro-
duction applications 1n 1976 emitted 34,900 megagrams of NOX emissions,
or nearly three times more NO, than was emitted by the 200,000 horsepower
of large-bore dlesel engine capacity (greater than 350 CID/cyl) sold for
the same application In that year (see Section 3.1). Thus, large-bore
gas engines are primary contributors of NOX emissions from new stationary
1C engines, and standards of performance should be directed particularly
at these sources.
If affected engines were defined as those greater than 350 CID/cyl,
then all manufacturers of gas engines greater than 500 hp, except Cater-
pillar, would be affected by proposed standards of performance. However,
large Caterpillar gas engines range from 225 to 930 horsepower, and
therefore, compete with the other large-bore manufacturers (particularly
Waukesha). Figures 9-11 and 9-12 show more clearly the overlap 1n horse-
power provided by manufacturers of engines of various cylinder displace-
ments. Therefore, a greater-than-350-CID/cyl limit would give one manu-
facturer an unfair competitive advantage over other large-bore engine
manufacturers. Thus, although a greater-than-350-CID/cyl limit would
subject most significant gas engine sources of NO, emissions to potential
standards of performance, this definition should be revised based on the
>
following considerations:
9-30
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
• The greater-than-350-CID/cyl definition excludes the only other
manufacturer (Caterpillar) of gas engines greater than 500 hp.
Caterpillar gas engines compete directly with the large gas
engines manufactured by Cooper, Waukesha, and White Superior,
which would be regulated.
t No emissions have been measured or control techniques demon-
strated for 1- and 2-cylinder engines which would be included
in potential standards of performance by the existing greater
than 350 CID/cyl limit.
The first observation suggests that the definition should be lowered,
or another definition adopted, to include the large Caterpillar engines
that compete in identical applications with Cooper, Waukesha, and White
Superior units. Although Caterpillar has not reported controlled emissions
data for their gas engines, control techniques have been demonstrated on
other similar gas engines and should be effective when applied to Cater-
pillar engines, since they are all similar in design (i.e., carbureted
and gas injected engines that are either turbocharged and aftercooled or
naturally aspirated).
Table 9-5 compares large Caterpillar gas engines with Waukesha
models that are greater than 350 CID/cyl. As this comparison illus-
trates, Caterpillar engines with smaller displacements per cylinder and
greater numbers of cylinders serve about the same power range as do the
larger Waukesha engines. On the basis of this table, either of the
following two steps would subject Caterpillar gas engines to potential
standards of performance:
9-33
-------An error occurred while trying to OCR this image.
-------
• Select a definition of greater than 240 CID/cyl
• Define affected gas engines as those greater than 350 CID/cvl
or greater than or equal to 8 cylinders and greater than 240
CID/cyl
Both measures would essentially include only Caterpillar engines
in the same power range as Waukesha. The second definition has a
slight advantage over the first since 1t includes only Caterpillar
engines that have Waukesha counterparts of about the same power (note
that the greater-than-240-CID/cyl definition alone would include the
Caterpillar G353, which has no large Waukesha counterpart). Therefore,
the greater-than-350-CID/cvl or greater-than-or-equal-to-8 c.yl inders
and greater-than-240-CID/cvl definition of affected gas engines is
recommended.
With regard to one and two cylinder engines, it is recommended
that they be excluded from potential standards of performance. This
suggestion can be supported considering:
• At present these engines account for less than 10 percent of
all gas engine horsepower and, therefore, are less signifi-
cant NOX emitters than the larger gas engines used for oil
and gas production
• These sources are numerous and widely dispersed in remote lo-
cations
• These engines are low rated* and therefore, probably have
lower NOX emissions than the larger higher-rated gas engines
In addition to these factors, consideration should be given to the
*0perate at a small fraction of their potential power output.
9-35
-------
undeveloped control technology for these engines. A spokesman for one
manufacturer noted that they are only currently preparing to measure
NO emissions from their one- and two-cylinder engines. Therefore, it
" .
is recommended that all one- and two-cylinder gas engines be exempted
from potential standards of performance.
In summary, then, it is recommended that the following criteria
define gas engines that are to be affected by standards of performance:
• Affected facilities are defined as engines that are either
greater than 350-CID/cyl or greater than 8-cylinder and
greater than 240-CID/cyl
t All one or two cylinder gas engines are exempt from standards
of performance
9.4 SELECTION OF BEST SYSTEM OF EMISSION REDUCTION
As discussed in Chapter 4, four emission control techniques, or
combinations of these techniques, have been identified as demonstrated
NO emission reduction systems for stationary large-bore internal
/\
combustion engines. These techniques are: (1) retarded ignition or
fuel injection, (2) air-to-fuel ratio changes, (3) manifold air cooling,
and (4) derating power output (at constant speed). In general, all
four techniques are applied by changing an engine operating adjustment.
Manifold air cooling, however, may require a larger heat exchanger, and
air-to-fuel changes may require turbocharger resizing.
These control techniques, described in Chapter 6, reduce NOX
emissions primarily by lowering peak flame temperatures. Some of the
techniques may result in increased fuel consumption and/or engine main-
tenance.
9-36
-------
Fuel Injection retard 1s the most effective NOX control technique
for dlesel-fueled engines, achieving maximum NOX reductions of about 65
percent. Similarly, air-to-fuel ratio change 1s the most effective NO
rt
control technique for natural gas engines, achieving maximum NOX
reductions of about 80 percent. Both retard and air-to-fuel ratio changes
are effective in reducing NOX emissions from dual-fuel engines, achieving
maximum NOX reductions of about 70 percent.
Other NOX emission control techniques exist but are not considered
feasible alternatives. These techniques, also described 1n Chapter 6,
Include exhaust gas ^circulation (EGR), combustion chamber modification
(CCM), water Induction, and catalytic reduction.
Exhaust gas redrculatlon tests have shown effective NO reduc-
n
tlonsj however, the necessity for cooling the redrculated gas may lead
to contamination of flow passages 1n the cooling heat exhanger as well
as 1n the engine turbocharger and aftercooler. At present there Is
Insufficient data on which to base conclusions and more development 1s
required. Therefore, 1t 1s not considered a demonstrated emission
control technique.
Data from smaller-bore dlesel engines Indicate that combustion
chamber configuration has a significant effect on NOV emissions.
A
However, none of the domestic large-bore engine manufacturers has
thoroughly studied the effects of modified combustion chamber geometries
on NOX emissions. Manufacturers have estimated that an extensive
development program of three to five years would be required to establish
the emission benefits of such a major engine redesign. The majority of
the existing engines are primarily long-established designs that have
9-37
-------
been refined over the years to Improve fuel economy and maintenance.
Since there 1s Insufficient data to draw conclusions, combustion cham-
ber modification 1s not considered a demonstrated emission control
technique.
The effect of water Induction 1n large stationary Internal combus-
tion engines 1s similar to the effect 1n gas turbines. Significant NOX
reductions are achieved due to the quenching effect of the presence of
water. However, as discussed 1n Section 4.4.7, tests with water Induc-
tion 1n large stationary Internal combustion engines have shown unaccept-
able deposit bund-up on the exhaust/Intake valves from the use of
untreated water, and severe lubricating oil contamination. Therefore,
water Induction 1s also not considered a demonstrated control technique.
Catalytic reduction of NOX 1n large stationary Internal combustion
engines 1s difficult to achieve and the capital cost could be high.
Most large stationary Internal combustion engines operate at air-to-
fuel (I.e., mass flowrate [g/hr] of air Into an engine divided by the
mass flowrate of fuel [g/hr]) ratios that are typically much greater
than sto1ch1ometr1c and consequently the engine exhaust 1s character-
ized by high oxygen concentrations. Existing automotive catalytic
converters, however, operate near sto1ch1ometr1c conditions (I.e., low
exhaust oxygen concentrations). These automobile catalysts are not
effective 1n reducing NO 1n the presence of high oxygen concentrations.
A
Consequently, entirely different catalyst systems would be needed to
reduce NO emissions from large stationary Internal combustion engines.
/\
Although such catalyst systems are currently under development and have
been demonstrated for one very limited application (i.e., fuel-rich
9-:
-------
naturally aspirated gas engines), they have not been demonstrated for
the broad range of 1C engines manufactured, such as turbocharged en-
gines, fuel-lean gas engines, or diesel engines. For these engines the
reduction of NOX by ammonia injection over a precious metal (e.g.,
Platinum) catalyst appears promising with NO, reductions of approximately
90 percent having been reported; however, the cost of such a system is
high.
For a typical 1000 horsepower engine approximately 2 cubic feet of
honeycomb catalyst (platinum based) would be required to ensure proper
operation of the system. The cost of the catalyst was estimated at
$1500/cubic foot (in 1973). Assuming that the engine costs $150/hp and
that the cost of the catalyst accounts for about one-half the cost of
the whole system (container, substrate, and catalyst), the capital
investment for this control system represents approximately four percent
of the engine purchase price.
The amount of ammonia required for an ammonia/catalyst NOX reduc-
tion system will depend on the NOX emission rate (g/hp-hr). Ba'sed on
uncontrolled NOX emission rates of 9 to 22 g/hp-hr, and the cost of
$150/ton for the ammonia, the cost impact of injecting ammonia is
approximately 5 to 15 percent of the total annual operating costs
($/hp-nr) for natural gas engines. When this operating cost is com-
bined with the capital cost of the catalytic system discussed above,
the total cost increase is about 25 percent. Therefore, in continuous
service applications this system is expensive compared to control
techniques such as retard or air-to-fuel ratio changes.
9-39
-------
It 1s also Important to note that the consumption of ammonia can
be expressed as a quantity of fuel since natural gas 1s generally used
to produce ammonia. Assuming a conservative NOX emission rate of 20
g/hp-hr, and engine heat rate of 7500 Btu/hp-hr, a heating value of
21,800 Btu/lb for natural gas, and a requirement for approximately 900
Ibs of gas per ton of ammonia produced, then the ammonia necessary for
the catalytic reduction has the same effect on the supply of natural
gas as a 2-percent Increase 1n fuel consumption. Additional fuel 1s
required to operate the plant which produces the ammonia.
Catalytic reduction, therefore, is currently not a demonstrated
NO emission control technique which could be used by all 1C engines.
A
Consequently, although catalytic reduction of NOX emissions could be
used in a few isolated cases to comply with standards of performance,
1t could not be used as the basis for developing standards of perfor-
mance which are applicable to all 1C engines.
In summary, four emission control techniques have been identified
as demonstrated NOX emission reduction systems for stationary large-
bore internal combustion engines. These techniques are: (1) retarded
ignition or fuel injection, (2) air-to-fuel ratio changes, (3) manifold
air cooling, and (4) derating power output (at constant speed). Fuel
injection retard is the most effective NOX control technique for diesel-
fuel engines and air-to-fuel ratio change is the most effective NOX
control technique for gas engines. Either technique is effective for
dual-fuel engines.
The data and information presented in Chapters 4 and 6 clearly
indicate that application of the control techniques mentioned above
9-40
-------
will reduce NOX emissions from Internal combustion engines. It Is not
immediately clear, however, from this data and Information whether the
application of these emission control techniques to all 1C engines
would reduce NOX emissions to a specific level due to Inherent differ-
ences 1n the uncontrolled emission characteristics of various engines.
In general, engines with high uncontrolled NOX emission levels have
relatively high controlled NOX emission levels and engines with low
uncontrolled NOX emission levels have relatively low controlled NO,
emission levels. To eliminate these Inherent differences in NO,
emission characteristics among various engines, the data were analyzed
In terms of the degree of reduction in NO, emissions as a function of
the degree of application of each emission control technique. Figures
4-27 and 4-31, reproduced here as Figures 9-13 and 9-14, Illustrate the
overall effectiveness of ignition retard and air-to-fuel ratio changes
for reducing NO, emissions. For example, in Figure 9-13, the applica-
tion of approximately eight of Ignition retard results in about 40
percent reduction of NOX emissions. Thus, the data presented in Chapters
4 and 6 demonstrate that the same degree of application of each of
these four NOX emission control techniques - I.e., (1) retarded ignition
or fuel Injection, (2) air-to-fuel ratio changes, (3) manifold air
cooling, and (4) derating power output (at constant speed) - will
result in essentially the same degree of reduction of NOX emissions
from all large stationary Internal combustion engines. Consequently,
the ability to achieve certain percentay, reductions in N0¥ emission
A
levels Is clearly demonstrated.
As can be seen from Figures 9-13 and 9-14, those Included In
9-41
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
Chapters 4 and 6, a wide range of regulatory options (I.e., basis for
standards) 1s available. The greatest reductions 1n NOX emissions,
achieved with some degree of consistency by the use of these four
demonstrated NO control techniques, was approximately 60 percent.
n
Therefore, 60 percent reduction was Initially selected 1n Chapter 6 as
the most stringent regulatory option that could serve as the basis for
the standards. Alternative regulatory options of 20 and 40 percent
reduction were selected as representative of less stringent basis for
standards.
Subsequent review and analysis, however, Indicate that technical/
hardware considerations limit the percentage NOX reduction that can be
achieved 1n practice. Generally Ignition retard 1n excess of eight
degrees 1n dlesel-fueled engines frequently leads to unacceptably high
exhaust temperatures, resulting 1n exhaust valve and/or turbocharger
turbine damage. Similarly, changes 1n the air-to-fuel ratio 1n excess
of five percent 1n gas engines frequently leads to excessive misfiring
or detonation which could lead to a serious explosion 1n the exhaust
manifold. As shown 1n Figures 9-13 and 9-14, eight degrees of Ignition
retard 1n dlesel-fueled engines and five percent change 1n air-to-fuel
ratios 1n gas-fueled engines yield about a 40 percent reduction 1n NOX
emissions. Consequently, 1n light of these limitations to the applica-
tion of these emission control techniques, it is apparent that a 40-
percent reduction 1n NOX emissions is the most stringent regulatory
option which could be selected as the basis for standards of performance,
An alternative of 20 percent NOX emission reductions was also considered
a viable regulatory option which could serve as the basis for standards.
9-44
-------
Environmental Impacts
Standards of performance based on alternative I (20-percent reduc-
tion) would reduce national NOX emissions of about 14.6 million mega-
grams per year for all stationary sources by 72,500 megagrams annually
In the fifth year after the standard went Into effect. In contrast,
standards of performance based on alternative II (40-percent reduction)
would reduce national NOX emissions by about 145,000 megagrams annually
In the fifth year after the standard went Into effect. Thus, standards
of performance based on alternative II would have a much greater Impact
on national NOX emissions than standards based on alternative I.
As discussed In Chapter 7, ambient air quality dispersion modeling,
based on "worst case" conditions, indicates uncontrolled ambient air
NOX levels near large stationary internal combustion engines can vary
from approximately 60 percent of the National Ambient Air Quality
Standard of 100 yg/m3 to over twice the standard depending on the size
of the engine. One calculation for a large gas engine yielded an
uncontrolled ambient NOX level of about 220 wg/m3.
These maximum concentrations, however, are located at distances ex-
tremely close to the source (0.3 km) because of the aerodynamic effects
on plume rise as well as the relatively low height of the exhaust stack
discharge. For example, it is estimated that at 1.0 km from the source,
those concentrations would be only 15 percent of the above cited levels,
well below the National Ambient Air Quality Standard.
In any event, standards of performance based on alternative I
would reduce the highest calculated ambient air concentration of 220
9-45
-------
wg/m3 to about 180 ug/ni3, while standards based on alternative II would
reduce this ambient NOX concentration level to about 100 ug/m3. Thus,
where ambient air NOX concentrations near large stationary internal
combustion engines would be significant, standards of performance based
on alternative II would be more effective in reducing ambient air NOX
levels than standards of performance based on alternative I.
Standards of performance based on either alternative would, with
the exception of naturally aspirated gas engines, not significantly
effect carbon monoxide (CO) or hydrocarbon (HC) emissions from most
engines. A typical diesel engine with a sales-weighted average uncon-
trolled CO emission level of approximately 2.9 g/hp-hr would experience
an increase in CO emissions of about 0.75 g/hp-hr (26 percent) to
comply with standards of performance based on alternative I, and an
increase of about 1.5 g/hp-hr (52 percent) to comply with standards of
performance based on alternative II. Total hydrocarbon emissions would
increase a sales-weighted average uncontrolled emission level of 0.3
g/hp-hr by about 0.06 g/hp-hr (20 percent) to comply with standards
based on alternative I, and would increase by about 0.1 g/hp-hr (33
percent) to comply with standards of performance based on alternative
II.
Similarly, a typical dual-fuel engine with a sales-weighted average
uncontrolled CO emission level of approximately 2.7 g/hp-hr would
experience an increase in CO emissions of about 1.2 g/hp-hr (44 percent)
and about 2.7 g/hp-hr (100 percent) to comply with standards of perfor-
mance based on alternatives I and II, respectively. Total hydrocarbon
9-46
-------
emissions, however, would decrease by about 0.3 g/hp-hr (11 percent)
from a sales-weighted average uncontrolled level of a approxlmatley 2.8
g/hp-hr to comply with standards of performance based on alternative I.
To comply with standards of performance based on alternative II total
hydrocarbon emissions would decrease 0.6 g/hp-hr (21 percent).
A typical turbocharged or blower scavenged gas engine with a
sales-weighted average uncontrolled CO emission level of approximately
7.7 g/hp-hr would experience an increase 1n CO emissions of about 1.9
g/hp-hr (25 percent) to comply with standards of performance based on
alternative I and about 3.8 g/hp-hr (49 percent) to comply with stan-
dards of performance based on alternative II. Total hydrocarbon emis-
sions would increase a sales-weighted average uncontrolled level of
approximately 1.8 g/hp-hr by about 0.2 g/hp-hr (11 percent) to comply
with standards of performance based on alternative I. To comply with
standards of performance based on alternative II total hydrocarbon
emissions would increase by about 0.4 g/hp-hr (22) percent.
A typical naturally aspirated gas engine with a sales-weighted
average uncontrolled CO emission level of approximately 7.7 g/hp-hr
would experience an increase in CO emissions of about 3.9 g/hp-hr (51
percent) to comply with standards of performance based on alternative I
and about 17 g/hp-hr (220 percent) to comply with standards of perfor-
mance based on alternative II. Total hydrocarbon emissions would
increase a sales-weighted average uncontrolled level of approximately
1.8 g/hp-hr by about 0.04 g/hp-hr (2 percent) to comply with standards
of performance based on alternative I. To comply with standards of
performance based on alternative II total hydrocarbon emissions would
9-47
-------
Increase by about 0.08 g/hp-hr (4 percent).
The increase in ambient air CO levels due to compliance with NOX
standards of performance based on either alternative would be small.
For most engines,the application of standards of performance based on
Alternative I would increase the maximum 8-hr ambient air CO concentra-
tion from about 0.23 mg/m3 to 0.29 mg/m3 for the typical diesel engine
mentioned above. The application of standards of performance based on
alternative II would increase the maximum 8-hour ambient air CO concen-
tration to 0.35 mg/m3. These values are insignificant compared to the
National Ambient Air Quality Standard of 10 mg/m3 for CO.
The application of standards of performance based on alternative I
would increase the maximum 8-hr ambient air CO concentration from an
3 3
uncontrolled concentration of 0.65 mg/m to 0.94 mg/m for the typical
dual-fuel engine mentioned above. The application of standards of
performance based on alternative II would increase the maximum 8-hr
3
ambient air CO concentration to 1.3 mg/m .
For a typical turbocharged or blower scavenged gas engine, the
increase in the maximum 8-hr ambient air CO concentration would be even
less. The application of standards of performance based on alternative
I would increase the maximum 8-hr ambient air CO concentration from an
o o
uncontrolled concentration of 0.15 mg/m to 0.19 mg/m . The applica-
tion of standards of performance based on alternative II would increase
the maximum 8-hr ambient air CO concentration to 0.22 mg/m . These
values are also insignificant compared to the National Ambient Air
Quality Standard.
9-48
-------
For a typical naturally aspirated gas engine, the application of
standards of performance based on alternative I would Increase the
maximum 8-hr ambient air CO concentration from an uncontrolled concen-
tration 0.77 mg/m3 to 1.16 mg/m3. The application of standards of
performance based on alternative II would Increase the maximum 8-hr
ambient air CO concentration to approximately 2.0 mg/m3.
Based on Industry growth projections, an Increase In national CO
emissions of about 63,000 megagrams annually would be realized in the
fifth year after the standards go Into effect as a result of standards
of performance based on alternative I. As a result of standards of
performance based on alternative II an Increase of about 216,000 mega-
grams annually would be realized 1n the fifth year after standards go
Into effect.
The large Increase 1n CO emissions Is due primarily to carbureted
or naturally aspirated gas engines. These engines operate closer to
Sto1ch1ometr1c conditions under which a small change 1n the air-to-fuel
ratio results In a large Increase in CO emissions. As shown 1n Figure
9-15, any significant NOX reduction 1s accompanied by a significant
Increase in CO.
As discussed earlier, though the total national CO emissions would
increase significantly, ambient air CO concentrations in the immediate
vicinity of these carbureted or naturally aspirated gas engines would
not be adversely affected. As a result of the standards of performance
based on alternative II, the maximum 1 hour ground level concentration
from a typical engine would increase to about 2 mg/m3 compared to the
National Ambient Air Quality Standard of 10 mg/m3. Ambient air NO
9-49
-------An error occurred while trying to OCR this image.
-------
concentrations from the same engine, however, would decrease concurrently
about 40 percent, to a level less than half of the National Ambient A1r
Quality Standard of 100 ng/m3.
Thus, there exists a trade-off between NOX emissions reduction and
CO emissions Increase, particularly for carbureted or naturally aspirated
gas engines. EPA recognizes this trade-off and 1s concerned about It's
attractiveness. It should be noted though that CO emissions are a local
problem since they rapidly oxidize to C02. Additionally, most naturally
aspirated gas engines are operated in remote locations where CO is not a
problem. NOX emissions, however, are linked to the formation of photo-
chemical oxidants and are subject to long range transport. N0¥ emissions
rt
reductions are also much harder to achieve than CO or HC emissions
reductions which may be achieved more easily from other sources.
One alternative is to propose a CO emissions limit based on the
use of oxidizing catalysts. These catalysts can provide CO and HC
emissions reductions on the order of 90 percent. Initial capital costs
are high, however, averaging about $7500 for a typical 1000 horsepower
naturally aspirated gas engine or about 15 percent of the purchase
price of the engine.
EPA feels these costs for control of CO emissions are unreasonable.
The trade-off between NOX and CO emissions, however, seems reasonable.
Therefore, CO was not selected for control by standards of performance.
Hydrocarbon emissions are also currently considered a pollutant of
concern due to the impact on ambient air oxidant concentrations.
However, although relationships are being developed to predict the HC
emissions impact on oxidant concentrations, these concentrations depend
9-51
-------
on the ambient air NO/HC ratio, Thus, 1t 1s difficult to estimate the
n
Impact of Increased HC emissions on ambient air oxldant concentrations.
Furthermore, based on data 1n Appendix C.4, 1t 1s estimated that more
than 90 percent of the total hydrocarbon emissions from gas engines and
75 percent of the total hydrocarbons from dual-fuel engines are methane,
which 1s nonreactlve and does not lead to oxldant formation. Finally,
the Increase 1n national total HC emissions based on either alternative,
for most engines, 1s very small.
Standards of performance based on alternative I would Increase
national total HC emissions of 10.2 million megagrams by about 2,300
megagrams annually 1n the fifth year after the standards went Into
effect compared to an Increase of about 4,600 megagrams annually associ-
ated with alternative II. Standards of performance based on alternative
I would Increase national reactive HC emissions by approximately 108
megagrams annually in the fifth year after the standards went into
effect, compared to an increase of approximately 216 megagrams annually
associated with alternative II.
As described in Chapter 3, stationary internal combustion engines
are sources of NO , HC and CO emissions, with both NOX and HC contri-
/\
buting to oxldant formation. With regard to regulation of emission
from 1C engines, NOX emissions are of more concern than emissions of
hydrocarbons for two reasons. First, NOX is emitted in greater quantities
from stationary internal combustion engines than hydrocarbons. Second,
a high priority has been assigned to development of standards of per-
formance limiting N0¥ emissions from stationary sources are projected
X
to increase by more than 40 percent in the 1975-to-1990 period. Apply-
9-52
-------
A
Ing best technology to all new sources would reduce this increase but
would not prevent it from occurring. This unavoidable increase in NO
emissions is attributable largely to the fact that current NO emission
A
control techniques are based on combustion redesign. In addition, few
NOX emission control techniques can achieve large (I.e., in the range
of 90 percent) reductions in NOX emissions. In contrast, HC emissions
are much easier to reduce. Large reductions from numerous sources are
achievable with the installation of existing add-on control equipment.
Consequently, EPA has assigned a high priority to the development of
standards of performance for major NOX emission sources wherever signi-
ficant reductions in MOX can be achieved.
The slight increase in HC emissions from 1C engines associated
with control of NOX from 1C engines can be offset from other sources
easier than NOX emissions can be reduced from other sources. There-
fore, the adverse environmental impact of increased HC emissions due to
the reduction in NOX emissions is considered small.
There would be essentially no water pollution impact of standards
of performance based on either alternative I or alternative II. Only
one control technique, increased manifold air cooling, could result in
an additional discharge of water. However, most newly installed engines
use a closed cooling system with no water discharge.
Standards of performance based on either alternative would also
have no solid waste impact.
There would also be no adverse noise impact resulting from stan-
dards of performance based on either alternative. Fan noise levels
from large-bore stationary engine installations could increase slightly
9-53
-------
as a result of Increased cooling requirements; however, 1n typical
Installations such as municipal generator plants, pipeline compressor
stations or industrial process plants, such Increases are insignificant
in comparison to existing noise levels.
Thus, as reflected in the summary Table 9-6, the environmental
Impacts of standards of performance based on either alternative are
small and reasonable.
Energy Impacts
The potential energy impact of standards of performance based on
either alternative is small. As discussed in Section 6.2, standards of
performance based on alternative I could increase the fuel consumption
of a typical blower scavenged or turbocharged gas engine by approxi-
mately one percent, whereas standards of performance based on alterna-
tive II could increase the fuel consumption by approximately two percent.
A typical 2000 horsepower blower scavenged or turbocharged gas engine
has an uncontrolled fuel consumption of approximately 343,000 scf
natural gas per day (the energy equivalent of 2842 gallons oil per
day). Increases translate into actual increased fuel consumption of
approximately 3500 scf and 7000 scf natural gas per day (the energy
equivalent of 25 and 50 gallons oil per day).
Standards of performance based on alternative I would increase the
fuel consumption of a typical naturally aspirated gas engine by approxi-
mately six percent. Standards of performance based on alternative II,
however, would increase the fuel consumption by approximately eight
percent.
A typical 2000 horsepower naturally aspirated gas engine has an
9-54
-------An error occurred while trying to OCR this image.
-------
uncontrolled fuel consumption of approximately 343,000 scf natural gas
per day (the energy equivalent of 2842 gallons oil per day).
These percentage increases represent actual fuel consumption
Increases of about 20,600 and 27,400 scf natural gas per day (the
energy equivalent of 142 and 189 gallons oil per day).
Standards of performance based on alternative I could Increase the
fuel consumption of a typical dual-fuel engine by approximately one
percent, whereas standards of performance based on alternative II
could Increase the fuel consumption by approximately three percent. A
typical dual-fuel engine rated at 2000 horsepower has an uncontrolled
fuel consumption of approximately 297,000 scf natural gas per day (the
energy equivalent of 2150 gallons oil per day). These percentage
Increases represent an actual fuel consumption increase of about 6100
and 12,200 scf natural gas per day (the energy equivalent of 43 and 86
gallons oil per day).
Standards of performance based on alternative I could Increase the
fuel consumption of a typical dlesel engine by approximately three
percent, whereas standards of performance based on alternative II could
Increase the fuel consumption by approximately seven percent. For a
typical dlesel engine rated at 2000 horsepower, with an uncontrolled
fuel consumption of approximately 2320 gallons of oil per day, these
percentage Increases represent actual fuel consumption Increases of
about 70 and 160 gallons of oil per day.
Thus, the potential energy Impact 1n the fifth year after the
standard goes Into effect, based on alternative I, would be equivalent
to approximately 1.03 million barrels of oil per year compared to an
9-56
-------
uncontrolled fuel consumption of 1C engines affected by the standard of
31 million barrels per year. The potential energy Impact 1n the fifth
year after the standard goes Into effect, based on alternative II,
would be equivalent to approximately 1.5 million barrels of oil per
year.
It should be noted that the largest Increase represents only 0.01
percent of the 1977 domestic consumption of crude oil and natural gas.
The largest Increase also represents only 0.03 percent of the projected
total oil Imported to the United States five years after the standards
go Into effect.
Thus, as reflected 1n the summary Table 9-7, the energy Impacts of
standards of performance based on either alternative are small and rea-
sonable.
TABLE 9-7. ENERGY IMPACTS OF ALTERNATIVES
Engine
Fuel
Type
Gasb
Gasc
Dual Fuel
Diesel
Totals of
all new
engines
after
5 years
Uncontrolled9
Fuel Consumption
(gal/day)
2842d
2842
2151
2317
31,000,000
bbls o1l/yr
Increase 1n Fuel Consumption (gal/day)
Alternative I Alternative II
25
142
43
70
1.03 million
bbls oll/yr
50
189
86
162
1.5 million
bbls o1l/yr
Jyplcal 2000 horsepower engine
.Blower scavenged or turbocharged
^Naturally aspirated
Expressed as equivalent oil consumption
9-57
-------
Economic Impact of Alternatives
Manufacturers of stationary internal combustion engines would incur
additional costs due to standards of performance. These costs however
would be small. As discussed in Section 6.3, these costs are a result
of one or more of the following activities that may be needed to manu-
facture engines which meet standards of performance: (1) extended
testing to verify the effectiveness of a particular control approach;
(2) development and application of NOX controls to existing engine
designs; and (3) engineering, tooling and pattern costs for minor re-
design of an engine family to accomodate application of a control tech-
nique. It is estimated that the total costs to the manufacturer for
each engine model family, including development, durability tests, and
retooling, would be approximately: (1) $125,000 for retard and air-to-
fuel change; (2) $150,000 for manifold air temperature reduction; and
(3) $25,000 for derate. For each manufacturer, therefore, total costs
would vary depending on (1) the number of engine model families pro-
duced, (2) their degree of advancement in emission testing, (3) the
uncontrolled emission levels of their engines, (4) the development and
durability testing required to produce engines that can meet proposed
standards of performance, and (5) the emission control technique selected,
As reported in Section 8.4, the economic impacts on manufacturers
arising from these cost penalties associated with standards of perfor-
mance based on either alternative would be small.
The manufacturer's total capital investment requirements for
developmental testing of engine models is estimated to be about $4.5
million to comply with standards of performance based on alternative I
9-58
-------
and about $5 million to comply with standards of performance based on
alternative II. These expenditures would be made over a two year period.
Analysis of the financial statements of engine manufacturers or their
parent companies Indicates that the manufacturers' overhead budgets are
sufficient to support the development of these controls without adverse
Impact on their financial position.
As discussed 1n Section 8.4.1.2, manufacturers would not experience
significant differential cost Impacts among competing engine model
families. The cost penalties summarized 1n Table 8-16 reflect the
range of total annual1zed cost penalties to the end use applications
for each engine fuel type and manufacture for each alternative. These
costs for each major end use, the cross-price Inelasticities, and the
Importance of each end use market to the companies' total Internal
combustion sales were analyzed to determine the relative sales advan-
tages between companies. Consequently, these analyses Indicated that
no significant sales advantages or disadvantages would develop among
competing manufacturers for standards of performance based on either
alternative. Based on "worst case" assumptions the maximum intra-
Industry sales losses would be about six percent as a result of standards
of performance based on either alternative. Thus, the intra-1ndustry
Impacts would be moderate and not cause any major dislocations within
the Industry.
These total annualized cost penalties imposed on 1C engines by
standards of performance would also have very little impact with regard
to increasing sales of gas turbines. Turbines do not compete with
internal combustion engines based on annualized costs alone, due to
9-59
-------
their higher operating costs (fuel). As discussed 1n Subsection 8.4.1.3,
the total annual1zed cost penalty associated with standards of perfor-
mance based on either alternative would bring the cost of owning Internal
combustion engines up to that of turbines 1n only one case -- dlesel
Internal combustion engines used 1n electric generation. This conclusion,
however, 1s based on limited data. It 1s quite likely, however, that
this manufacturer's line of dlesel engines, through minor combustion
modifications, could reduce their NOV emission to levels comparable to
A
that of other manufacturers. Further, due to technical limitations,
economic considerations, and customer preference, 1t 1s unlikely that
1C engine users will switch to gas turbines. For example, 1t 1s unlike-
ly that turbines would replace dlesel engines 1n plants using banks of
smaller engines, unless the entire bank were replaced with one turbine.
Standards of performance based on alternative I would result 1n no
loss of sales to gas turbines whereas standards of performance based on
alternative II would result in the possible loss of sales for one
dlesel manufacturer.
Thus, the economic Impacts on the manufacturers arising from
standards of performance based on either alternative are considered
small and reasonable.
The application of NO controls will also increase costs to the
rt
engine user. The magnitude of this increase will depend upon the
amount and type of emission control applied. As was shown in Section
6.3, various control approaches affect Initial costs, fuel consumption,
and maintenance differently. Fuel penalites, though, are the major
9-60
-------
factor affecting this Increase for high usage engines.
The following four end uses were selected to represent the major
applications of dlesel, dual-fuel, and natural gas engines: (1) dlesel
engine, electrical generation; (2) dual-fuel engine, electrical genera-
tion, (3) gas engine, oil and gas transmission and (4) gas engine, oil
and gas production.
The total annual1zed cost of a typical uncontrolled dlesel fuel,
electrical generation engine 1s about 2.5^/hp-hr. For a typical 2000
horsepower engine operating 8000 hours per year, this total annual1zed
cost would be about $450,000 per year. Standards of performance based
on alternative I would Increase this total annual1zed cost by about
0.04^/hp-hr (1.5 percent). Similarly, standards of performance based on
alternative II would Increase the total annual 1 zed cost by about
O.lltf/hp-hr (4.5 percent). For the engine mentioned above, these
values translate Into dollar amounts of about $6,400 and $17,600 respec-
tively.
The total annual1zed cost of a typical uncontrolled dual-fuel,
electrical^generation engine 1s about 2.8^/hp-hr. For a typical 2000
horsepower engine operating 8000 hours per year this would be about
$448,000 per year. As a result of standards of performance based on
alternative I this total annual 1zed cost would Increase by about 0.07
-------
fuel, oil and gas transmission engine 1s about 2.2<£/hp-hr. For a
typical 2000 horsepower engine operating 8000 hours per year, this
total annual1zed cost would be about $354,000 per year. Standards of
performance based on alternative I would Increase this total annual1zed
cost by about 0.02
-------
would have to expend additional capital to purchase more expensive
engines. This capital cost penalty however, 1s small. A two percent
Increase In engine price would be expected on the average as a result
of standards of performance based on either alternative. Typical
Initial costs for uncontrolled dlesel and dual-fuel, electrical generation
engines, and natural gas, oil and gas transmission engines are about
$150/hp-hr. Initial costs for natural gas fuel, oil and gas production
engines are about $50/hp. For typical 2000 horsepower engines, these
Initial capital costs would be about $300,000 and $100,000, respectively.
Standards of performance based on either alternative would Increase
the Initial capital cost of a typical dlesel or dual-fuel, electrical
generation engine or natural gas fuel, oil and gas transmission engine
rated at 2000 horsepower by about $6000.
In contrast, standards of performance based on either alternative
would Increase the Initial capital cost of a typical natural gas, oil
and gas production engine rated at 2000 horsepower by about $2000.
The total additional capital cost for all users would equal about
$9.6 million per year on a cumulative basis on either alternative
compared to total uncontrolled costs of all new engines of $450 million
per year.
As discussed 1n Section 8.4, the economic Impacts on users arising
from the cost penalties associated with standards of performance based
on either alternative would be small. In general, these Impacts
translate Into price increases for the end products or services provided
by the Industrial and commercial users of large stationary Internal
combustion engines. The electric utility industry would realize a
9-63
-------
price increase after five years of 0.02 percent to comply with stan-
dards of performance based on either alternative. After five years,
delivered natural gas prices would increase 0.02 percent due to the
application standards of performance based on alternative I and 0.04
percent due to the application standards of performance based on al-
ternative II.
Even after a full phase-in period of 30 years, during which new
controlled engines would replace all existing uncontrolled engines, the
electric utility industry would realize a price increase of only 0.1
percent to comply with standards of performance based on either alter-
native. Similarly, delivered natural gas prices would increase only
0.1 percent due to the application of standards of performance based on
alternative I and 0.3 percent to comply with standards of performance
based on alternative II. Thus, the economic impacts of standards of
performance based on either alternative are considered small and rea-
sonable.
Conclusions
Based on this assessment of the impacts of each alternative, and
given the fact that alternative II achieves a greater degree of NOY re-
A
duction, it is selected as the best technological system of continuous
emission reduction of NO from stationary large-bore 1C engines considering
A
the cost of achieving such emission reduction, any nonair quality
health and environmental impact and energy requirements.
Table 9-8, which follows, summarizes the economic impacts of each
alternative.
9-64
-------
TABLE 9-8 ECONOMIC IMPACTS OF ALTERNATIVES
Impact
Impact on Manufacture
Capital budget requirements
Uncontrolled
Level of Cost
• *
Alternative I
$4.5 million over two years;
able to generate Internally
from profits.
Alternative II
$5 ml 11 (on over two years j able
to generate Internally from
profits.
Intra-lndustry competition
Competition from gas turbines
Impact on End-Use Aoolleitlorn
Total annual lied costa
Diesel fuel, electrical
generation
Dual-fuel, electrical gen
eratlon
Natural gas fuel, oil and
gas transmission
Natural gas fuel, oil and
gas production
Totals of all new engines
after 5 years
Capital Cost Penalty8
Diesel fuel, electrical
generation or dual fuel,
electrical generation or
natural gas fuel, oil and
gas transmission
Natural gas fuel, oil and
gas production
Totals etc.
.Impact on Product Prices and
Users ~~
Electricity prices
Gas prices
2,5
-------
9.5 SELECTION OF FORMAT FOR THE STANDARDS
A number of different formats could be used to limit NO emissions
^
from large stationary Internal combustion engines. Standards could be
developed to limit emissions 1n terms of: (1) percent reduction, (2)
mass emission per unit of energy (power) output, (3) mass emissions per
unit of energy (fuel) Input, or (4) concentration of emissions 1n the
exhaust gases discharged to the atmosphere.
Analysis of the effectiveness of the various demonstrated NO
A
emission control techniques clearly shows that what 1s demonstrated 1s
the ability to achieve a percent reduction 1n NO emissions. In other
^
words, application of each emission control technique to the same
degree (I.e., eight degrees of ignition retard to five percent change in
air-to-fuel ratios) will result 1n essentially the same percentage reduc-
tion in NO emissions. However, a percent reduction format is highly im-
rt
practical for two reasons. First a reference uncontrolled N0¥ emission le-
o
vel would have to be established for each manufacturer's engine, a diffi-
cult task since some manufacturers produce as many as 25 models which
are sold with several ratings. Second, a reference uncontrolled NOY
^
emission level would have to be established for any new engines developed
after promulgation of the standard. This would be quite simple for
engines that employed NO control techniques such as ignition retard or
A
air-to-fuel ratio change to comply with standards. Emissions could be
measured without the use of these techniques. For engines designed to
comply with the standards through the use of combustion chamber modifi-
cation, however, this would not be possible. Thus, new engines would
receive no credit for the NO emission reduction achieved by combustion
A
9-66
-------
chamber redesign.
A mass-per-un1t-of-energy-output format, typically referred to as
brake-specific emissions (g/hp-hr), relates the total mass of NO
A
emissions to the engine's productivity. Although brake-specific mass
standards (g/hp-hr) appear meaningful because they relate directly to
the quantity of emissions discharges Into the atmosphere, there are
disadvantages In that enforcement of mass standards would be costly and
complicated in practice. This can be Illustrated by examining the
relationship between brake-specific mass emissions (BSNO ) and the
/\
parameters that would have to be measured 1n the field:
BSNOX^ NOX (me) (1/w)
where:
BSNOX * grams N0x/horsepower-hour
NOX » concentration of NOX in exhaust, parts per million (ppm)
me « exhaust mass flowrate, Ib/hr
w - power output, horsepower
Thus, exhaust flow and power output would have to be determined 1n
addition to NOX concentration. For example, to determine exhaust gas
flowrate, one of three methods can be used: (1) directly measure exhaust
volume flowrate: (2) measure inlet air and fuel flowrate; and (3)
measure all exhaust carbon constituents (primarily HC, CO, C02), and
fuel flow, and conduct a fuel analysis.
Since large internal combustion engines have very large exhaust
flowrates (in excess of 50,000 cfm), exhaust flowrates are difficult to
determine accurately in either the field or laboratory. Similarly, the
accurate measurement of inlet airflow is difficult. Thus, methods (1)
9-67
-------
or (2) are unlikely to be used 1n practice for large engines.
Although method (3) has been used 1n the field to determine brake-
specific mass emissions, the measurement of three additional exhaust
gas constituents and the fuel analysis considerably complicate the test
procedure^31). Moreover, 1t 1s difficult to accurately measure fuel
flow over a short Interval (typically less than 1/2 hour, which would
be required for a performance test). Thus, the determination of exhaust
gas flowrates 1n the field 1s difficult and complicated.
Another disadvantage of the brake-specific mass emissions format 1s
that power output must be determined. Power can be determined from an
engine dynamometer 1n the laboratory, but dynamometers cannot be used 1n
the field. Power output could be determined by: (1) Inferring the
power from engine operating parameters (fuel flow, rpm, manifold pressure,
etc.) or (2) Inferring engine power from the output of the generator or
compressor attached to the engine. In practice, however, these approaches
are time consuming and are less accurate than dynamometer measurements.
A format limiting NOX emissions per unit of energy (fuel) Input
would be specified 1n terms of grams NOX per joule fuel Input (equiva-
lent to Tb NOV/M Btu). The advantage of this format 1s that no power
A
measurement would be required, thereby simplifying enforcement. How-
ever, as with a brake-specific mass emission format, total exhaust gas
mass flowrates must be calculated, and as was discussed earlier, all
methods for this determination are difficult under field conditions.
In addition, standards of performance based on fuel Input could
penalize more efficient engines, which typically operate at higher
temperatures and pressures, leading to higher NOX emissions. For ex-
9-68
-------
ample, given two engines with the same brake-specific emissions, the
more efficient engine, which consumes less fuel, will have a higher
fuel-based emissions level because the ratio of mass NOX emissions per
unit of fuel energy Input has a smaller denominator. Thus, this format
could offset other Incentives manufacturers have to develop more effi-
cient engines.
Another possible format would be to limit the concentration of NOX
emissions 1n the exhaust gases discharged to the atmosphere. Concentrations
would be spedfed 1n terms of parts of NO, per million (ppm) parts of
exhaust (volumetric). The major advantage of this format 1s In the
simplicity of Its enforcement. As compared to the formats discussed
previously, only a minimum of data and calculations are required, which
decreases testing costs and minimizes errors In determining compliance
with an emission standard. Measurements are direct; only NOX and 02
concentration measurements of exhaust gas must be made. The NOX measure-
ment reads out directly in ppm of dry exhaust, and the oxygen measure-
ment, required to prevent a user from diluting the exhaust gas with air
and lowering the NOX concentration, reads out in percent-02. A refer-
ence concentration of oxygen, however, must be established for this
format.
The primary disadvantages associated with concentration standards
are: (1) a standard could be circumvented by dilution of exhaust gases
discharged Into the atmosphere, which lowers the concentration of the
emissions but does not reduce the total mass emitted, and (2) a concen-
tration standard could penalize high efficiency engines because more
efficient engines generally discharge higher concentrations of NO
x
9-69
-------
emissions due to higher operating temperatures and pressures as mentioned
above (although the mass emission rate may be the same as a lower effi-
ciency engine). A concentration standard based on low efficiency engines,
therefore, could penalize high efficiency engines. Both these problems,
however, can be overcome through the use of appropriate "correction"
factors.
Since the percent reduction format 1s Impractical, and the problems
associated with the enforcement of mass standards (mass per unit energy
output or Input) appear to outweigh the benefits, the concentration for-
mation was selected as the format for standards of performance for large
stationary Internal combustion engines.
As mentioned above, because a concentration standard can be circum-
vented by dilution of the exhaust gases, measured concentrations must
be expressed relative to some fixed dilution level. For combustion
processes, this can be accomplished by correcting measured concentra-
tions to a reference concentration of oxygen. The oxygen concentration
1n the exhaust gases 1s related to the excess (or dilution) air.
Typical oxygen concentrations 1n large-bore Internal combustion engines
can range from 8 to 16 percent but are normally about 15 percent.
Thus, referencing the standard to a typical level of 15 percent oxygen
would prevent circumvention by dilution. (Section 9.6 discusses the
correction factor for adjusting measured NOX concentrations to an
oxygen concentration of 15 percent).
As also mentioned above, selection of a concentration format
could penalize high efficiency internal combustion engines. These
highly efficient engines generally operate at higher temperatures and
9-70
-------
pressures, and as a result discharge gases with higher N0x concentrations
than less efficient engines, although both engines' brake-specific mass
emissions could be the same. Thus, a concentration standard based on
low efficiency engines could effectively require more stringent con-
trols for high efficiency engines. Conversely, a concentration stan-
dard based on high efficiency engines would require no controls.
Consequently, selecting a concentration format for standards of per-
formance requires an efficiency factor to permit higher N0y emission
A
from more efficient engines.
The Incentive for manufacturers to Increase engine efficiency 1s
to lower engine fuel consumption. Therefore, the objective of an effi-
ciency adjustment factor should be to give an emissions credit for the
lower fuel consumption of more efficient internal combustion engines.
Since the fuel consumption of internal combustion engines varies linearly
with efficiency, a linear adjustment factor is selected to permit
increased NOX emissions from highly efficient internal combustion
engines. A linear efficiency adjustment factor also effectively limits
NOX emissions to a constant mass emission rate per unit of power output.
The efficiency adjustment factor needs to be referenced to a
baseline efficiency. Most large existing stationary internal combustion
engines fall in the range of 30 to 40 percent efficiency. Therefore,
35 percent is selected as the baseline efficiency. The efficiency of
internal combustion engines is usually expressed in terms of heat rate.
The heat rate of engines operating at 35 percenc en iciency is about
7270 Btu/hp-hr. Thus, the following linear adjustment factor is selected
to permit increased NOX emissions from high efficiency large stationary
9-71
-------
internal combustion engines:
7270
where:
xa = x
a Y
X • Adjusted NO emissions permitted at 15 percent oxygen, ppmv
a x
X = NO emission limit specified in the standards at 15 percent
A
oxygen on a dry basis
Y = LHV heat input per unit of power output (Btu/hp-hr)
NOTE: Above adjustment is made at standard atmospheric
conditions of 29.92 m Hg, 85°F, and 75 grains moi-
sture per pound of dry air.
This efficiency adjustment factor permits a linear increase in NOX
emission with increased efficiencies above 35 percent. This adjustment
would not be used to adjust the emission limit downward for internal
combustion engines with efficiencies of less than 35 percent. This
efficiency adjustment factor also applies only to the 1C engine itself
and not the entire system of which the engine may be a part. Since
Section 111 of the Clean Air Act requires the use of the best system of
emission reduction in all cases, this precludes the application of the
efficiency adjustment factor to an entire system. For example, 1C
engines with waste heat recovery may have a higher overall efficiency
than the 1C engine alone. Thus, the application of the efficiency
adjustment factor to the entire system would permit greater NOX emissions
because of the system's higher overall efficiency, and would not necessarily
require the use of the best demonstrated system of emission reduction on
the 1C engine.
9-72
-------
9.6 SELECTION OF NUMERICAL EMISSION LIMITS
Overall Approach
As mentioned earlier, It 1s difficult to select , specific NO
emission limit which ,11 1C engines could meet primarily through the
use of ignition retard or air-to-fuel ratio change. Because of In-
herent differences among various 1C engines with regard to uncontrolled
NOX emission levels, there exists a rather large variation within the
data and Information Included 1n the SSEIS concerning controlled NOX
emission levels. Generally speaking, engines with relatively low "
uncontrolled NOX emissions achieved low controlled NOX emission levels.
and engines with high uncontrolled NO, emission levels achieved relatively
high controlled NOX emission levels. Consequently, the following
alternatives were considered for selection of the numerical concentra-
tion emission Hmlts based on a 40 percent reduction 1n NOX emissions:
1. Apply the 40 percent reduction to the highest observed
uncontrolled NOV emission level.
*\
2. Apply the 40 percent reduction to a sales-weighted average
uncontrolled NO emission level.
/\
3. Apply the 40 percent reduction to this sales-weighted average
uncontrolled NOX emission level plus one standard deviation.
The highest observed uncontrolled NOX emission levels for dlesel,
dual-fuel and gas engines discussed In alternative I above can be found
in Figures 4-49{.-c). respective^. The uncontrolled levels for dual-
fuel engines are generally lo«er than those for dlesels, which are gen-
erally lower than those for gas engines. The highest uncontrolled
9-73
-------
levels for each fuel type are as follows: (1) gas, 29 g/hp-hr, (2)
dual-fuel, 15.0 g/hp-hr, and (3) diesel, 19 g/hp-hr.
The sales-weighted uncontrolled N0¥ emission levels which are used
A
as the base levels in the second alternative are discussed in Section
4.3.4. It was noted that uncontrolled NOX emission levels vary among
both engines of the same fuel type and of different fuels (even after
considering the effects of ambient conditions and measurement methods).
Therefore, an average uncontrolled level was determined by applying a
sales-weighting to each manufacturer's average uncontrolled NOX emis-
sions for engines of each fuel type (see Section 4.3.4). The sales-
weighting, based on horsepower sold, gives more weight to those engine
models which have the highest sales. The sales-weighted average uncon-
trolled N0¥ emission level for each engine fuel type are as follows:
J\
(1) gas, 15 g/hp-hr, (2) dual-fuel, 8 g/hp-hr, and (3) diesel, 11 g/hp-
hr.
The third alternative incorporates a "margin for engine variabili-
ty" by adding one standard deviation to the sales-weighted average
uncontrolled NO emission level and then applying the 40 percent reduc-
^
tion. Standard deviations discussed were calculated from the uncontrolled
NO emission data included in the SSEIS, assuming it had a normal distri-
J\
bution. A subsequent statistical evaluation of the data indicated that
this assumption was valid (see Appendix C for a complete discussion).
The standard deviations for each engine fuel type are as follows: (1)
gas, 4 g/hp-hr, (2) dual-fuel, 3.2 g/hp-hr and (3) diesel, 3.7 g/hp-hr.
The standard deviation of the uncontrolled NOX emission data base
is relatively large compared to the sales-weighted average uncontrolled
9-74
-------
Internal combustion engines:
x - y 727°
Aa " * ~
where:
Xg • Adjusted N0x emissions permitted at 15 percent oxygen, ppmv
X = NOX emission limit specified in the standards at 15 percent
oxygen on a dry basis
Y = LHV heat input per unit of power output (Btu/hp-hr)
NOTE: Above adjustment is made at standard atmospheric
conditions of 29.92 m Hg, 85°F, and 75 grains moi-
sture per pound of dry air.
This efficiency adjustment factor permits a linear increase in NO
A
emission with increased efficiencies above 35 percent. This adjustment
would not be used to adjust the emission limit downward for internal
combustion engines with efficiencies of less than 35 percent. This
efficiency adjustment factor also applies only to the 1C engine itself
and not the entire system of which the engine may be a part. Since
Section 111 of the Clean Air Act requires the use of the best system of
emission reduction in all cases, this precludes the application of the
efficiency adjustment factor to an entire system. For example, 1C
engines with waste heat recovery may have a higher overall efficiency
than the 1C engine alone. Thus, the application of the efficiency
adjustment factor to the entire system would permit greater NO emissions
n
because of the system's higher overall efficiency, and would not necessarily
require the use of the best demonstrated system of emission reduction on
the 1C engine.
9-72
-------
concentration emission limit.
The first alternative 1s to apply the 40 percent reduction to the
highest uncontrolled NOX emission level within a fuel category, For
example, Figure 4-49(c), which summarizes NOX emission reductions
achieved by gas engines, lists 29 g/hp-hr as the highest uncontrolled
NOX emission level. The application of a 40 percent reduction would
lead to an emission level of about 17 g/hp-hr, As Illustrated 1n
Figure 9-16, 1f this level were selected as a standard of performance,
99 percent of production gas engines could easily meet the emission
limit by reducing emissions by 40 percent or less. However, 69 percent
of production engines would not have to reduce NOX emissions at all.
Only one percent of production engines would have to reduce NO, emissions
by more than 40 percent.
The second alternative 1s to apply the 40 percent reduction to the
sales-weighted average uncontrolled NOX emission level. For example,
the sales-weighted average uncontrolled NOX level for gas engines 1s 15
g/hp-hr. The application of a 40 percent reduction would lead to an NOX
emission level of 9 g/hp-hr. As illustrated in Figure 9-16, if this
level were selected as a standard of performance, 50 percent of produc-
tion gas engines could meet the standard with 40 percent or less reduc-
tion in NOX emissions. However, 50 percent of production gas engines
would be required to reduce NOX emission by greater than 40 percent.
Only seven percent of production gas engines would not have to reduce NOX
emissions at all.
The third alternative is to base the standard on a 40 percent
reduction in NO emissions from the sales-weighted average uncontrolled
X
N0¥ emission level plus one standard deviation. For example, the sales-
^
9-76
-------
weighted average uncontrolled NO, emission level for gas production gas
engines 1s 15 g/hp-hr and the standard deviation of the production gas
engine data base (Appendix C) Is 4 g/hp-hr. Thus, the appl,cation of a
40 percent reduction to the sum of these two values would lead to an
emission level of 11 g/hp-hr. As Illustrated 1n Figure 9-16, 1f this
level were selected as a standard of performance, 84 percent of the pro-
duction gas engines would not have to reduce N0X emission at all. Only
16 percent of the production gas engines would have to reduce NOX
emissions by more than 40 percent.
Similarly, applying the three alternatives to dual-fuel engines
s results similar to those for gas engines. The highest uncontrolled
NOX emission level from a dual fuel engine is ,5 g/hp-hr. The appli-
cation of a 40 percent reduction would lead to an emission level of 9
9/hp-hr. ,f this level were selected as a standard of performance, 98
percent of production dual-fuel engines could easily meet the emission
limit by reducing NO, emissions by 40 percent or less. However, 62
percent would have to achieve no reduction In NO, emissions. Only two
percent of production engines would have to reduce NO, emissions by more
than 40 percent.
The sales-weighted average uncontrolled NO, emission level for
dual-fuel engines is 8 g/hp-hr. The application of a 40 percent reduc-
tion would lead to a NO, emission level of about 5 g/hp-hr. If this
leve, were selected as a standard of performance, 64 percent of the
production dual-fuel engines could meet the standard by reducing NO
en-issions 40 percent or less. Only ,8 percent of the production dull-
fuel engines would not have to reduce NO, emissions at all. Also, 46
9-77
-------
percent of the production dual-fuel engines would be required to reduce
NO emissions by greater than 40 percent.
The standard deviation of the production dual-fuel engine data base
1s 3.2 g/hp-hr. Thus, the application of a 40 percent reduction to the
sum of the sales-weighted average uncontrolled NOX emission level (8
g/hp-hr) and the standard deviation (3.2 g/hp-hr) would lead to an
emission level of 7 g/hp-hr. If this level were selected as a standard
of performance, 87 percent of the production dual-fuel engines could
easily meet the emission limit by reducing emissions by 40 percent or
less. However, 48 percent of the production gas engines would not have
to reduce N0y emission at all. Only 13 percent of the production gas
A
engines would have to reduce NOX emissions by more than 40 percent.
Finally, the application of the three alternatives to diesel engines
also yields very similar results. The highest uncontrolled NOX emission
level from a diesel engine is 19 g/hp-hr. The application of a 40
percent reduction would lead to in an emission level of about 11 g/hp-
hr. If this level were selected as a standard of performance, 98 percent
of production diesel engines could easily meet the emission limit by
reducing emissions by 40 percent or less. However, 40 percent would
have to achieve no reduction in NOX emissions. Two percent of production
engines would be required to reduce NOX emission by more than 40 percent.
The sales-weighted average uncontrolled NOX emission level for
diesel engines is 11 g/hp-hr. The application of a 40 percent reduction
would lead to a NOX emission level of about 7 g/hp-hr. If this level
were selected as a standard of performance, 56 percent of the production
diesel engines could meet the standard by reducing NO, emissions 40
9-78
-------An error occurred while trying to OCR this image.
-------
percent or less. Only 14 percent of the production diesel engines would
not have to reduce NO emissions at all. However, 44 percent of the
X
production diesel engines would be required to reduce NO emission by
A
greater than 40 percent.
The standard deviation of the production diesel engine data base 1s
3.7 g/hp-hr. Thus, the application of a 40 percent reduction to the sum
of the sales-weighted uncontrolled NOX emission level (11 g/hp-hr) and
the standard deviation (3.7 g/hp-hr) would lead to a NOV emission level
A
of 9 g/hp-hr. If this level were selected as a standard of performance,
86 percent of the production diesel engines could easily meet the emis-
sion limit by reducing emissions by 40 percent or less. However, 29
percent of the production gas engines would not have to reduce NOX
emissions at all. Only 14 percent of the production gas engines would
have to reduce NOV emissions by more than 40 percent.
A
Table 9-9 presents a summary of the statistical analysis of stan-
dards of performance based on each alternative for each engine fuel
type. If standards of performance were based on Alternative I, essen-
tially all engines could achieve the emission limit by reducing NOX
emissions 40 percent or less. A significant reduction in NOX emissions
would not be achieved, however, since 50 to 70 percent of the 1C en-
gines would not have to reduce NO emissions at all. If the standards
A
of performance were based on Alternative II, about 50 percent of the
1C engines (in all categories) would have to reduce NOX emissions by
greater than 40 percent. Less than 10 percent would not have to reduce
NO emissions at all. Thus, this alternative would achieve a signifi-
A
cant reduction in NO emissions from new sources. If standards of
A
9-80
-------
TABLE 9-9. SUMMARY OF STATISTICAL ANALYSES OF ALTERNATE
EMISSION LIMITS
GAS ENGINES
Alternative
Standard
Percent required to apply
less than or equal to
40 percent control
17
99
9
50
11
84
Percent required to do
nothing
Percent required to apply
more than 40 percent con-
trol
1 50
DUAL-FUEL ENGINES
Alternative
—
Standard
••• • I l.l_
Percent required to apply
less than or equal to
40 percent control
"^ •"•"' '
Percent required to do
nothing
98 56
Percent required to apply
more than'40 percent con-
trol
18
16
Percent required to apply
less than or equal to
40 percent control
f— . — .
Percent required to do
nothing
Percent required to apply
more than 40 percent con-
trol
I ••••Bin .... -.- -
DIESEL ENG
'""•""—'—•"' -!-..• ^_
Alternative
~
Standard
98 54
62 18
2 46
INES
I II
i.iii.ni. __.„—,, i i •__,.».„..,,..
11 7
87
48
13
III
i,.l
9
86
29
14
9-81
-------
performance were based on Alternative III, the results would be similar
to those achieved with Alternative I. About 85 percent of engines
could easily meet the standards by reducing N0¥ emissions by less than
A
40 percent. About 20 to 30 percent of 1C engines would not have to
reduce NOY emissions at all and about 15 percent of 1C engines would
n
have to reduce NO emissions by more than 40 percent.
A
In light of the high priority discussed earlier which has been
directed toward reducing NOV emissions and the significance of 1C
A
engines 1n terms of their contribution to NOX emissions from stationary
sources, the second alternative was chosen for selecting the MOV
A
emission concentration limit. This approach will achieve the greatest
reduction 1n NOX emissions from new 1C engines.
Selection of Limits
A concentration (ppm) format was selected for the standards. Con-
sequently, the brake-specific NOX emission limits corresponding to the
second alternative for selecting numerical emission limits (I.e., gas -
9 g/hp-hr; dual-fuel - 5 g/hp-hr; dlesel - 7 g/hp-hr) must be converted
to concentration limits (corrected to 15 percent oxygen). This may be
done by dividing the brake-specific volume of NOX emissions by the brake
specific total exhaust gas volume. Determining the brake-specific volume
of NOV emissions is straightforward. Determining the brake-specific
A
total exhaust gas volume is more complex, in that the brake-specific
exhaust flow and the exhaust gas molecular weight are unknown. Knowing
the fuel heating value and composition, the brake-specific fuel consumption,
and assuming 15 percent excess air, however, defines these unknowns.
(The complete derivation is explained in detail in Appendix C-5).
9-1
-------
Combining these factors leads to the following conversion factor:
(ir)x BSNO>
NO 1^6 / * —"x
x
15
16.6 + 3.29 Z!
x BSFC
12.0 + Z
fl
where:
NOX a NOX concentration ppmv corrected to 15-percent oxygen
on a dry basis.
BSNOX - Brake-specific NOX emission, g/hp-hr.
BSFC « Brake-specific fuel consumption, g/hp-hr.
Z - Hydrogen/Carbon ratio of the fuel.
The numerator 1s the brake specific volume of NOY emissions multl-
6
pi led by 10 in order to convert the decimal equivalent to ppm. In the
denominator, the brake-specific total dry exhaust gas volume with 15
percent excess oxygen 1s expressed as a function of the fuel's hydrogen/
carbon ratio and the brake-specific fuel consumption. The fuel consump-
tion has been converted from Btu/hp-hr to g/hp-hr using the fuel's
lower heating value (LHV).
For natural gas, a hydrogen to carbon (H/C) ratio of 3.5 and an LHV
of 20,000 Btu/lb was assumed. Diesel ASTM-2 has a H/C ratio of 1.8 and
an LHV of 18,320 Btu/lb.
Using the above equation, plots of BSNOX versus ppm were generated
for each fuel type. The uncontrolled values of BSNO and BSFC were
/\
used for each engine, producing plots similar to Figure 9-17. As
shown, agreement between this equation and actual emission data relating
9-83
-------An error occurred while trying to OCR this image.
-------
NOX emissions concentrations to BSNOX 1s excellent, Comparison of this
conversion factor with available raw data also shows excellent agree-
ment (see Appendix C.5). Applying this conversion factor to the brake-
specific emission limits associated with the second alternative for NO
o
emission limits leads to the NOX concentration emission limits for
large stationary Internal combustion engines summarized 1n Table 9-10.
These emission limits have been rounded upward to the nearest 100
ppm to Include a "margin" to allow for source variability. The standard
for dlesel engines has also bean applied to dual-fuel engines. If a
separate emission limit had been selected for dual-fuel engines, the
corresponding numerical NOX concentration emission limit would be 400
ppm. Sales of dual-fuel engines have ranged from 17 to 95 units annually
over the past five years, with a general trend of decreasing sales. Dual-
fuel engines serve the same applications as diesel engines, and new
dual-fuel engines will likely operate primarily as diesel engines due
to increasingly limited natural gas supplies. Thus, combining of
dual-fuel engines with diesel engines for standards of performance will
have little adverse impacts and will simplify enforcement of the standards
of performance.
TABLE 9-10. NUMERICAL NO CONCENTRATION EMISSION LIMITS FOR LARGE
STATIONARY INTERNAL COMBUSTION ENGINES
Engine
•••••^
Gas
Diesel/dual-fuel
x
Emission
Limit
700 ppm
600 ppm
9-85
-------
As discussed in Section 4.2.1, the effect of ambient atmospheric
conditions on NO emission from large stationary internal combustion
A
engines can be significant. Therefore, to enforce the standards uni-
formly, NO emissions must be determined relative to a reference set of
A
ambient conditions. All existing ambient correction factors were
reviewed that could potentially be applied to large stationary internal
combustion engines to correct NOX emissions to standard conditions. A
detailed discussion of this review is presented in Section 4.2 and
Appendix C.2.
The correction factors that were selected for both spark ignition
(SI) and compression ignition (CI) engines are presented in Table 4-2,
which is reproduced below as Table 9-11. For the compression ignition
engines (i.e., diesel and dual-fuel), a single correction factor for
both temperature and humidity was selected. Constants for use in this
correction factor, which have been experimentally derived, are summarized
in the table. For spark ignition engines (i.e., gas), separate correc-
tion factors were selected for humidity and temperature, and measured
NO emissions are corrected to reference ambient conditions by multiplying
A
these two factors together. No correction factor was selected for
changes in ambient pressure, because no generalized relationship could
be determined for the very limited data that were available. These
correction factors represent the general effects of ambient temperature
and relative humidity on NOX emissions, and will be used to adjust
measured NO emissions during any performance test to determine compli-
A
ance with the numerical emission limit.
9-86
-------
TABLE 9-11 EXISTING INTERNAL COMBUSTION ENGINE AMBIENT CORRECTION
FACTORS FOR APPLICATION TO LARGE-BORE ENGINES
Fuel
Correction Factor
Diesel and
Dual-Fuel
+ 0.00235(H - 75) + 0.00220 (T - 85))
Gas
K - (KH) (KT)
KH» 0.844 + 0.151 () +0.075
- (T - 85)(0.0135)
where:
H • observed humidity, grains f^O/lb dry air
T a observed inlet air temperature, °F
Since the recommended factors may not be applicable to certain
engine models, as an alternative to the use of these correction factors,
engine manufacturers, owners, or operators may elect to develop their
own ambient correction factors. All such correction factors, however,
must be substantiated with data and then approved for use by EPA in
determining compliance with the NOX emission limits. The ambient
correction factor will be applied to all performance tests, not only
those in which the use of such factors would reduce measured emission
levels.
9-87
-------
As discussed In the Standards Support and Environmental Impact
Statement, "Proposed Standards of Performance for Stationary Gas Turbines,"
EPA-450/2-77-017a, the contribution to N0¥ emissions by the conversion of
/\
fuel-bound nitrogen in heavy fuel to N0¥ can be significant for stationary
A
gas turbines. The organic NOV contribution to total gas turbine NO
« x
emission is complicated by the fact that the percentage of fuel-bound
nitrogen converted to NOX emissions varies with the fuel-bound nitrogen
level. Figure 9-18 illustrates the variation in conversion of fuel-
bound nitrogen to NOX emissions with the fuel-bound nitrogen level of
the fuel. While this figure is based on very limited data, it indicates
that the percentage of fuel-bound nitrogen converted to NOY emission
A
decreases as the fuel-bound nitrogen level increases. Below a fuel-
bound nitrogen level of about 0.05 percent, essentially 100 percent of
the fuel-bound nitrogen is converted to NOV. Above a fuel-bound nitro-
A
gen level of about 0.4 percent, only about 40 percent is converted to
NOV. Using Figure 9-18, an estimate of the effect on controlled NO
* x
emission levels of firing fuels with various fuel-bound nitrogen levels
can be made.
As discussed in the Standards Support and Environmental Impact
Statement, "Proposed Standards of Performance for Stationary Gas
Turbines," EPA-450/2-77-017a, assuming a fuel with 0.25 weight percent
fuel-bound nitrogen (which allows approximately 50 percent availability
of domestic heavy fuel oil), controlled NOX emission would increase by
about 50 ppm due to the contribution to N0¥ emissions of fuel-bound ni-
A
trogen. In gas turbines, this contribution was significant when compared
to the proposed emission limit of 75 ppm. It can be assumed that the
9-1
-------An error occurred while trying to OCR this image.
-------
conversion of fuel-bound nitrogen to NO In large internal combustion
/\
engines is similar to that in gas turbines. Specifically, in both
facilities, fuel and air combustion occurs at high temperatures and at
similar sto1ch1ometr1c conditions. Based on this assumption, the
contribution of fuel-bound nitrogen to NOX emissions (I.e., 50 ppm), is
likely to be very small (I.e., approximately 10 percent). Given that
the emission limits have been rounded upward to the nearest 100 ppm and
the potential contribution of fuel-bound nitrogen to NOX emissions is
very small, no allowance has been included for the fuel-bound nitrogen
content of the fuel in determining compliance with the standards of
performance.
9.7 SELECTION OF COMPLIANCE TIME FRAME
Manufacturers of large-bore 1C engines are generally committed to
a particular design approach and, therefore, conduct extensive research,
development, and prototype testing before releasing a new engine model
for sale. Consequently, these manufacturers will require some period
of time to modify or reoptimize and test 1C engines to meet standards
of performance. As discussed in Section 6.3, the estimated time span
between the decision by a manufacturer to control NOX emissions from an
engine model and start of production of the first controlled engine is
about 15 months for any of the four demonstrated emission control tech-
niques. With their present facilities, however, testing can typically
be conducted on only two to three engine models at a time. Since most
manufacturers produce a number of engine models, additional time is
required before standards of performance become effective. In addition,
a number of manufacturers produce their most popular engine models at
9-90
-------
a fairly steady rate of production and satisfy fluctuating demand from
Inventory. Consequently, additional time Is necessary to allow manufac-
turers to sell their current Inventory of uncontrolled 1C engines
before they must comply with standards of performance.
It 1s estimated that about 30 months delay 1n the effective date
of the standard 1s appropriate to allow manufacturers time to comply
with the proposed standards. In addition, 1n light of the stringency
of the standards (I.e., many models will have to reduce NO emissions
A
by more than 40 percent), this time period provides the flexibility of
manufacturers to develop and use combinations of the control techniques
upon which the standards are based or other control techniques. Con-
sequently, 30 months from the date of proposal 1s selected as the delay
period for Implementation of these standards on large stationary Inter-
nal combustion engines.
9.8 MODIFICATION/RECONSTRUCTION
A discussion of the modification and reconstruction regulations
and how they pertain to the Internal combustion engine Industry can be
found 1n Chapter 5. Since few modified or reconstructed Internal
combustion engines are anticipated, the modification and reconstruction
regulations will have little impact. The demonstrated NO reduction
n
techniques, however, are as effective in reducing emissions of NO from
A
modified or reconstructed Internal combustion as from new Internal
combustion engines. Thus, modified or reconstructed internal combustion
engines merit no special allowance in the standards of performance.
9.9 SELECTION OF PERFORMANCE TEST METHOD
A performance test method 1s required to determine whether an
engine complies with the standards of performance. Reference Method
9-91
-------
20, "Determination of Nitrogen Oxides, Sulfur Dioxide, and Oxygen
emissions from Stationary Gas Turbines," 1s proposed as the performance
test method for 1C engines. Reference Method 20 has been shown to
provide valid results. Consequently, 1t was decided that rather than
a totally new reference test method being developed, Reference Method
20 would be modified for use on 1C engines.
The changes and additions to Reference Method 20 required to make
1t applicable for testing of Internal combustion engines include (by
section):
1. Principle and Applicability. Sulfur dioxide measurements
are not applicable for internal combustion engine testing.
6.1 Selection of a sampling site and the minimum number of
traverse points.
6.11 Select a sampling site located at least five stack diameters
downstream of any turbocharger exhaust, crossover junction, or recircu-
lation take-offs and upstream of any dilution air inlet. Locate the
sample site one meter or three stack diameters (whichever 1s less) upstream
of the gas discharge to the atmosphere.
6.1.3 A preliminary 02 traverse is not necessary.
6.2 Cross-sectional layout and location of traverse points
use a minimum of 3 sample points located at positions of 16.7, 50 and
83.3 percent of the stack diameter.
6.3.1.4 Record the data required on the engine operation record
on Figure 20.6 of Reference Method 20. In addition, record (a) the
intake manifold pressure; (b) the intake manifold temperature; (c) rack
position; (d) engine speed; and (e) Injector or spark fuming. (The
water or steam injection rate 1s not applicable to internal combustion
engines.)
9-92
-------
NO emissions measured by Reference Method 20 will be affected by
A
ambient atmospheric conditions. Consequently, measured NOX emissions
would be adjusted during any performance test by the ambient condition
correction factors discussed earlier, or by custom correction factors
approved for use by EPA.
The performance test may be performed either by the manufacturer
or at the actual user operating site. If the test is performed at the
manufacturer's facility, compliance with that performance test will be
sufficient proof of compliance by the user as long as the engine operating
parameters are not varied during user operation from the settings under
which testing was done.
9.10 SELECTION OF MONITORING REQUIREMENTS
To provide a means for enforcement personnel to ensure that an
emission control system installed to comply with standards of performance
is properly operated and maintained, monitoring requirements are generally
included in standards of performance. For stationary internal combustion
engines, the most straightforward means of ensuring proper operation
and maintenance would be to monitor NOX emissions released to the
atmosphere.
Installed costs, however, for continuous NOX monitors are approxi-
mately $25,000.(32^ Thus, the cost of continuous NOX emission monitoring
is considered unreasonable for 1C engines since most large stationary
1C engines only cost from $50,000 to $3,000,000 (I.e., 1000 hp gas
production engine and 20,000 hp electrical generation engine).
A more simple and less costly method of monitoring is measuring
various engine operating parameters related to NOX emissions. Conse-
9-93
-------
quently, monitoring of exhaust gas temperature was considered since
this parameter could be measured just after the combustion process
where NOX 1s formed. However, a thorough Investigation of this approach
showed no simple correlation between NOX emissions and exhaust gas
temperature.
A qualitative estimate of NOX emissions, however, can be developed
by measuring several engine operating parameters simultaneously, such
as spark Ignition or fuel Injector timing, engine speed, and a number
of other parameters. ' These parameters are typically measured at
most Installations and thus should not Impose an additional cost Im-
pact. For these reasons, the emission monitoring requirements Included
1n the proposed standards of performance require monitoring various
engine operating parameters.
For dlesel and dual-fuel engines, the engine parameters to be
monitored are: (1) Intake manifold temperature; (2) Intake manifold
pressure; (3) rack position; (4) fuel Injector timing; and (5) engine
speed. Gas engines would require monitoring of: (1) Intake manifold
temperature; (2) Intake manifold pressure; (3) fuel header pressure;
(4) spark timing; and (5) engine speed.
Another parameter that could be monitored for gas engines 1s the fuel
heat value, since 1t can affect NOX emissions significantly. Because
of the high costs of a fuel heating value monitor, and the fact that many
facilities can obtain the lower heating value directly from the gas
supplier, monitoring of this parameter would not be required.
The operating ranges for each parameter over which the engine
could operate and 1n which the engine could comply with the emission limit
9-94
-------
would be determined during the performance test. Once established, these
parameters would be monitored to ensure proper operation and maintenance of
the emission control techniques employed to comply with the standards of
performance.
For facilities having an operator present everyday, these oper-
ating parameters would be recorded dally. For remote facilities where an
operator 1s not present every day, these parameters would be recorded weekly.
The owner/operator would record the parameters, and 1f these parameters
Include values outside the operating ranges determined during the
performance test, a report would be submitted to the Administrator on a
quarterly basis Identifying these periods as excess emissions. Each
excess emission report would Include the operating ranges for each
parameter as determined during the performance test, the monitored
values for each parameter, and the ambient air conditions.
9-95
-------
REFERENCES FOR CHAPTER 9
(1) Bartok, W., et al. System Study of Nitrogen Oxide Control Methods for
Stationary sources, Final Report. Volume II. National Air Pollution
Control Administration. NTIS Report No. PB-192-789. 1969.
(2) McGowin, C.R. Stationary Internal Combustion Engines in the United
States. Environmental Protection Agency. EPA-R2-73-210. April 1973.
(3) Brown, R.A., H.B. Mason, and R.J. Schreiber. Systems Analysis
Requirements for Nitrogen Oxide Control of Stationary Sources.
Environmental Protection Agency. EPA-650/2-74-091. June 1974.
(4) Preliminary Environmental Assessment of the Application of Combustion
Modification Technology to Control Pollutant Emissions from Major
Stationary Combustion Sources. Volume II-Technical Result. Acurex
Corporation. Report TR-77-28. pp. 5-46 February 1977.
(5) Computer printout of Nationwide Emissions Report, National Emissions
Data System, as of January 10, 1975. Obtained from the Monitoring and
Data Analysis Division, OAQPS, U.S. Enviornmental Agency.
(6) Youngblood, S.B. and G.R. Offen. Acurex interoffice Memorandum.
Emissions Inventory of Currently Installed Stationary Reciprocating
Engines. September 23, 1975.
(7) Op. Cit., Reference 5.
(8) Hopper, T.6. and W.A. Marrone. Impact of New Source Performance
Standards on 1985 National Emissions from Stationary Sources. The
Research Corporation of New England. October 24, 1975.
(9) Habegger, L.J., et.al. Priorities and Procedures for Development of
Standards of Performance for New Stationary Sources of Atmospheric
Emissions. Argonne National Laboratory. May 1976.
(10) Air Quality Criteria for Hydrocarbons. National Air Pollution
Control Administration. AP-64. Washington, D.C. 1970.
(11) Goodwin, D.R., EPA/Emission Standards and Engineering Division.
Request for Information. June 16, 1976.
(12) Snyder, W.E. (Waukesha) and D.R. Goodwin (EPA) Private
Communication. July 20, 1976.
9-96
-------
(EPA)'
'1976: (GMC) ami D'R- S0°dw1n -
G00 the Files, Meeting Ronnn- .
Of I «r / O •
9-97
-------
APPENDIX A
EVOLUTION OF PROPOSED STANDARDS
This study to develop proposed standards of performance for new
stationary reciprocating internal combustion engines began on June 24,
1974. In March 1976, draft Standards Support Document and Environmental
Impact Statement (SSEIS) was presented to the National A1r Pollution Control
Techniques Advisory Committee (NAPCTAC). This draft had been prepared
by the Aerotherm Division of Acurex Corporation, Mountain View, California
under Contract 68-02-1318, Task No. 7. Randy Seiffert, Standards Support
Criteria Pollutants Section, Industrial Studies Branch, Emission Standards
a,nd Engineering Division (ESED), Office of Air Quality Planning and Standards
(OAQPS), U.S. EPA, was the Project Engineer. A revised draft SSEIS was
prepared by Acurex/Aerotherm under Contract 68-02-2530 beginning in October
1976, under the direction of Mr. Douglas Bell (Project Engineer) and
Mr. Fred Porter (Section Chief), Standards Development Branch, Regulations
Preparation Section Branch, ESED/OAQPS, U.S. EPA. Preliminary work for this
revision was accomplished under Contract 68-01-3158, Task 14, beginning in
July 1976, under the direction of Mr. John McDermon, Standards Development
Branch, ESED/OAQPS, U.S. EPA.
The first step 1n this investigation was to Initiate a two-part
literature survey, one part being directed toward technical questions
and the other toward a characterization of the Industry. The technically-
oriented survey sought Information on the best control technologies available
A-l
-------
for stationary engines. The bibliographical services of APTIC, the Engineering
Index, and the Indices of the Transactions of the Society of Automotive
Engineers (SAE) and of the American Society of Mechanical Engineers (ASME)
were studied, and appropriate papers obtained. As a result of this survey,
contacts were initiated with staff members at Southwest Research Institute
(SwRI), who had conducted a study for the American Gas Association (A6A)
on natural gas fueled engines used in pipeline compressor stations. SwRI
has also investigated emissions from a variety of engines used in mobile
applications, ranging from small gasoline units to large diesel locomotive
engines. The Bartlesville Energy Research Center of the Bureau of Mines,
U.S. Department of Interior, was found to have conducted a significant
amount of R&D work on emissions reductions from engines, mainly 100- to
500-hp diesel and gasoline units. In addition to these studies, most
published research was concerned with emissions from automobile engines.
Business indices, such as Moody's and Dunn and Bradstreet, were
consulted to identify the domestic manufacturers of engines for stationary
applications. However, the trade journals were found to be much more helpful
for this purpose.-7 These included the Oil and Gas Journal. Automotive
Engineering, and Power. In addition, the Diesel and Gas Turbine Worldwide
Catalog was especially useful^). Other sources were Federal Power Commission
(FPC) publications, which contained information about engine-driven electric
generators, various AGA documents, which gave information on the gas pipeline
manufacturers.
A-2
-------
industry's use of engines, and statistical surveys by the U.S. Department
of Agriculture, which provided some data on agricultural pumping requirements.
Some annual production statistics were obtained from the Commerce Department's
series of Current Industrial Reports, but frequently these documents did
not divide production figures into sufficient subgroups to correlate a
production trend with a particular application. Therefore, the majority
of the production, marketing, and application data were obtained by direct
telephone or written contacts with marketing representatives, application
engineers, or environmental managers within companies who sell engines
for stationary application.
Direct contacts with engine manufacturers also provided much of
the data on emissions from uncontrolled engines, control technologies,
and emissions from controlled engines. In addition, a report by McGowin(2)
provided useful information about large-bore engine applications and utili-
zation rates and one by Roessler, et al.,(3) presented a thorough discussion
of control technologies used in mobile applications. Additional information
regarding stationary applications, control technologies, and their costs
were received from manufacturers in response to official requests for
data by the Director, Emission Standards and Engineering Division, U.S.
EPA. These requests were sent to manufacturers under the authority specified
in Section 114 of the Clean Air Act. Most of the data on the effectiveness
of the various controls come from laboratory experiments at the manufacturer's
plants or, in a few cases, at special test centers such as at SwRI or
Bartlesville.
An extensive telephone survey was conducted during July and August
1974 among local and state air pollution control authorities (see Table A-l)
in an attempt to locate potential examples of "best demonstrated control
A-3
-------
Table A-l. AIR POLLUTION CONTROL AGENCIES CONTACTED IN
SEARCH OF CONTROLLED STATIONARY ENGINES
State
Arizona
California
Colorado
Florida
Illinois
Missouri
New Jersey
New York
Oklahoma
Oregon
Pennsylvania
Texas
Office Contacted*
1. Department of Health, Division of Air Pollution
Control
1. Bay Area APCD
2. Kern County APCD
3. Los Angeles APCD
4. San Bernardino APCD
5. San Diego APCD
1. Department of Health
2. City and County of Denver Health Department
3. Tri-County Health Department
1. Department of Pollution Control: A1r, Noise, and
Solid Waste
2. Dade County
1, Illinois EPA - Air Pollution (Chicago)
1. St. Louis APCD
1. Department of Environmental Protection (Trenton
Central Office)
1. Department of Environmental Conservation
2. New York City Department of Air Resources
1. Oklahoma City - County Health Department
2. Tulsa City - County Health Department
1. Department of Environmental Quality
2. Lane County Air Pollution Authority
3. Mid-Willamette Valley Air Pollution Authority
1. Alleghany County Air Pollution Control
1. Air Control Board -Austin
2. Air Control Board - Houston
aAPCD-Air Pollution Control District
A-4
-------
technology." Based on this survey, which was planned to include all the
geographical areas known to have relatively stringent regulations, it
was concluded that no installed engines were currrently subjected to any
emission limitations other than for visible emissions. The only exception
was the 140 Ib/hr rule for NOX from new engines in Southern California
(see Section 2.2). Since most uncontrolled engines under 4000 hp can
meet this regulation, and since no larger engines are known to have been
installed in that area recently, that rule did not lead us to any examples
of field-installed controlled engines. Therefore, the evaluation of
control technology was to be based upon published emission data and data
obtained from engine manufacturers.
Since no installed engines were found which could be cited as
examples of units with "best" controls, no emission tests were conducted.
However, a San Jose, California sewage plant (San Jose Water Pollution
Control Plant) was visited to give a better understanding of the operational
flexibilities available to users of large-bore engines.(4) This plant
had six gas engines and six dual-fuel units.
The direct contacts were supplemented by visits to the manufacturers
listed in Table A-2. The purpose of these visits was to obtain more information
concerning the status of their R&D efforts in emission reductions, their
experience with the commonly proposed control technologies, their estimates
of the cost and time required to incorporate such controls on their engines,
and the importance of the stationary market place in their total sales.
The group meetings that were held with representatives of engine
manufacturer's associations are listed in Table A-3. In general, the
meetings served to inform these representatives of the purpose of performance
standards for new sources, the process of developing a standard, including
A-5
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
informal and formal opportunities for input from all interested parties,
and the reasons for considering that stationary engines should be regulated
by performance standards. The meetings were also used to request information
of a general nature, such as data on the effects of variable ambient condi-
tions (temperature, humidity, etc.) on emissions, appropriate test cycles
(based on their knowledge of engine usage patterns) possible engine operating
parameters that are known to be directly and uniquely related to emission
levels and could be used in lieu of emission testing for compliance monitor-
ing. As is also shown in Table A-3, representatives of three industry
organizations met with EPA and commented on the draft SSEIS after it was
presented at the March 1976 NAPCTAC meeting in Washington, D.C.
Several meetings were held between Acurex/Aerotherm and EPA staff
during the development of the draft document. In the first half of 1978,
two public meetings were held to discuss the proposed standard?, one before
the NAPCTAC committee and the other with the engine manufacturers. These
are listed 1n Table A-4 along with a brief description of the purpose, con-
clusion, and any redirections.
In early 1978, Acurex was assigned the responsibility for writing the
rationale chapter of the SSEIS, the preamble to the standard, and the regula-
tion itself. In addition, Acurex was also assigned responsibility for es-
tablishing monitoring and emission testing procedures. This work was per-
formed with the advice and review of EPA/OAQPS.
Review of the original draft Standards Support Document and Environ-
mental Impact Statement by the Industrial Studies Branch, Emission Standards
and Engineering Division, OAQPS was completed on December 20, 1975, and
the final draft was forwarded to the Industrial Studies Branch by Acurex/
Aerotherm on March 17, 1976. The revised draft SSEIS was reviewed by the
A-8
-------
Standards Development Branch, Emission Standards and Engineering Division,
OAQPS by March 1978 and a final SSEIS was forwarded to the Standards
Development Branch by Acurex/Aerotherm in July 1978.
A-9
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
REFERENCES FOR APPENDIX A
(1) Diesel and Gas Turbine Worldwide Catalog. Diesel and Gas Turbine
Progress. Milwaukee, Wisconsin. 39. 1974.
(2) McGowin, C. G., Stationary Internal Combustion Engines in the United
States. EPA-R2-73-210. April 1973.
(3) Roessler, W. V., A. Muraszew, and R. D. Kopa. Assessment of the Appli-
cability of Automotive Emission Control Technology to Stationary Engines
EPA-650/2-74-051. July 1974.
(4) Offen, G. R. (Acurex/Aerotherm). Trip Report. San Jose Water Pollution
Control Plant. (Interoffice Memorandum to F. E. Moreno, Acurex/Aerotherm)
June 5, 1975. '
A-15
-------
APPENDIX B
Agency Guidelines for Preparing
Regulatory Action Environmental
Impact Statements (39 FR 37419)
Location Within the Standards
Support and Environmental
Impact Statement
1. Background and Description of
Proposed Action
Summary of Proposed Standards
Statutory Basis for the Standard
Facility Affected
Process Affected
Availability of Control
Existing Regulations at State
or Local Level
2. Alternatives to the Proposed
Action
Environmental Impacts
Costs
The standards are summarized in
Chapter 1.
The statutory basis for the standard
is given in Chapter 2.
A description of the facility to be
affected is given in Chapter 3.
A description of the process to be
affected is given in Chapter 3.
Information on the availability of
control technology is given in
Chapter 4.
A discussion of existing regulations
on the industry to be affected by
the standard is included in Chapter 3,
Section 2.3.
Environmental effects of delaying
the standards are discussed in
Chapter 7, Section 6.
The costs of alternative control
techniques are discussed in Chapter
8, Sections 2, 3, and 4.
B-l
-------
Agency Guidelines for Preparing
Regulatory Action Environmental
Impact Statements (39 FR 37419)
Location Within The Standards
Support and Environmental
Impact Statement .
3. Environmental Impact of
Proposed Action
A1r Pollution
Water Pollution
Solid Waste Disposal
Energy
Other
The air pollution Impact of the
standards 1s considered 1n Chapter 7,
Section 1.
The water pollution Impact of the
standards 1s discussed 1n Chapter 7,
Section 2.
The solid waste disposal Impact of
the standards 1s discussed 1n
Chapter 7, Section 3.
The energy Impact of the standards
1s considered 1n Chapter 7,
Section 4.
The environmental Impacts related
to noise and thermal pollution
are discussed 1n Chapter 7,
Section 5.
B-2
-------
APPENDIX C
EMISSION SOURCE TEST DATA
C.I DATA
This appendix tabulates all the quantitative emission data that were
obtained for large-bore engines during the preparation of the Standard Support
and Environmental Impact Statement. These data include emission and fuel
consumption rates for both baseline and controlled conditions. Since no controlled
stationary reciprocating engines are known to exist in field installations,
all the data were obtained from engine manufacturers, who reported on the
results of experiments in their laboratories. In some cases, these data have
been made public, but in most instances they were received directly from the
manufacturer as a private communication.
Large stationary engines are expensive to purchase and operate,
and controlling their emissions has not previously received high priority.
For these reasons, virtually no government, commercial (other than manufacturers),
or university test labs have obtained one or reported on possible emission reduction
technologies that might apply to currently produced engines.l/ However,
Southwest Research Institute has a two-cylinder version of an Electromotive
Division (EN)) locomotive engine which they have used for emission control
c-i
-------
investigations(l). This engine is no longer being produced by EMD, although
it represents most of the current locomotive engine population. According
to data supplied by EMD, this model's successor has higher baseline NOX emissions
(14.7 g/hp-hr for the newer model compared to 9.1 g/hp-hr for the older two-
cylinder version at rated conditions) and lower total hydrocarbon emissions
(0.49 g/hp-hr compared to 0.92 g/hp-hr). These differences are consistent
with the increased efficiency of the engine. New EMD data (Engines 17, 18
and 19) showed the same trends with the application of controls that the SwRI
engine data did; consequently, SwRI's results are not included here.
The data include engines which are currently marketed for stationary
application and also several units used in mobile and marine applications.
However, data from installed engines were excluded because these older units
are not the same as those new units that would have to meet any promulgated
standards.
NOX emissions were measured using one of the four procedures and
several kinds of NOx analyzers. These procedures and instruments are
identified and discussed in Section 4.2 for each large-bore engine
manufacturer that reported emissions data. NOx emissions were measured by
either a nondispersive analyzer or a chemiluminescent instrument with a thermal
reactor to convert the N02 to NO. Results were reported as grams N02-
Nondispersi ve infrared analyzers (NDIR) were used to measure the CO
concentrations. Since the exhaust was "dry," interference due to water was
negligible. When C02 and oxygen were measured, an NDIR analyzer was used for
the C02 and a paramagnetic analyzer for the oxygen. These constituents were
generally sampled to check the other measurements or to calculate total
exhaust flowrate if the inlet air flowrate into the engine was not measured.
C-2
-------
Total hydrocarbon emissions were measured in a heated flame ionization
detector (FID). The tubing and FID were maintained at 3750F to prevent
condensation, and hence, removal of any of the heavier hydrocarbons.
Therefore, these measurements were made on a "wet" basis because the water
had not been removed. The results were corrected for the water content to
make them correspond with the NOx and CO data. Since the exhaust from gas
engines does not contain these easily condensible heavy hydrocarbons, the
lines and FID were not heated in seme tests on gas engines. In addition to
these total hydrocarbon measurements, a few manufacturers measured the
nonmethane emissions from gas and dual-fuel engines. Their methods and
results are discussed in Appendix C.4.
Four engine manufacturers (Colt, Cooper, DeLaval, and White Superior),
who are members of the Diesel Engine Manufacturers Association (DEMA),
measured emissions data according to the measurement procedure published by
DEMA(2). Although the DEMA methodology suggests that a three-mode cycle be
used for engines that will be used at constant speed (essentially all
stationary applications) and a six-mode cycle for units that will see
variable speed duty (marine and locomotive), most of the results reported
here were collected prior to the final publication of the DEMA procedures.
Hence, emissions were usually measured only at rated load and speed. However,
some engines were also tested at selected part load and speed points. These
additional data points appear on the following tables (see Tables C-l through
C-82) as derated operating points or as variations in the engine operating
speed.
Several of the engines (Engine Nos. 30 through 33) at the low
horsepower end of the large engine category were tested according to the
C-3
-------
California 13-mode cycle. However, only the emissions at rated conditions
were used here since the 13-mode automotive cycle does not represent
stationary engine usage, and no relationship was found between emission rates
based on the average of the California 13-mode cycle and those at rated
conditions.
The data were submitted in either tabular or graphical form. In
the latter case the data points were clearly marked and could be converted
readily to the tabular format for Inclusion in this appendix. Other
operational data were either supplied by the manufacturer or obtained from
product Information literature. Fuel consumption values 1n all cases are
specific to the test that measured the emissions and are not from general
product specifications.
Fuel consumption was generally reported 1n terms of Ib/hp-hr for
dlesels and Btu/hp-hr for gas and dual-fuel engines. Conversion to kcal/hp-
hr was accomplished using a lower heating value of 18,320 Btu/lb for dlesel
fuel and a units conversion of 0.252 kcal/Btu. All data presented in this
document on emissions from diesel and dual-fuel engines are based on operation
with No. 2 distillate (diesel oil).
In several cases, a number of tests were performed on the same engine,
and consequently an uncontrolled value was reported for each control
technique. In all cases the uncontrolled level precedes any controlled data
reported for a particular test. In the tabulations which follows, a separate
table 1s used for each engine, and the data are printed in numerical order
based on engine code number. These data come from References 1 through 22.
These references are noted on the tables except where an engine manufacturer
has requested anonymity.
C-4
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
C.2 REVIEW OF AMBIENT CORRECTION FACTORS FOR APPLICATION TO LARGE-BORE
ENGINE NOX DATA
Ambient correction factors have been developed for automobile- and
truck-size gasoline spark ignition (SI) and compression ignition (CI)
engines(23). The following sections will discuss the application of the
existing SI factors to natural gas engines (C.2.1) and the existing CI
factors to large-bore diesel and dual-fuel engines (C.2.2). In addition, gas
turbine ambient correction factors will be examined for application to
large-bore engine emissions (C.2.3). As these sections will illustrate, no
satisfactory ambient temperature correction factor has been developed for any
size of SI internal combustion engines, and only one study has considered
ambient temperature corrections for CI engines. Therefore, Section C.2.4
will discuss an analytical approach to correct emissions for variations in
ambient temperature.
C.2.1 Ambient Correction Factors Developed for SI Engines
A survey of the literature showed that correction factors for
gasoline-fueled engines have been developed only for ambient humidity
variations(24). one of the studies also evaluated the effect of ambient
temperature and barometric pressure variations on exhaust emissions, but
found that engine-to-engine variations were too great to generalize a
correction factor for either temperature or pressure(25). These studies
will be briefly discussed in the following paragraphs, and then the ambient
humidity correction factors will be compared and evaluated for application to
the large-bore natural gas-fueled engines in this study.
The Automobile Manufacturer's Association contracted with the Ethyl
Corporation to conduct a study on the effect of ambient air humidity,
temperature, and pressure on heavy-duty (HD) gasoline engines^26). The
C-£
-------
results of this study for the effect of humidity later became the correction
factor designated 1n the Federal Register for HD gasoline eng1nes(27).
This correction factor was derived from emission measurements conducted
on seven engines 1n accordance with the Gasoline-Fueled, Heavy-Duty Engine
procedures 1n the 1970 Federal Register. The test engines were Installed
1n an engine dynamometer test cell where humidity, temperature, and pressure
were varied. The engines tested were gasoline-fueled, spark-ignited, carbureted,
heavy-duty truck engines ranging in size from 38 to 75 CID per cylinder.
The compression ratios varied from 7 to 9.4, and air-to-fuel ratios varied
from 14.4 to 15.7. The objective of the study was to develop factors to adjust
composite mass emissions to a standard condition of 75 grains H20/lb dry air,
90°F inlet air temperature, and a barometric pressure of 29.92 inches of mercury.
Figure C-l(28) illustrates the effect of humidity on both A/F ratio
and NOX emissions (Federal nine-mode composite cycle). As this figure
illustrates, a reasonably good correlation was established between changes in
ambient humidity and NOX emissions. Note also that A/F ratios are
essentially constant or decrease slightly with increasing inlet air humidity.
The composite cycle ambient correction factor was of the form
K = 0.634 + 0.00654(H) - 0.0000222(H)2 (C-l)
where NOX corrected = (K) NOX observed
H = specific humidity, grains H20/lb dry air
Figures C-2(29) and c-3(30) illustrate the effect of temperature and pressure
on A/F and NOX emissions for these same engines. Obviously, engine-to-engine
variations were too.great to generalize a correction factor for either temperature
or barometric pressure.
C-89
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
Both physical reasoning and these results indicate that changes in
ambient temperature and pressure affect the A/F ratio more than do changes in
humidity. Moreover, it is well known that the effect of A/F ratio variations
on NOX emissions depends on the engine operating point on the NOX vs. A/F
curve (see Figure C-4)(31). For example, a decrease in A/F ratio for an
engine operating at A in Figure C-4 would result in a decrease in NOX
emissions, whereas a decrease in A/F ratio for an engine at C would cause an
increase in NOX emissions. The author concluded, therefore, that engines
operating at different A/F ratios with different metering characteristics
should exhibit varying (possibly contradictory) effects on exhaust emissions
for similar changes in inlet air conditions.
The author also indicated that changes in inlet air conditions could
affect fuel distribution to the engine, and this in turn affects NOX
emissions. This effect is illustrated in Figure C-5(32). In addition,
changes in inlet air conditions which change the engine A/F ratio also change
the minimum spark advance for best torque (mbt). Figure C-6(33) illustrates
how NOX emissions vary with A/F for the mbt setting and other settings
retarded from mbt. Since some of the engines in that study had no vacuum
advances, they experienced, in effect, retarded mbt settings as their A/F
ratio changed. Thus, their emissions would respond differently to A/F changes
than engines with vacuum advances. Therefore, the development of a general
correction factor for ambient temperature and pressure was not possible, although
variations in these parameters affected emissions significantly.
In two other studies, ambient humidity corrections were developed from
LD gasoline vehicles(34>35). In the first study, correction factor was
derived from emission tests at 76°F and at four ambient humidities on a fleet
of eight passenger cars operated on a chassis dynamometer set to simulate
C-93
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
derived from emission tests at 76°F and at four ambient humidities on a fleet
of eight passenger cars operated on a chassis dynamometer set to simulate
seven different road loads of the Federal seven-mode composite cycle for LD
vehicles. These passenger cars were powered by gasoline-fueled, spark-
ignited, carbureted engines ranging in size from 42 to 59 CID per cylinder.
The compression ratios varied from 8.5 to 10.5, and air-to-fuel ratios from
14.6 to 16.4.
Corrections for ambient humidity were derived for both composite
cycle and constant load. The Federal Test Cycle constant load factor is
considered more applicable to large-bore carbureted SI engines which typically
operate at a constant load (nearly rated load). These factors are
K = 0.7% + 0.175(H/100) + 0.129(H/100}2
composite factor (C-ZaJ
K = 0.844 + 0.15KH/100) + 0.075(H/100)2
constant load, 50 mph (C-Zb)
The results of this study were adopted by California to correct emissions for
anbient humidity for gasoline-powered vehicles under 6000 pounds <36).
In the second study, a similar test program on gasoline vehicles was
conducted to develop an ambient humidity correction factor that was adopted
by EPA to correct LD gasoline vehciles(37>38) . This factor is
K = 1/(1 - 0.0047(H - 75)) (C-3)
In yet another study, an ambient correction factor for humidity was
developed for typical research engines where A/F and mbt settings were held
constant^39). The author of this study also noted the effect of A/F ratio on
NOX emissions as illustrated in Figure C-7(4°). Based on this figure,
C-96
-------An error occurred while trying to OCR this image.
-------
Krause(41) reasoned that composite correction factors would vary from
constant load factors since NOX emissions vs. humidity curves changed with
A/F ratio. Despite the validity of these observations, the correction factor
derived in this study is probably less applicable to large-bore engines than
those derived for gasoline vehicles, since in practice A/F ratios and spark
settings of large-bore engines are not necessarily held constant as they were
in this study (e.g., spark timing is fixed as load changes for some engines,
and varies with load for other engines).
Comparison of Existing Humidity Correction Factors for SI Engines
Based on the preceding discussion, three ambient humidity correction
factors are potentially applicable to large-bore, natural gas fueled engines,
particularly four-stroke, carbureted versions. These factors are summarized
in Table C-83(42>43>44'45). Figure C-8 is a comparison of the three factors
over a typical range of ambient humidities. Note that only one of the factors
plotted is a constant load factor (Equation (C-2b); the other three are based
on composite test cycles.
As Figure C-8 illustrates, there is a considerable difference in
correction depending on the study. The results from the EPA/Scott study show
the greatest sensitivity to ambient humidity variation, while the results for
HD gasoline engines show the least. This is not unexpected in view of the
previous discussion regarding the variation in engine responses to changes in
inlet conditions depending on A/F ratio, fuel metering and distribution, and
ignition (distribution operation) characteristics of different engines. The
correction based on constant load has been chosen as the most suitable
correction to be applied to large-bore SI engines, since these engines are
C-98
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
typically operated at constant load. Thus, this correction factor is most
applicable to carbureted (4-NA) natural gas engines.
Application of any of the correction factors in Figure C-8 to other
engine types (e.g., turbocharged units) is questionable due to major
differences in inlet air intake systems. For>example, data from the draft
SSEIS indicate that NOX reductions due to water induction are directly
proportional to water-to-fuel (W/F) rates, up to W/F ratios of 1. Moreover,
W/F rates due to ambient humidity are a function of A/F ratio, as Figure
C-9 illustrates. Based on this figure, the W/F rate of a carbureted natural
gas engine (trapped A/F « 17) is about 30 percent lower for a specific
humidity of 100, than the W/F ratio of a turbocharged engine with a trapped
A/F ratio of 25 (trapped A/F ratios for turbocharged SI engines typically
range from 20 to 25)(«,47). Note§ also> that the curves Qf constant numidUy
diverge with increasing A/F ratio. Therefore, it can be anticipated that NOV
n
emissions of engines with different A/F ratios will respond differently to
identical changes in ambient humidity. Thus, the application of the constant
load humidity correction factor (based on carbureted gasoline engines) to
other than carbureted, large-bore SI engines is questionable.
Similar conclusions can also be reasoned regarding the application of
ambient temperature and pressure correction factors to NOX emissions from
engines whose air intake as well as fuel systems differ substantially. As
discussed earlier, however, no factors have been developed for SI engines for
either temperature or pressure.
C.2.2 Ambient Correction Factors Developed for CI Engines
A survey of the literature established that two sources have reported
ambient correction factors for truck-size diesel engines(48,49). A study by
Krause, et al., was sponsored by the Automobile Manufacturer's Association
C-101
-------An error occurred while trying to OCR this image.
-------
and the Engine Manufacturer's Association and resulted in correction factors
for both temperature and humidity. These corrections were later adopted by
the EPA for HD diesel engines(5°), The other source reported corrections
for humidity only.
In Krause's study, a correction factor was derived for six different
engines run over a 13-mode cycle. Emission procedures used were similar to
"California Procedures for Diesel Engines in 1973 and Subsequent Model Year
Vehicles Over 6001 Pounds Gross Vehicle Weight." The heavy-duty truck
engines tested were diesel-fueled, compression-ignition engines ranging in
size from 69 to 155 CID. The compression ratios varied between 14.9 and
18.7. Inlet air conditions were controlled to one of 19 combinations of
humidity and temperature. The humidity conditions ranged from 35 to 125
grains ^O/lb dry air, and the temperature was varied between 70°F and 115°F.
Barometric pressure was controlled to 28.00 ±0.2 inch Hg at the air cleaner
inlet.
The engines in the study were of the following types:
• Four-stroke turbocharged, direct injection (4-TC)
• Four-stroke turbocharged, prechamber (4-TC, PC)
• Four-stroke naturally aspirated, MAN chamber (4-NA, MAN)
• Four-stroke naturally aspirated, (4-NA)
• Two-stroke blower scavenged (2-BS)
• Four-stroke turbocharged, aftercooled (4-TC, AC)
Unlike turbocharged large-bore engines, which are nearly always
aftercooled, only one of the turbocharged engines in this study was
aftercooled. That engine as well as the 4-NA and the 2-BS units were similar
in design to large-bore engines. A correction factor was developed for all
of the engines as well as for individual units. The factor was of the form
C-103
-------
K = 1/(A(H - 75) + B(T - 85))
where NO corrected = (K)x (NO observed) and A and B are related to A/F ratio
(load). The reference ambient conditions are 75 grains H20/lb dry air and
85°F. Figure C-lo(51) is a plot of the coefficients A and B for each engine as
a function of load (or A/F ratio).
Average values of A and B for all engines (as function of A/F) were
also determined. This result does not appear wholly justified given the
variations 1n response to ambient humidity and temperature exhibited by the
different engine types depicted in Figure C-10. Note that the 4-TC, AC unit
is significantly less sensitive to ambient temperature variations over the
load range than the other designs. Therefore, separate ambient correction
factors for 2-BS, 4-NA, and 4-TC, AC units from this study were used on the
corresponding large-bore designs instead of just one average value of all
engines. The coefficients A and B were determined for rated load conditions
using average rated load A/F ratios that were reported for the large-bore
engines.
The second source(52) which examined humidity effects reported
correction factor based on experimental tests of a 2-BS and a 4-NA engine.
Ambient humidity corrections from both of these sources (assuming
inlet temperature is held constant, hence, the B term of Krause's factor
drops out) are illustrated in Figure C-ll(53»54). The SI (gasoline) factors
discussed in Section C.2.1 are also plotted for comparison. Note that there
is little variation in correction factors for different diesel engine types
with the exception of the 2-BS unit from the CRC report. Moreover, the
diesel-fueled (CI) engines appear less sensitive to ambient humidity changes
C-104
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
than SI units, particularly for humidities less than 75 grains H20/lb dry
air.
Initially, one might expect the diesel units to be more responsive to
inlet humidity variations, since these units operate at higher A/F ratios
(approximately 25 to 40), and therefore, induct more ambient water than SI
engines which operate more nearly at stoichimetric ratios (*15 to 17,
depending on the fuel). Diesel engines, however, have a greater thermal
inertia than SI engines due to their higher trapped A/F ratios (as well as a
different combustion process). Apparently the higher thermal inertia in the
diesel units more than offsets their higher effective water induction rate;
thus, their NOX emissions are less sensitive to changes in ambient humidity
than SI engines. This explanation is corroborated by experiments which have
demonstrated that water injection, as a control technique, produces
significantly greater NOX reductions in SI units than it does in CI (diesel)
units (see Section 4.4.7 of the draft SSEIS).
The Krause study also investigated the effect of ambient temperature
on NOX emissions. Figure C-12(55) presents the correction factors that were
derived for engine types similar to those in the present study (humidity is
assumed constant, hence, the A term of Krause1s factor drops out). The
reference temperature is taken as 85<>F. This figure shows that the naturally
aspirated and blower-scavenged engines NOX emissions are more sensitive to
inlet air temperature changes than the aftercooled design. Since Krause's
study systematically examined the effect of both inlet temperature and
humidity for a number of CI engine types, his factors were selected for
application to similar large-bore engine types.
C-107
-------An error occurred while trying to OCR this image.
-------
C'2'3 Potential Application of Gas Turbine Ambient Correction Factors to
reciprocating ic Engines ~ ~— ——
Since no ambient correction factors have been developed specifically
for large-bore engines, all existing correction factors for internal
combustion engines were examined including those for gas turbines. There is
a considerable amount of technical literature on gas turbine ambient
correction factors as a result of the promulgation of emission standards for
aircraft engines and the proposed emission standards for stationary gas
turb1nes(56-6l)t Th1s Sect1on describes these various factors and their
potential application to the large-bore engine data 1n the present study.
Gas turbine combustion characteristics have some common features with
those in diesel engines. Both the gas turbine primary combustion zone and
the diesel combustion chamber can be characterized as a well-stirred reactor,
and both use similar distillate fuels or natural gas. On the other hand, the
gas turbine combustion is a steady-state, constant pressure process, whereas
the diesel is an unsteady, variable pressure combustion process.
Nevertheless, the similarities of the two systems warrant an investigation of
existing gas turbines correction factors.
Many investigators have developed ambient humidity correction factors
based on a model that relates NOX formation parameters of temperature,
pressure, equivalence ratio and residence time (using the kinetic rate
equations for NOX formation)(62). Humidity enters this model through its
effect on reaction (flame) temperature. Most researchers have shown that the
effect of humidity on NOX formation takes the form
NOX corrected/NOx observed = exp (K (Hobserved - Hreference))
C-109
-------
where Href = specific humidity at reference (standard) conditions
K = empirical constant that ranges from 14 to 30, generally
taken as 19(63)
Figure C-13(64'65) is a plot of this correction factor and shows that
it agrees reasonably well with the HD diesel (4-TC, AC) factor discussed in
Section C.2.2 (the gas turbine factor was adjusted to a reference humidity of
75 grains HaO/lb dry air to correspond with the reference humidity of the HD
factor).
Other sources have shown similar agreement of humidity effects on NOX
emissions of diesel engines and gas turbines as illustrated in Figure
C-14(66). The gas turbine humidity correction factor was derived from
empirical data based on water injection as a means of NOX control in gas
turbines. The ambient humidity was converted into an effective water-to-fuel
ratio by multiplying the ambient humidity loading by the near stoichiometric
A/F ratio existing at combustion. Then the empirically derived water
injection correction factor of Ambrose(67) was used to calculate the
percentage reduction in NOX. It is reasoned that the overall A/F ratio is
inappropriate since much of the water vapor in the inlet air never reaches
the primary combustion zone because it is vented for engine cooling or enters
downstream as dilution or wall cooling air. Based on this result, one can
conclude that changes in humidity appear to affect NOX formation in gas turbines
in much the same way as diesel (1C) engines. However, it should be remembered
that a number of adequate humidity correction factors have been developed
for 1C engines; therefore, this gas turbine result is of limited value, but
serves to reinforce the more global application of these corrections.
C-110
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
Gas turbine temperature corrections were also examined for application
to reciprocating engine data. Figure C-15(68,69) 1s a comparison of the HD
diesel (4-TC, AC) correction factor with various gas turbine factors.
Obviously, there 1s little agreement with the exception of one factor.
Presumably these differences arise from differences in inlet air compressor
and combustor design. The gas turbine factors, however, have some
flexibility in that they are related to the combustor inlet parameters of
pressure and temperature. Therefore, an attempt was made to relate changes
1n NOX formation of reciprocating engines to changes in ambient temperature
by estimating the cylinder temperature before combustion in a typical
reciprocating engine and using the compressor pressure ratio and the
assumption of 1sentrop1c compression to calculate temperature. This
calculated temperature was then used with the gas turbine factors. Figure
C-16(70,71) presents the results of this approach for several different forms
of gas turbine factors. Again, these results do not correlate well with the
HD diesel factor, probably due to the empirical nature of the gas turbine
equations and the large differences in air intake systems between engines and
turbines.
On the basis of this brief review, gas turbine ambient correction
factors do not make a useful contribution as potential ambient correction
factors for reciprocating engines.
C>2>4 L."i*"f.]j!*1cal Approach for th«» Ambient Temperature Correction of NO,
emissions "
Since no temperature correction factor has been reported In the
literature (see Section C.2.1) for SI engines, and no systematic emissions
data exist (either SI or CI) from which to base a temperature correction for
large-bore engines In general, an attempt was made to develop a
C-113
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
semi analytical approach for relating changes 1n ambient temperature to changes
1n NOX level from large-bore engines. This approach was used to estimate how
changes In ambient temperature affect NOX levels.
The analytical approach 1s based on the fact that ambient temperature
changes affect NOX emissions by their direct effect on both fuel/air (F/A)
ratio and peak flame temperature. As the ambient temperature rises, the
inlet air becomes less dense. Since the air Intake volume 1s essentially
constant, the engine will Inject a smaller mass of air; that 1s, the F/A
ratio will Increase. This change 1n F/A ratio Is related to a change 1n NOX
level for both SI and CI engines.
The temperature of the air in a diesel engine or fuel/air mixture in
a dual-fuel or natural-gas engine after compression is correspondingly
changed by a change in ambient (inlet) air temperature. That is, an increase
in inlet air temperture results in an even greater charge temperature after
compression. This increase in charge temperature leads to a higher peak
flame temperature and, consequently, greater NOX levels. The increase in NOX
!evel due to an inlet air temperature increase is related to specific engine
design parameters such as F/A ratio, degree of aftercooling, and compression
ratio, as well as the fundamental nature of the combustion process (i.e., CI
or SI). Therefore, the relationship between NOX level and inlet air
temperature is anticipated to be highly dependent on the particular engine
design.
The discussion above indicates that NOX level is primarily a function
of both F/A ratio and inlet air temperature. That is,
NOX = g(f,T)
C-116
-------
where f = F/A, fuel/air ratio
T = Ta, ambient inlet temperature
N = NOX, oxides of nitrogen
The subscript R denotes a reference ambient condition
Since we are interested in predicting changes in NOX level rather than
absolute NOX level we may mathematically represent a change in NOX to a
change in ambient temperature as
dN = (aN/3f)T df + (8N/aT)f di
or
(C-4)
dN/dT = aN/af df/dT + aN/gT
where the derivatives on the right side of Equation (C-4) remain to be
evaluated. The first partial derivative, 3N/3f, represents the change in NOX
emissions due to a change in F/A at constant aibient temperature, while the
second partial, 3N/3T, represents the change in NOX due to temperature
variations at constant F/A. This mathematical formulation can be portrayed
graphically as shown in Figure C-17. The diagram shows NOX vs. fuel-to-air
ratio, the downswing curves, and load vs. fuel-to-air ratio, the upsloping
curves, for a typical turbocharged diesel engine. The basic problem is to
find the change in NOX at constant load due to some ambient temperature
change. For example, assume the ambient temperature, T, is greater than the
reference temperature, TR. Starting at the uncorrected fuel-to-air ratio,
one locates Point A on the NOX production curve and Point B on the load
curve. Then moving over, at constant load, to tne rererence temperature load
curve, one locates Point C and, hence, the reference fuel-to-air ratio. Now,
C-117
-------An error occurred while trying to OCR this image.
-------
knowing the reference fuel-to-air ratio, one can move from Point D to Point
E on the reference temperature curve. Thus, the change in NOX production
resulting in this change of ambient temperature is indicated as ANO¥.
A *
The change in NOX level with F/A for a constant ambient temperature,
3N/3f, is presumed known for a given engine or engine type. This
relationship can be established frcm data that can be obtained in the
laboratory on a sample engine. In addition, it is assumed that the NOX vs.
F/A plots of Figure C-12 have similar slopes for different ambient
temperatures. Therefore, only NOX vs. F/A data at one ambient temperature
are required to evaluate the derivative.
The derivative df/dT, expresses the rate of change of F/A ratio with
a change in ambient temperature. Wu and McAulay have derived the following
expression relating F/A to ambient temperature (at constant load) for both
turbocharged and naturally aspirated engines*7*).
f - 'R (T/TR)B
and
(C-5)
B • (1 + n) - (1 - 61) n (k - 1/nK)
where k * Cp/Cv, ratio of specific heats
n • turbocharger adiabatic compressor efficiency
n = turbocharger exponent frcm equation TAn rc = constant where
0.5 < n < 1 depending on turbocharger compressor pressure
ratio
rc - turbocharger compressor pressure ratio
ei • (Tc - Tm)/(Tc - Ta)
C-119
-------
where Tc, Tm are compressor exit and intake manifold inlet temperature,
respectively. Then, differentiating Equation (C-5) with respect to inlet
temperature we can evaluate the term df/dT in Equation (C-4).
df/dT - B(fR/TR) (I/!' - B(fR/TR) (1 +
* B(fR/TR) (1 + (B - 1) (AT/TR)) (C-6)
Since B is always less than 2 and TR is 7QOF = 530°R, a nominal change in
ambient temperature of 25°F makes
(B - 1) T/TR < 1/20
Since this term is small compared to unity, we can approximate the derivative
by
df/dt a B (fR/TR)
Thus, the term from Equation (C-4) that predicts the change in NOX due to a
change in F/A is evaluated as
3N/af df/dt = (9N/9f) (B) (fR/TR) (C-7)
An expression is now required to relate the change in NOX level to a
change in ambient inlet air temperature, i.e., the partial derivative 3N/3T
of Equation (C-4). This dependence can be evaluated by first relating
changes in ambient temperature to changes in flame temperature using the
following relationship from Williams' Combustion Theory(73) suggested by
Wilson, Muir, and Pellicciotti^74).
C-120
-------
Tfl • TB + 1/Cp ((Q - L) YM/1 + Cp(Te8 - TB))/(1 + Yox/1) (C-8)
where Tfl = droplet diffusion flame temperature (75)
T« s 1sentrop1c compression temperature
CR s compression ratio of engine
TM s manifold air temperature, related to TA by degree of
turbocharglng and aftercoollng
TB = boiling point temperature of the fuel
Q = heat of combustion per unit mass of fuel
L = latent heat of vaporization of the fuel
YOX = ambient oxygen mass fraction
1 = stoichiometric oxygen-to-fuel ratio (by mass)
Cp = specific heat of the fuel/air mixture
Then, the change in flame temperature is related to a change in NOV
/\
production using the Arrhenius relation
dNOx/dt « exp(-K/RTf1) ; K/R = 123,000°R
If it is assumed that the rate of NOX production is independent of time, one
can readily integrate the Arrhenius equation to yield the following expression
for the ratio of NOX produced at a given ambient temperature, T, and a
corresponding flame temperature, Tfl, to that at the reference condition.
= exp (K(Tfl - TflR)/R(Tfl x Tf]R)) (c.9)
The partial derivative can then be approximated in finite difference fom by
3N/9T,N(1-NR/N)/(T-TR) (c
C-121
-------
Now by substituting Equations (C-7), (C-9), and (C-10) into Equation (C-4) we
obtain:
dN . Bf R 3N .
N
1 - exp
'lC]l
R Tfl xTf1R Jl
(C-ll)
T-TR
A preliminary evaluation of this expression for a turbocharged diesel
engine predicts a 1-percent change in NOX per W change in ambient inlet air
temperature. For a constant intercooler effectiveness and turbocharger
compressor efficiency, a 1-degree change in ambient temperature corresponds
to a 1-degree change in manifold inlet temperature. Therefore, this
analytical expression can be checked using NOX emissions data vs. manifold
air temperature (at constant load) for a turbocharged, diesel engine.
Emissions data for the manifold air cooling control (see Figure 4-34 of
the draft SSEIS) indicates that this technique produces a 0.1- to 0.3-percent
change in NOX per degree Fahrenheit for diesel engines. Emissions data
reported by Ingersoll-Rand (and from Figure 4-34 of the draft), show about a
1-percent change in NOX per 1°F change in manifold air temperature for
turbocharged SI engines^). Since NOX emissions from SI engines are more
responsive to changes in manifold air temperature, it appears that this
analytical approach overestimates the effect of inlet air temperature on NOX
emissions from this turbocharged CI engine. The assumption of a constant NOX
production rate with peak flame temperature may be more valid for SI engines
than CI engines owing to differences in the combustion processes.
C-122
-------
The approach outlined here has the potential to incorporate many
different parameters into a single correction factor. Additional terns could
be added for any other operating parameters that have a significant effect on
NOX emissions, either directly or by their effect on F/A or Tfl. Once the
appropriate relations are established, it may be possible to predict changes
in NOX levels as a result of changes in any one of these variables caused by
ambient variations.
These goals are somewhat ambitious; therefore, the most logical first
step is to gather available data taken at different ambient conditions and
apply the method to prove its validity. At this time, however, there is
insufficient information for either SI or CI engines to substantiate this
approach.
C-123
-------
C.3 REVIEW OF NOX MEASUREMENT METHODS
Previous studies have indicated that sampling instrumentation and
procedures significantly affect emission levels. Furthermore, no one
standard procedure has been adopted by all of the manufacturers of stationary
1C engines reporting emissions. That is, eight large-bore engine
manufacturers who reported emission data used either chemiluminescent (CL) or
nondispersive infrared (NDIR) and ultraviolet (NDUV) instruments and one of
four emission measurement procedures (SAE, EPA, EMD, or DEMA).-
Therefore, this section will present a detailed discussion of the instrument
and sample acquisition practices used by the engine manufacturers. The
purpose of this discussion is to identify possible variations in exhaust
emission levels attributable to particular measurement equipment and/or
procedures. Specifically, this section will:
, Establish which instruments, sampling trains, and procedures were
used by each manufacturer who reported NOX emissions
• Summarize the potential sources of uncertainty relating to each
measurement practice
• Evaluate the variability among manufactuers' emissions data due to
the following discussion will begin by illustrating typical measurement
uncertainties in present sampling practices. This analysis will show that
significant uncertainties in measurements can arise due to differences in both
instrument and sampling procedures. Therefore, both of these sources of error
will be discussed as they apply to the instruments and procedures used by the
2/Alc0j cott, Cooper, DeLaval, Electro Motive, Ingersoll-Rand, Waukesha,
and White Superior.
C-124
-------
eight large-bore engine manufacturers. Then these practices will be compared
and the uncertainty in each manufacturer's NOX measurements estimated.
c'3-1 Previous Studies of Exhaust Measurement Variation
Before evaluating each manufacturer's sampling practice, it is of
interest to examine the present "state-of-the-art" of internal combustion
engine exhaust measurements of NOX. Three recent studies have been conducted
to compare the measurements of NOX made by different laboratories from the
same emission source. In two of these studies, NOX measurements were made,
simultaneously, in the same laboratory using identical procedures. In the
other study, the same emission source was sent to each laboratory. These
studies illustrate the magnitude of emission variations attributable to
instruments and sampling procedures. They also indicate the reproducibility
of emissions measurements, both within a laboratory and among different
laboratories. The results of these studies will serve as a basis for
comparing potential data variations due to the measurement practices of the
eight large-bore engine manufacturers considered in the current study.
A series of cooperative emissions tests was conducted by the
Coordinating Research Council (CRC) to evaluate measurement methods used to
analyze diesel exhaust emisions from truck-size engines. In Phase III of
this program, six laboratories sent sampling teams to one location to make
simultaneous measurements (using NDIR analyzers) of a multicylinder
engine(77). The engine used in this study was a six-cylinder, 300-cubic-
inch, four-stroke, direct-injection diesel. The procedures that were used
during this test program to measure NO, CO, and C02 evolved into SAE
Recommended Practice J-177<78). In Phase IV of that program> ^
was circulated to 15 laboratories to: (1) verify that the generally good
C-125
-------
agreement of Phase III emission levels were the result of improved sampling
procedures, and (2) obtain NO/NOX data using CL analyzers^).
Table C-84(80>81) shows the results of these two cooperative tests.
Although the Phase III variations appear reasonable, the Phase IV results
indicate poor agreement among the laboratories. These larger uncertainties
were attributed to poor calibration procedures (span gases out of spec,
instruments not calibrated) and possibly some variation in engine
performance. In addition, it was noted that the average NO concentration
measured at rated load by the CL analyzers was approximately 23 percent lower
than that measured by NDIR.
A more recent cooperative test program conducted by the CRC evaluated
EPA's revised heavy-duty diesel engine NO/NOX measurement methods and
instrumentation^). Six participants made a series of NO/NOX measurements
on a multicylinder engine, simultaneously, and produced the range of
uncertainty shown below. A range is shown because the results were analyzed
in three different groupings: (1) all data, (2) those which remained after
eliminating questionable results from participants who encounter sampling
problems, and finally, (3) those which were left after excluding both
questionable data and results obtained from instruments with long sample transfer
times.
13-mode Standard Deviation Standard Deviation
(q/hp-hr) (q/hp-hr) (Mean,_%)
NO 8.03 - 8.21 0.42 - 0.29 5.9 - 3.5
NOX 8.05 - 8.16 0.30 - 0.17 3.8 2.1
All but one of the participants used a CL analyzer, and the results from the
one NDIR analyzer (a new reduced interference design with a NOX converter)
C-126
-------An error occurred while trying to OCR this image.
-------
were equivalent to those from the CL analyzers. Thus, based on simultaneous
tests using the same instruments, sampling practices, and emissions sources,
uncertainties in emission measurements among laboratories ranged from 3 to 6
percent for NO, and 2 to 4 percent for NOX. In addition, the repeatability
within a laboratory ranged from 2 to 7 percent for NO and 1 to 3 percent for
NOX.
In contrast to the above program, the eight large-bore engine
manufacturers used one of four test procedures (SAE, EPA, DEMA, or EMD) and
either CL or NDIR/NDUV instruments. Furthermore, the NDIR's were not reduced
interference designs. Hence, their data are likely to vary more due to
measurement practices than the truck engine results. The following
discussion will first examine the sources of such variations from differences
in both instruments and sampling practices. Then an attempt will be made to
suggest uncertainty bounds for the emisions data from each manufacturer.
C.3.2 Variations in NOv Emissio^ Related to the NOv Instrument
Since the development of commercial chemiluminescent analyzers,
(1971), various studies have been conducted to compare their operation with
the already established NDIR analyzer. All of these studies have shown the
NDIR analyzer to record consistently higher levels of NO than the CL analyzer
for a given source(83-85). Three of the large-bore engine manufacturers
reported NOX emissions using NDIR's and the other five used CL analyzers.
Table C-85 shows the NOX instruments used by each of the
manufacturers. Scott Research Laboratories made the emission measurements
for Ingersoll-Rand and White-Alco, since neither manufacturer presently owns
emission measurement equipment. Note that an electrochemical instrument was
used by Shell Oil Research in 1971 to measure NOX from one Cooper-Bessemer
C-128
-------An error occurred while trying to OCR this image.
-------
and one Ingersoll-Rand engine. None of the manufacturers has used this
instrument since then; therefore, no attempt will be made to correlate those
emissions with instrument differences. Only Waukesha and GMC/EMD continue to
measure NOX emissions with NDIR analyzers, and Waukesha has recently acquired
a CL analyzer. Thus, all but one manufacturer will be using CL instruments
in the future. In the following paragraphs, the principle of operation of
the NDIR and CL instruments and associated sources of error will be briefly
reviewed. Then the results of NDIR/CL comparisons will be summarized to seek
a method of expressing emissions on an "equivalent instrument" basis.
C.3.2.1 NDIR Instrumentation
The NDIR instrument was introduced nearly 20 years ago and has
continued to be used widely as a CO and 0)3 detector. In addition, it was
used extensively to detect NO, HC, and S02, but other methods are not
supplanting it for these species. Its principle of operation depends upon
absorption of infrared radiation by the gaseous sample. Figure C-18
illustrates a typical NDIR instrument. Built-in optical and gaseous filters
are used to produce a narrow infrared beam band width to compensate for
interference (absorption) by other constituents.
Despite these precautions, water vapor and C02, to some extent, may
cause positive interferences (high readings) even though refrigerant and
chemical driers are used to remove water vapor. Desiccants, however, have
been found to cause significant interferences as well as water vapor (see
later discussion and Section C.3.3). In addition to these problems, the
response of the NDIR instrument to the specie of interest is nonlinear in
some instrument designs, necessitating a carefully constructed calibration
curve using at least four, and preferably six to eight, calibration gases.
C-130
-------An error occurred while trying to OCR this image.
-------
C.3.2.2 Chemllumlnescent Analyzers
Chemiluminescent analyzers, 1n contrast to NDIR instruments, have been
developed only recently (1971) for source sampling. Nevertheless, the CL
instrument has gained increasing application for the measurement of NOX
(NO + N02) and NO. In this type of instrument ozone, 03, is reacted with
nitric oxide, NO, to produce a chemically excited state of N02*. which emits
light as it decays to stable N02. The intensity of this emitted light is
proportional to the NO concentration present in the sample. The analyzer
uses a photomultiplier to detect the light.
Figure C-19 is a schematic of a CL analyzer and illustrates the
reaction of ozone with NO in the reaction chamber of the instrument. The
Instrument is easier to calibrate than some NDIR's due to its linear response
to NO, and thus requires fewer calibration gases. Note that a N02 * NO
converter is depicted 1n Figure C-19. There are basically two types on the
market. One device consists of a stainless steel tube which is heated to
«1200°F and essentially converts all N02 in the sample to NO. The other
device catalytically converts N02 to NO. In this way just NO or both NO and
N02 can be measured, depending on whether the sample gas is fed directly to
the reaction chamber or passed through the converter first. The N02 level is
then deduced by subtracting the NO level from the total NO* reading.
Potential problems encountered in this instrument include quenching of
the excited N02 by other species, converter inefficiencies, and interferences
caused by chemiluminescence of other gases. In general, quenching (by C02 or
H20) is negligible in CL instruments; particularly if water is removed from
the sample before analysis and low pressure (vacuum) reaction chambers are
used to reduce quenching by C02. Nevertheless, some quenching problems have
been observed during measurement of fuel-rich automotive exhausts. Quenching
C-132
-------An error occurred while trying to OCR this image.
-------
effects, however, have not been observed during measurements on truck-size
diesels, whose exhausts are similar to those of large bore stationary engines
(i.e., *15 percent oxygen).
Converter problems are avoided by thermal conditioning of new
converters and by making regular checks of converter efficiency (>90
percent). Interferences are minimized by the instrument manufacturer through
the choice of spectral filters and photodetectors.
C.3.2.3 Sources of Instrument Error
Sources of error for each of the instruments listed in Table C-85 are
summarized in Table C-ee^86"94). Note that neither the quenching effect nor the
converter problems of the CL analyzer should occur when measuring large-bore
engine exhausts with a properly operated and maintained CL analyzer. Present
NDIR analyzers, on the other hand, are susceptible to errors due to
interference despite sample conditioning to remove water. This Interference,
combined with the relative ease of operation and accuracy of CL analyzers, has
led to an increasing preference of CL's for NOX measurements. As an example,
EPA's proposed revisions to the Heavy Duty Diesel and Gasoline Engine
Sampling Procedure specify CL analyzers. The following section summarizes
comparisons of CL and NDIR/NDUV measurements of NOX. These comparisons will
be the basis for expressing NDIR/NDUV measurements as equivalent CL levels.
C.3.2.4 Correlations of CL to NDIR/NDUV
As discussed above, NDIR analyzers are subject to water vapor
interference despite sample conditioning to remove water vapor. These
interferences cause the instrument to indicate higher NOX levels than are
actually present. A comprehensive study conducted by TRW (with Scott
C-134
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
C.3.3 Variations 1n NOv Levels Related to Sampling Procedures
In addition to the choice of instrumentation, differences in sampling
procedures can cause variations in reported emission levels. This was
clearly illustrated in Table C-84 for the Phase IV cooperative tests where 15
laboratories made measurements of an identical source. Since four basic
sampling procedures (DEMA, SAE, EMD, EPA) were used by the eight large-bore
engine manufactuers, some variation between their reported emission levels
can probably be attributed to differences in their sampling practices. The
sources of these uncertainties will be identified and the four sampling
procedures compared and evaluated in this section.
C.3.3.1 Sources of Sampling Error
The two major problems encountered in measuring engine exhausts are:
(1) chemical changes that occur during transfer of the sample to the
analyzer, and (2) error due to improper operation of the analyzer. Figure
C-20 is a simple schematic of an engine sampling system. In transferring the
exhaust gas sample to the analyzer, care must be taken to ensure that all of
the NOX (NO + N02) or NO (when only an NDIR is used) originating from the
engine exhaust reaches the analyzer. For these reasons sample lines are
heated to prevent condensation or kinetic conversion of constituents. High
sample flowrates (short sample residence times) are maintained to minimize
sample degradation during its transfer from the engine to the analyzer, and
water removal devices are employed to minimize instrument interferences from
water vapor. Similarly, it is important that the analyzer be calibrated and
all components, such as N02 + NO converters (on CL instruments), be
functioning properly. Therefore, adequate analyzer specifications and
calibration procedures are essential for accurate emission measurements.
C-138
-------An error occurred while trying to OCR this image.
-------
Table C-88(a)(101-107) presents a more detailed summary of sources of
sample transfer error. As this table Indicates, heated sampling lines
coupled with low sample residence times are essential to preserve the Initial
amounts of NO and N02 contained in the sample gas. Furthermore, leak checks
of the entire sampling system (particularly on the vacuum side) should always
be performed to assure that the sample gas reaches the analyzer undiluted.
Water removal 1s required with NDIR analyzers; therefore, care must be taken
to minimize the time available for NO to be converted to N02 which can then
be absorbed by the condensed water from the sample gas. Chemical driers
(deslccants) are an unacceptable water removal device since they promote the
N02 •»• NO reaction and absorb N02 (see Section C.3.2).
Table C-88(b)(108) summarizes Important analyzer related procedures
required for accurate measurements. Care must also be exercised in operating
emission measurement instruments. Frequent calibrations should be performed
using accurate, certified blends of calibration gases. Zero and span checks
also serve to Indicate potential instrument problems as well as necessary
gain adjustments. Span gases should be frequently cross-checked with
calibration gases or checked against a NBS standard since even certified
gases can be in error.
Some calibration curves for NDIR analyzers are nonlinear; therefore,
several calibration gases (minimum of six) should be used to calibrate the
instrument(109). These instruments also experience some hysterysis after
sitting unused; therefore, frequent calibration is necessary. Also,
calibration points should be curve-fit using a higher order polynomial. In
addition, efficiencies of NOX converters (on CL analyzers) should be checked
regularly. Finally, strip chart recordings of data are superior to visual
C-140
-------
TABLE C-88(a). SOURCES OF SAMPLING ERROR: THE SAMPLE TRANSFER
Sampling System
Source
Error
Correction
• Unheated
sampling line
• Long sample
residence times
0 Sample line
connections and
fittings
• Water removal
— Refrigerant
Chemical
drier (desiccant)
NO •*• NOg; N02 adsorbed
during water removal:
Observed 40-percent
loss of NO when sample
residence times were
35-40 sec (101)
Heated Line;
ad-
sorbed during water re-
moval : 6-7% NOy loss
for system response
times3 greater than
25 sec (102)
Cold Line: 12-percent
NOX loss for sample
residence time of
13 sec (103)
Leaks that dilute sample
gas: Observed 25-percent
loss of NOX due to pre-
fliter leak (104)
N02 adsorbed in condenser;
(105)
N02 ->• NO and drier "eats"
NOg. Negative or positive
errors with cold lines
(SAE procedure) depending
on how drier conditioned
(106, 107)
Heat line to 375°F
Use short sample line
and/or flowrates to
limit system response
to 15 sec or less.
Leak check system be-
fore testing
CL: Locate condenser
after converter.
NDIR: Maintain high
sample flow through
condenser, remove
H20 as it forms.
Do not use chemical
driers
System response time « sample residence time + Instrument response time
C-141
-------
TABLE C-88(b). SOURCES OF SAMPLING ERROR: INSTRUMENT RELATED
Source
• Instrument drift,
ozone shortage (CL),
plugged capillary
• Converter
malfunction (CL)
t Visual rather than
stripchart readings
t Calibration and
span gases
Instrumentation
Error
Low NO readings
A
Not all N02 converted to
NO, results in low NOX
reading also possible for
NO to exceed NOX levels
Analyzer meter only ac-
curate to ±3 percent.
Visual averaging less
consistent than chart
averaging (108)
Change of constituents
with time, or erroneous
certification
Correction
Calibrate instrument
frequently. Zero and
span before and after
each measurement for
all instrument ranges.
Instrument should meet
minimum performance
criteria.
Perform converter
checks regularly. Use
known standard to check
converter efficiency.
Use stripchart, aver-
age levels over an
interval for which
steady state condi-
tions exist.
Use certified gases of
specified blends.
Cross-check span and
calibration gases.
C-142
-------
analyzer meter readings since a permanent record is produced, and emission
levels can be averaged over a time interval more accurately and consistently.
C.3.3.2 Sampling Procedures: EPA, DEMA, SAE, EMD
Having completed this brief review of sources of sampling error, we
can examine each of the four measurement practices presently used by the
eight large-bore engine manufacturers. Table C-89 is a comparison of the
EPA, DEMA, SAE, and EMD emission measurement practices^10-113). The
information for the EPA procedures is based on the revised EPA Heavy Duty
Diesel Emissions Regulations as proposed in the Federal Register, Volume 41,
No. 101, May 24, 1976. This revised procedure requires CL analyzers for
NO/NOX measurement, establishes instrument and calibration specifications,
and defines sample transfer configurations and practices.
This comprehensive EPA regulation was recently verified through
cooperative testing by both manufacturers of mobile diesel engines and the
EPA-Ann Arbor Mobile Sources LaboratoryUH). The six participants in this
program (see Section C.3.1) made essentially equivalent NOX measurements
for a series of eight tests. Two participants, however, did experience a
small, but consistent, NO/NOX crossover (NO levels greater than NOX levels).
Nevertheless, the standard deviation of measured NOX levels was generally
small, ranging from 2 to 4 percent (of the mean level). Therefore, the EPA
practice will be the basis of comparison for the other measurement practices.
The potential sources of measurement uncertainty of the other three
procedures (DEMA, SAE, and EMD) are summarized in Table C-90. The sampling
trains for these three procedures, as well as the EPA setup, are illustrated
in Figure C-21. The DEMA procedure specifies a CL analyzer and the SAE/EMD
C-143
-------
TABLE C-89. COMPARISON OF SAMPLING PRACTICES
Test Procedure
DEMA: No time between modes specified
EPA: 5 minutes
SAE: 20 minutes
EMD: 30 minutes
DEMA: Data recorded and averaged over 10 minutes
EPA: Data recorded over last 2 minutes
SAE: Data recorded over last 5 minutes
EMD: Data recorded over last 3 minutes
Instrument Calibration - DEMA and EPA are CL; SAE 1s NDIR; EMD 1s NDIR/NDUV
DEMA: Analyzers calibrated semimonthly. No details of calibration given.
EPA- Check NOX converter once per week (must be at least 90-percent
efficient); leak check system, calibrate analyzer, check sample
line residence time, and quench check every 30 days
SAE: Calibrate monthly
EMD: Checked and calibrated monthly, semiannually and annually
DEMA: Number of calibration gases not specified
EPA- Span analyzer with calibration gases having nominal concentrations
of 30, 60, and 90 percent of full scale concentrations
SAE: Use calibration gases that are 25, 50, 75, and 100 percent of
instrument range used.
EMD: Use 4 calibration gases
DEMA- No specifications for calibration or span gas blends and dilutents
EPA- Blends and dilutents specified and accurate within 2 percent of
true concentration or traceable within 1 percent of NBS blends.
Span gas traceable to within 1 percent of calibration gas.
SAE: ±2 percent accuracy on gas analysis certification
EMD: ±2 percent accuracy blends and dilutents defined by EMD
DEMA: No analyzer specifications required
EPA: Specifications for response-time, precision, noise, zero ana
span drift, and linearity ,.,*,,. u ++-
SAE: Accuracy should be ±2 percent full-scale deflection or better
EMD: Same as SAE
C-144
-------
TABLE C-89. Concluded
~— •—+****— m
Sample Transfer
lt-!1 (S'S.'} probe* configuration not specified
number and s1ze of n°les In S.S. probe
l' Probe» no configuration specified
s.S. probe, configuration specified
DEMA: Sampling Une material not specified
a n ess stee1' teflon. or P^ven Inert
°r tefion
EPA? !
SAE: Unheated
END: Unheated
DEMA: No sample Une length specified
' 1nstrurent response
SAE: No length specified
EMD: No length specified
DEMA: Chemical driers can be used
EPA: Not allowed
SAE: Refrigerant and chemical drier specified
EMD: Same as SAE
,
SAE: Condenser and drier before analyzer
EMD: Same as SAE for NDIR. NDUV does not use wfr rernoval.
C-145
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
procedures require an NDIR Instrument. The EMD practice also utilizes an
NDUV analyzer, 1n series after the NDIR, to measure N02 In the sample.
Table C-90 clearly Indicates that there are several sources of
measurement uncertainty possible for the DEMA procedure, largely as a result
of its failure to define instrument and sample transfer practices. Its most
obvious shortcoming is the potential degradation of the NO and N02 in the
sample gas as it passes from the probe through an unheated sample line (of
unspecified length) and condenser before entering the CL analyzer. High
residence times would also promote NOX loss for this configuration. One
large-bore engine manufacturer demonstrated that the sample residence time
must be less than 3 seconds to prevent significant loss of N02 in the sample
(12-percent loss for a 13-second residence time) as it passed through a cold
sample line and condenser (see Table C-88).
Similarly, SAE and EMD practices can also lead to a loss of NOX as the
sample passes through an unheated sample line, a condenser, and a desiccant
before reaching the analyzer. As discussed previously, the desiccant
converts N02 to NO and absorbs N02.
Comparisons of the SAE sampling train with the proposed EPA train
indicated that SAE measured concentrations of NO are understated by 15 to 40
percent during the first few measurement modes of the federal 13-mode
composite cycle, then are equal in the seventh or eighth modes, and finally
are overstated by about 10 percent in the last modesd"). These results and
the study conducted by TRW suggest that the desiccant requires a period of
time to equilibrate. If a new drier were used before each 13-mode test
(permissible under present EPA regulations), NO levels would tend to be low
(since levels would be understated until the drier had equilibrated, possibly
not until the last, rated speed, low power modes were being measured). If
C-148
-------
the drier was not replaced, or if it was preconditioned, the NO levels would
tend to be overstated. This is probably the case for those large-bore engine
manufacturers who use driers, since their engines are typically operated
about an hour before they are stabilized and measurements are begun.
The EMD procedure is similar to the SAE procedure with the exception
that both NO and N02 are measured using NDIR and NDUV analyzers. Results
from the TRW study(H6), based on automotive exhausts> .^.^ ^
NDIR/NDUV NOX measurements were about 20 percent higher than CL readings (no
water removal for CL sample). Apparently, the positive biases (interference
and drier) more than offset potential NOX loss due to the cold sampling in
that study, m addition to the positive NDIR biases, NDUV analyzers have a
history of stability problems, usually manifested as a drifting (high)
reading. One study attributed this error to interference due to carbon
particle buildup^!?).
C.3.3.3 Summary and Conclusion of DEMA, SAE, and EMD Practices
In sugary, it appears that the DEMA practices will generally lead to
negative errors due to NOX loss in transporting the sample from the engine to
the analyzer. At this time the absolute uncertainty associated with this
practice is unknown relative to the EPA procedure. Limited results (see
Table C-87) suggest this error could be as much as -10 to -15 percent
depending on the amount of N02 in the sample (estimated to be 2 to 10 percent
of the total NOX) and the extent that NO is converted to N02 during the
sample transfer.
The positive bias of the SAE procedure gives a 12-percent error in the
NO reading. The EMD procedure, which measures both NO and N02 causes a 20-
percent error in the NOX reading. Nevertheless, both procedures could
C-149
-------
experience as much as a -40-percent error in the NO or NOX reading if the
chemical drier has not stabilized. Thus, the uncertainties in NOX
measurements for the SAE and EMD practices appear to be significantly greater
than the DEMA practice.
Finally, the EPA procedure defined for the heavy-duty diesel and
gasoline engines appears to have minimized the various sources of measurement
error and should give more accurate and consistent NOX readings. Therefore,
each of the manufacturer's sampling practices will be evaluated in the follow-
ing section by comparing it to the EPA procedure.
C.3.3.4 Comparison of Manufacturer's Measurement Uncertainty
Table C-91 summarizes each manufacturer's practice in terms of
procedures that could lead to measurement error. Note that three of the four
DEMA manufacturers using cold sample lines have located their water traps
near the analyzer instead of at the engine as is recommended In the DEMA
practice. Therefore, this setup will promote additional loss of NOX in the
exhaust sample. In addition, Colt and Waukesha have relatively high sample
residence times which will also promote NOX loss.
The potential errors in measurement resulting from these practices are
estimated in Table C-92U18-122) relative to the EPA procedure. (Blanks
appear in Table C-92 for items which do not apply to a particular
manufacturer.) These uncertainties are depicted for each manufacturer in
Figure C-22. Errors due to system leaks are not included in the figure since
this error cannot be generalized. Also, the uncertainty bands were
constructed assuming that errors of the same sign were additive. Note that
C-150
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
these uncertainties substantially exceed the 4-percent scatter in NOX levels
observed during verification tests of the EPA procedure.
NDIR and NDIR/NDUV instrument biases have been included in this figure
and are based on the TRW study(123K Note that the potential negative 40-
percent error of Table C-92 is not included in Figure C-22 since the
manufacturers generally operate the engine long enough to condition the drier
before measurements are taken. The data for Waukesha indicates that sampling
errors can be as large as instrument bias, but in the case of Cooper and
GMC/EMD, instrument biases predominate.
Conclusions
The above analyses of instruments and measurement practices suggest
the following conclusions about the reported data:
• The exhaust data which were reported using the DEMA, SAE, or EMD
practices should be banded by the appropriate uncertainty level.
In general, measurement uncertainties are expected to range from
-15 to +5 percent for DEMA data and -20 to +20 percent for the
SAE/EMD data.
• Waukesha's data contains the greatest uncertainty since: (1) only
NO was measured, (2) other significant sampling problems could
have been present, and (3) the amount of N02 in the exhaust was
unknown; thus correction of NO to NO + N02 would be speculative.
• The Ingersoll-Rand and White-Alco exhaust emissions data appear to
have the least measurement uncertainty (±5 percent) since their
sampling procedure was essentially identical to the EPA procedure.
C-154
-------
The previous analyses also lead one to conclude that the EPA heavy-
duty engine method should be preferred over any other method previously used
for sampling stationary reciprocating engines. Therefore, any test method
developed for these engines should follow the same basic sampling procedures
and utilize the same basic sampling system. In the mean time, the DEMA and
SAE/EMD sampling practices should include more carefully specified sampling
procedures, and the use of NDIR/NDUU instruments should be carefully re-
viewed because of their bias relative to CL instruments.
C-155
-------
C 4 THE EFFECT OF NOX CONTROL TECHNIQUES ON NONMETHANE HYDROCARBON
EMISSIONS
It is well known that the application of NOX controls to 1C engines
can cause the emissions of other criteria pollutants to increase. Therefore,
CO and HC emissions may require control as well as NOX levels. However,
attempts to regulate HC emissions from large-bore engines are not
straightforward since most data for these engines are in terms of total
hydrocarbon (THC), rather than reactive HC which is the criteria pollutant.
Moreover, the Office of General Council (OGC) has indicated that EPA
should establish, if feasible, standards for the criteria pollutant, NMHC, to
avoid legal challenges of a THC standard under Section 111 of the Clean Air
Act(124). The OGC added, however, that EPA could set standards on THC if
such an approach: (1) simplifies sampling procedures, and (2) allows the use
of existing data. If EPA follows this approach, they must show that a
standard for THC would cost manufacturers and users no more than an standard
for NMHC. The purpose of this memo is to investigate this issue.
Data from Section 4.4 indicated that, in general, the application of
N0x controls to large-bore 1C engines resulted in increases of 5 to 50 percent
in THC levels (see Figure 4-52). However, the THC levels reported for the
gas and dual-fuel engines contained both methane, (CH4> * noncriteria
pollutant, and NMHC, the regulated HC pollutant. (NMHC emissions are defined
here as all other hydrocarbons in the exhaust sample.) If NMHC levels change
differently than THC levels with the application of NOX controls, separate
control regulations may be required for NMHC levels. Therefore, the data for
THC and NMHC (where measured) from gas and dual-fuel engines are reexamined
, * • K MOV control techniques affect NMHC levels relative to
here to determine how NOx uu
THC levels.
C-156
-------
In the following sections, uncontrolled levels of THC and NMHC will be
compared for both new (laboratory) and installed (field engines). Then these
emissions will be evaluated after the application of NOX control techniques
that were reported in Section 4.4. In addition, the measurement techniques
for both NMHC and THC will be briefly reviewed. Based on this evaluation, a
recommendation will be presented concerning the separate control of NMHC
emissions.
C'4>1 Comparison of Controlled THC and NMHC Emissions from Gas and Dual-
t-uei Engines "~ —
Previous investigators have shown that NMHC levels are generally less
than 20 percent of the THC emissions from natural gas engines. Figure C-23
illustrates this conclusion using data reported by Southwest Research
Institute (SwRI), and two engine manufacturers, Colt and Cooper. The SwRI
data were from engines installed on gas transmission pipelines(125), whereas
the manufacturer's data were from new engines that were measured in
laboratories. All the THC data were measured with flame ionization detectors
(FID's). Nonmethane emissions were determined as the difference in THC and
methane levels. Cooper and Southwest measured methane emissions with a gas
chromatograph. Colt physically removed all nonmethane portions of the sample
and then measured the methane remaining with an FID.
Figure C-23 shows that THC levels for pipeline engines ranged from 1
to 9 g/hp-hr and were lower for four-cycle designs (1 to 3 g/hp-hr) compared
to two-cycle levels (3 to 9 g/hp-hr). THC emissions from the engines tested
by the two manufacturers were generally lower than those from the pipeline
engines for both two- and four-cycle designs.
Although the THC levels for the new gas engines were lower than those
for the installed engines, the NMHC levels were greater. As a consequence,
C-157
-------An error occurred while trying to OCR this image.
-------
a greater fraction of the THC emission from the new engines were NMHC as
compared to the pipeline units. On the average, NMHC emissions were only 6
percent of THC levels for the installed engines, but 23 percent for the new
engines. These differences may be related to the condition of the installed
engines which were tested as found. That is, older, installed gas engines
may have had more blowby of unburned fuel and have been operating off peak
combustion (i.e., requiring maintenance or an overhaul), thereby causing
greater emissions of unburned fuel (methane). Furthermore, lower peak
temperatures associated with less than optimum combustion may have resulted
in fewer emissions of partially oxidized fuel (i.e., reactive hydrocarbons or
NMHC).
Figure C-24 shows uncontrolled THC and NMHC levels for Colt and Cooper
dual fuel engines. (No measurements of installed dual-fuel engines were
available.) These data are similar to the limited data for manufacturer's
gas engines, indicating that NMHC levels from these engines range from 10 to
20 percent of THC levels.
In the following sections, the levels of NMHC and the ratios of
NMHC/THC for these gas and dual-fuel engines will be evaluated after the
application of NOX controls. These comparisons will show whether NMHC
levels respond differently to controls than to THC levels, and, therefore,
whether any standards would have to be written in terms of NMHC.
Ct4'2 The Effect of NQK Control Techniques on THC and NMHC Levels
Figures C-25 through C-30 illustrate how the application of NOX
control techniques affect THC and NMHC levels from gas and dual-fuel engines.
Note that data were reported for both pipeline engines and those tested in
manufacturers' laboratories. In general, these figures show that derate
C-159
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
and air-to-fuel ratio changes cause THC emissions to increase, but retard
(R), external EGR, and manifold air temperature (MAT) reductions do not
change the uncontrolled THC. Furthermore, for all of these techniques, NMHC
levels remained about the same percentage of THC as they were in the
uncontrolled engines. That is, NMHC emissions from both gas and dual-fuel
engines change in the same proportion as do THC levels, or remain in the same
proportion (where THC remained unchanged) with the application of NOX
controls.
Figure C-25 shows that, in general, derating caused THC to increase
less than 50 percent, although extreme amounts of derate (>25 percent) caused
uncontrolled THC levels to increase from 1 to 3 g/hp-hr to 2 to 5 g/hp-hr.
In one case, THC emissions increased from 5 to 18 g/hp-hr for an installed
pipeline engine that, possibly, required maintenance. In all cases, NMHC
levels ranged from 5 to 25 percent of THC levels and did not change with
derating. The increases in THC levels are probably related to the lower
exhaust and peak temperatures of part load operation.
Similarly, Figure C-26 shows that increases in A/F cause moderate
(10- to 25-percent) increases in THC levels. One 2-TC installed gas engine
was the exception; THC levels increased 75 percent from 11 to 19 g/hp-hr, and
may be related to a worn condition of the engine requiring maintenance (i.e.,
excessive blowby). Increases in A/F of gas engines that are already
operating lean tend to lower peak combustion temperatures and prevent complete
combustion. Consequently, HC emissions increase, although not as rapidly as
with decreases in A/F from rich (less than stoichiometric) operation (see
Figure 4-29). Note that the ratio of NMHC to THC remains essentially
constant with increases in A/F.
C-167
-------
Figures C-27 through C-29 indicate that retard, EGR, and MAT do not
cause THC or NMHC levels to increase. The application of retard delays the
combustion of the air/fuel mixture, resulting in higher exhaust temperatures
which, combined with an excess of oxygen, effectively oxidizes most
remaining unburned fuel. The application of this technique, however, causes
fuel consumption to increase (see Figure 4-27).
In contrast to retard, EGR and MAT controls usually cause no increase
in fuel consumption. It follows that if complete combustion is maintained
(evidenced by unchanged fuel consumption) HC emissions should remain low or
unchanged. With the application of EGR, unburned HC are recirculated or
trapped and then combusted during the next cylinder firing. The limited data
for EGR, illustrated in Figure C-28, show that NMHC and THC levels remain
unchanged.
With the reduction of manifold air temperature, however, HC emissions
can be expected to increase as the reaction of this specie proceeds at a
lower temperature. Data shown in Figure 4-52 indicate that THC levels
increase, but less than 25 percent, with manifold temperature reductions.
Nevertheless, the limited data for both NMHC and THC levels shown in Figure
C-29 indicate that these emissions remain unchanged.
Figure C-30 shows how the application of combinations of controls
affects NMHC and THC emissions. The results of this figure are mixed. It
should be noted that this data is limited, representing six engines from two
manufacturers. For three of the six engines shown, combined controls cause
THC and NMHC emission to increase in the same proportion. However, data for
two 2-TC designs (one dual fuel, the other gas) indicated that NMHC emissions
decreased (when THC levels increased) with the applications of: (1) retard
C-168
-------
and air-to-fuel for the dual fuel unit, and (2) EGR and derate for the gas
engine.
Opposing this result and of more concern, one 4-TC design showed that
NMHC emissions increased from 0 to 0.2 g/hp-hr when THC emissions decreased
from 5 to 3 g/hp-hr with the application of retard and MAT. Nevertheless,
the NMHC level is small, representing only 5 percent of the THC level. It
should also be noted that, despite precautions to avoid sample degradation,
the manufacturer of this engine experienced some inconsistencies in
determining NMHC levels (i.e., measured methane levels exceeded THC levels).
Therefore, the following section will briefly consider the procedures for
measuring both NMHC and THC before a recommendation concerning the separate
control of NMHC emissions is presented in Section C.4.4.
c-4.3 Measurement Practices for NMHC and THC Emissions
The development of a separate standard for NMHC emissions is largely
dependent on the demonstration of techniques for detecting NMHC directly, or
those which detect NMHC emissions indirectly by measuring methane and THC.
The NMHC level is then determined by the difference of the two measured
levels. It should be noted that the accuracy of indirect techniques depends
on the accurate determination of both methane and THC levels.
At the present time, there are basically three methods available for
determining NMHC emissions. These are(126):
• Gas Chromatographs (GC): Provide direct measurement of methane by
separation from other hydrocarbons. NMHC measurement is obtained
indirectly by subtraction of methane from THC or directly by
measurement of remaining nonmethane hydrocarbons. When NMHC is
measured directly, the methane value is also obtained and can be
C-169
-------
subtracted from a THC measurement (with FID) as a check of the
NMHC measurement.
t Nondispersive Infrared (NDIR): Also provides direct measurement
of methane. NMHC measurement is obtained indirectly by subtrac-
tion of methane measurement from independent THC measurement.
• Selective Combustor (SC): Gives a direct measurement of methane
by selective combustion of all nonmethane hydrocarbon, allowing
measurement of the remaining methane on a conventional THC
analyzer such as a flame ionization detector (FID). NMHC
measurement is obtained indirectly by subtraction of methane from
THC.
All of these NMHC techniques use a conventional heated flame ionization
detector (FID) to make one measurement of either NMHC using a GC, or THC
using a GC, NDIR, or SC.
Gas chromatographs are commercially available and were used by
Cooper(127) and SwRI to report NMHC levels from their engines. This method,
however, cannot perform real-time analysis (continuous rather than bag or
batch sampling), a capability desired by development engineers and present in
other major emission instruments (e.g., NOX, CO, SOX analyzers). In
addition, this instrument requires specialized training to operate and
maintain.
Selective combustors have only recently become available commercially.
They are simple and inexpensive compared to a GC but apparently require more
development. This is because not all lighter paraffins (e.g., ethane and
propane) are eliminated from the test sample during the selective removal of
all NMHC emissions. This results in higher methane (or smaller NMHC) levels
than were actually present. Colt used this method to report NMHC levels from
C-170
-------
their engines. However, no comparisons with GC data were available to
determine if this effect may have caused the Colt NMHC data to be
understated.
The third NMHC instrument, the NDIR, exists as a prototype and also
requires more development before becoming available commercially.
Nevertheless, the instrument has demonstrated accuracy and freedom from
interference for either batch or continuous sampling of methane emissions. It
is also believed that it will require greater maintenance than existing
NDIR's since it is more complex.
It should be noted that Fisher and Goodwin(128) supported the
development of standards for NMHC's. They noted, however, that although
measurements of NMHC levels were feasible, significant time and expense would
be required to adequately develop the necessary instruments and procedures.
Cooperative studies of the SAE measurement method have also
indicated that futher development of heated FID's is required to reduce
measurement uncertainties(129). The most recent study consisted of a
comparison of data from one diesel engine emission source measured during
round-robin tests. The results of this study showed that THC levels varied by
10 percent (on the average) for a given manufacturer and 22 percent among
manufacturers. It was concluded that further development of this measurement
method was necessary to reduce this scatter.
The preceding discussion indicates that measurement methods for NMHC
have not been adequately demonstrated at this time. In addition, studies of
FID instruments indicate that considerable measurement uncertainty can arise
from these instruments even though an accepted practice is followed.
C-171
-------
C.4.4 Recommendations for the Control of NMHC
Based on the preceding results, it 1s recommended that any control of
NMHC levels be based on the measurement of THC levels. This can be supported
by recognizing that: (1) in general, NMHC levels change in the same manner
as THC levels with the application of NOX control techniques; and (2)
monitoring NMHC levels accurately is considerably more difficult than for THC
since gas chromatographs or physical conditioning of the sample is required.
Therefore, if standards for hydrocarbons are deemed necessary, they should be
based on the reduction of THC, rather than NMHC, levels.
C-172
-------
C.5 DEVELOPMENT OF CONVERSION BSNOtf TO NOV (ppm) CORRECTED TO 15
PERCENT 02 xx
All emissions data in Alpendix C.I had been reported in terms of
brake specific emissions, BSNOX (g/hp-hr). In addition, applying
percent emissions reduction to the sales weighted uncontrolled emissions
levels (section 4.3.4) results in emission limits having the
brake-specific N0x format. Since a concentration format (ppm) has been
selected for the standard, a method must be derived to convert the BSNO
A
limits to concentration limits.
C-5-1 Converting BSNOX to N0;c Concentration Format
By definition, the volumetric concentration of NO in the exhaust
A
is simply the volume of NOX divided by the total exhaust volume. If the
mass concentration is known, dividing each (the mass of NO and exhaust)
A
by the respective molecular weights will give the volumetric
concentration. This may be equally done on a brake-specific basis.
Therefore, since the molecular weight of NOX will be defined equal to
N02 ( 46), the volumetric concentration of NO can be written as
A
BSNOX
NOX (ppm) = - HE
concentration Brake specific exhaust gas, g/hp-hr
molecular weight of exhaust gas
Both brake-specific N0x limits and the molecular weight of NO
A
are known. However, the brake specific exhaust gas flow and the modecular
weight of the exhaust gas are in general not available. Furthermore, to
set emission limits for each fuel type, the average ratio of exhaust flow
to exhaust gas molecular weight must be determined. This may be done by
C-173
-------
assuming properties of the fuel and a reference 15 percent 02
concentration in the exhaust gas. The following procedure has been usedto
determine the ratio of exhaust gas flow to exhaust gas modecular weight,
using data obtain from available brake-specific fuel consumption.
For any hydrocarbon fuel, the chemical reaction may be written as:
CxH + Z 02 + 3.76Z N2 * x CO,, + y/2 H.,0 + (Z - x - y/4) 02 + 3.76 N2 (2)
This is a generalized equation, ignoring the small concentrations of other
constituents such as NO HC, CO etc., which are less than 1.0 percent.
A
The moles of exhaust gas produced per mole of fuel burned is therefore
[x+ y/2 + (Z - x - y/4) + 3.76 Z]; and for dry products water is removed
leaving [x + (Z - x - y/4) + 3.76 z] moles of dry exhaust. This reduces
to [4.76 Z - y/4] total moles of exhaust gas. The molecular weight of the
gas is therefore the ratio of each constituent multiplied by the
respective constituent molecular weight:
M
x 44 + (Z - x - y/4) 32 +
'exhaust gas [4.76Z _ y/4] [4.76Z - y/4] [4.76Z - y/4]
or
M
exhaust gas [4>?6Z . y/4]
(44x + 32Z - 32x - 8y + 105.28Z) (3a)
C-174
-------An error occurred while trying to OCR this image.
-------
The exhaust mass flow must be determined next. Since the total
mass flow remains constant, the mass of exhaust gases per mole of fuel
burned is:
m dry exhaust « m fuel + m air - m water formed
(8)
From equation (1) above, (Z) moles of air are burned with each mole of
fuel, and (y/2) moles of water are formed, By multiplying by the
molecular weight of each, the mass flows are as follows:
1 x (12x + Y) + Z[32 + 3.76(28)] -(y/2)(18) (9)
Substituting equation (5) again (assuming 15 percent oxygen in exhaust);
m
m drv exhaust _ {12x +y) + (3.5x + 0.743y)(32 + 3.76(28)) - (y/2)(18) (10)
mole fuel
m dry exhaust _ ^x + 94y
mole fuel
Dividing by the molecular weight of the fuel yields the mass of exhaust
gases per mass of fuel, assuming 15 percent 02 in the exhaust stream
C-176
-------
ITI dry exhaust , 492x + 94y (12)
m fuel 12x + y
By multiplying this value by the brake-specific fuel consumption (BSFC)
the brake-specific exhaust flow 1s determined.
dry exhaust - BSFC
^2x + 94y)
TZxTy /
ZxT U3;
The ratio of exhaust mass flow to exhaust gas molecular weight can
now be determined. From equation (13) and (7):
BSFC x 4?2x *
Brake specific exhaust flow \ '2x + y
Molecular wt. of exhaust gas * ( 492x +~—^ (14)
This reduces to
BSFC x (16'6ftxV;29y) (15)
Substitution Into equation (1) and converting to parts per million (ppm)
gives:
C-177
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
the limits of 5 g/hp-hr for dual fuel, 9 g/hp-hr for gas, and 7 g/hp-hr
for diesel, (and the above fuel properties) the following NOX emission
limits were determined:
654 ppm for gas
518 ppm for diesel
370 ppm for dual-fuel
As a check of the validity of the conversion method, all raw ppm
versus BSNO data was plotted (corrected to 15 percent 02). By using
A
equation (16), the average fuel consumption, and a mean value of the
hydrogen/carbon ratio the average conversion curve was determined and is
shown in Figure C-34. The standard deviation (expressed in percent)
between this curve and the actual data was found to be 10 percent.
C-182 .
-------An error occurred while trying to OCR this image.
-------
C.6 STATISTICAL ANALYSIS FOR ALTERNATIVE EMISSION LIMITS
In selecting the numerical emission limits for each fuel type, it
was necessary to determine the tradeoffs associated with standards of
performance based on each of the three alternatives of applying a 40
percent NO emissions reduction. Thus, a statistical approach was
/\
considered the most logical approach.
The uncontrolled NOV emission from each engine fuel type were
A
assumed to follow a theoretical normal distribution. Figures C-35, C-36.
and C-37 illustrate the theoretical normal distribution curves for diesel,
dual-fuel, and gas engines, respectively. Note that the curves have
truncated ends at positive points on the axis as no engines emit 0
g/hp-hr. The areas under each curve represent the population if each
engine fuel type emitted up to a particular limit. The mean values which
bisect the normal distribution were assumed to be the sales-weighted
uncontrolled average NOX emission levels discussed in Section 4.3.4 and
the standard deviations, for each fuel type were assumed to be equal to
the standard deviation calculated from the data base in Appendix C. That
is, the standard deviation is not sales weighted as no such information
was available. If the standard deviation were calculated for increasingly
Targe samples and increasing small class intervals, its value would be
expected to approach the theoretical standard deviation. The data
tabulated in Appendix C was used to calculate sample standard deviations
for each fuel type: (1) Gas, a- 4 g/hp-hr, (2) Dual-fuel, a = 3.2
g/hp-hr, and (3) Diesel,
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
Standards of performance based on each of the three alternatives
were analyzed relative to the theoretical normal distribution curves to
determine: (1) the percentage of the engines that would have to reduce
NO emissions by 40 percent or less to meet the standard; (2) the
percentage of engines that would be required to do nothing to meet the
standard; and (3) the percentage of engines that would be required to
reduce NO emissions by more than 40 percent to meet the standard. The
X
results are illustrated on each figure and summarized in Table C-93.
As a check on the accuracy of the assumption of a normal
distribution, the actual data base in Appendix C was analyzed relative to
percentages of engines for each alternative determined from the normal
distribution. The results are tabluated in Table C-93 as the numbers in
parentheses. The actual sample values show very good agreement with the
values determined by the pure normal distribution approach.
Thus, it can be concluded that the assumption that the uncontrolled
NOX emission levels, for a given fuel type, follow a theroetical normal
distribution curve is essentially true and the statistical approach is
deemed to be a valid approach.
C-188
-------
TABLE C-93. SUMMARY OF STATISTICAL ANALYSES OF ALTERNATE EMISSIONS LIMITS
GAS ENGINES
Alternative
Standard
Percent meeting standard
with f_40 percent control
Percent required to do
nothing to meet standard
Percent required to apply
40 percent control to
I II III
17 9 11
99 (100) 50 (50) 84 (91)
69 (67) 7 (6) 18 (9)
1 (0) 50 (58) 16 (9)
DUAL-FUEL ENGINES
Alternative
Standard
Percent meeting standard
with £40 percent control
Percent required to do
nothing to meet standard
Percent required to apply
40 percent control to
I II HI
9 5 7
98 (100) 54 (44) 87 (89)
62 (67) 18 (4) 48 (44)
2 (0) 46 (56) 13 (11)
DIESEL ENGINES
Alternative
Standard
Percent meeting standard
with < 40 percent control
Percent required to do
nothing to meet standard
Percent required to apply
40 percent control to
I II in
11 7 9
98 (100) 56 (63) 86 (90)
50 (65) 14 (18) 29 (45)
2 (0) 44 (37) 14 (10)
C-189
-------
C 7 COMPARISON OF SMALL-BORE, AUTO ENGINE EMISSION CONTROL TECHNIQUES
FOR LARGE BORE-STATIONARY 1C ENGINES
C.7.1 Introduction
During a review meeting with EPA in September 1977, several
questions were raised concerning the application to large-bore stationary
engines of control techniques that have been shown to be effective on
automotive engines. In addition, questions were asked why catalytic
reduction of NO (which has been successfully used in Japan) and water
/\
induction (the primary N0x control technique for stationary gas
turbines) were not alternative controls for stationary i, ^ngines. This
memo briefly summarizes these three questions.
C.7.2.1 Comparison of Control Techniques for Large-Bore Stationary
and Automotive Engines
Essentially the same NOX control techniques that have been used
by engine manufacturers to meet emission regulations for mobile sources
are also effective when applied to large-bore engines. These techniques
include derating*, retarded ignition or fuel injection, manifold air
cooling (with turbocharging), air-to-fuel ratio changes, and exhaust gas
recirculation. Of these controls, retard (for diesel or dual fuel
engines) and air-to-fuel changes (for natural gas engines) are
particularly effective when applied to large-bore engines.
Nevertheless, inherent differences in the design and operating
modes of large stationary and smaller, mobile engines have dictated
'Derating is accomplished by using a larger engine than necessary for a
particular car.
C-190
-------
different approaches to exhaust emission control. Figure C-38, which
shows the effect of A/F ratio on exhaust emissions, can be used to
illustrate this point. Typically, spark ignition automotive engines are
operated at slightly less than stoichiometric A/F ratios. This operation
is largely dictated by the difficulty in closely controlling the A/F
ratios to individual cylinders within these carbureted engines.
Therefore, A/F ratios must be maintained richer than stoichiometric to
avoid detonation and misfiring.
With the advent of strict emision'regulations (primarily aimed at
reducing HC and CO emissions), automotive manufacturers concentrated on
increasing the A/F ratio (e.g., "lean burn" engines), which, as
Figure C-35 illustrates, decreases HC and CO emissions significantly and
also improves fuel economy. N0x emissions, however, increase initially
as A/F ratios are increased beyond stoichiometric. As emission
regulations became even more stringent, additional controls were,
therefore, required. These controls included exhaust gas recirculation
(EGR) and returning to richer A/F ratios to reduce N0x. To compensate
for the resulting increase in HC and CO emissions, oxidizing catalysts
were added. These catalysts also enabled automobiles to reach the more
stringent requirements for HC and CO that became mandatory in 1975.
Improvements in carburetion and mixing were also effective in optimizing
exhaust emissions over a range of loads and speeds.
Lage bore engines, in contrast, typically operate leaner than
stoichiometric. Consequently, as shown in Figure C-35, NO emissions
/\
C-191
-------An error occurred while trying to OCR this image.
-------
are relatively high, but HC and CO emissions are now low*. These
operating conditions result from the need to optimize fuel consumption of
large-bore engines because they see high usage, high load service.
Therefore, the emphasis of exhaust emission control for large-bore engines
is in reducing NOX emissions. Although the same types of combustion
modifications that are effective in reducing NO from automotive engines
A
could be used on large stationary engines, the selection of optimum
controls and the direction and degree of application of the selected
controls is different due to differences in uncontrolled A/F ratios and
fuel charging methods.
C.7.2.2 Catalytic Reduction
As of this date no large-bore engine manufacturers are using
catalysts to reduce NO emissions. NO reduction catalysts, until
^ A
1977, have not been used on gasoline or diesel vehicles either. The
catalysts that have been and are still currently in use on automobiles are
primarily oxidizing catalysts, whose purpose is to lower HC and CO
emissions. Recently, however, three-way catalysts have been developed for
the reduction of HC, CO, and NOV emissions and the first commercial
A
application was for a 1977 Volvo sold in California. (Three-Way
Conversion Catalysts — Part of the New Emission Control System; SAE
Paper, 770365). These three-way catalysts are similar to the earlier
oxidizing catalysts (precious metals coated on monoliths or pellets) with
*The exceptions to this generalization are those large-bore engines which
are naturally aspirated or carbureted; these units have emissions
characteristics which are similar to automotive engines since they operate
at A/F ratios closer to stoichiometric.
C-193
-------
the addition of rhodium which selectively reduces NOX when the A/F ratio
1s maintained very near sto1ch1ometr1c. These conditions exist 1n
automotive exhausts but not in the discharge from large bore engines.
Therefore, this control approach shows little promise for large bore
engines.
As discussed on page 4-139 of the SSEIS, however, an
ammonia/catalyst N0¥ control system has been suggested in the literature
J\
and 1s based on the successful application of this technique to nitric
acid plant tail gases and a number of large stationary combustion sources
in Japan. An excellent discussion of technique is presented in a paper
"Status of Flue Gas Treatment Technology for Control of NOX and
Simultaneous Control of SOX and NOX", Mobley, J. David, and Stern,
Richard D., U.S. Environmental Protection Agency, Report No.
EPA-600/7-77-033C. This paper is an excellent review of this technology,
its effectiveness, and its costs. Based on information in this report it
would appear that selective catalytic reduction is a technically feasible
approach for gaseous fuel engines, but the fouling (particulate) and
catalyst-poisoning (sulfur) problems associated with oil or coal
combustion would require further development. Moreover, this approach
appears to be very expensive. If the published costs are extrapolated for
1C engines, the data indicate that annual ownership and operating costs
would increase from between 3 to 30 percent. These costs are based on
sources that are at least an order of magnitude larger than a typical
engine installation. In addition, the costs of maintenance and catalyst
replacement may be somewhat understated for smaller sources, such as 1C
engines. Thus, it is probable that the actual costs for an engine
application would be even greater than these estimates.
C-194
-------
A similar N0x reduction scheme has been developed by Exxon
(Thermal Denox Process) for N0x reduction from stationary boilers and
furnaces. In this process, ammonia is injected into the flue gas at
locations where the temperature exceeds 1500°F. There appears to be no
application for this process to 1C engine exhausts, however, since exhaust
temperatures are generally less than 1000°F.
C.7.2.3 Water Induction
As discussed in the SSEIS (Section 4.4.7), water induction has been
investigated by four large bore engine manufacturers (White Superior,
Ingersoll-Rand, Cooper, and GMC/Electromotive). All reported serious
concern about the feasibility of this technique based on observations of
water in the crankcase and lubricating oil, rapid build-up of mineral
scale (untreated water) around intake/exhaust valves and other components,
and combustion deposits. All of the tests were of short duration (less
than 25 hours of engine operation), and manufacturers believe much longer
tests (2000 to 8000 hours) would be necessary to establish the effect of
water induction on wear rates and operating reliability.
In addition to these manufacturers, two smaller bore engine
manufacturers, GMC/Detroit Diesel and Caterpillar, have reported tests of
water induction in truck-size diesel engines. Caterpillar believes the
technique is viable if the water is treated for freeze protection and
mineral content. Caterpillar tests showed significant deposits, but they
concluded these could be prevented by demoralizing the water.
GMC/Detroit Diesel also experienced significant deposits and concluded
that these deposits originated from condensation of combustion products
rather than water minerals. Based on their exprience, GMC/Detroit Diesel
concluded that water induction was not feasible in their 2-stroke designs.
C-195
-------
C.8 Determination of Sales-Weighted Average Uncontrolled CO and HC
Emissions
Since the source of variability caused by engine design cannot be
specifically Identified, a procedure 1s required to characterize uncon-
trolled CO and HC emissions levels of engines which are sold for simi-
lar applications.
The procedure adopted here is to compute a weighted, average
uncontrolled emission level for engines in the diesel, dual-fuel, or
natural gas categories. The three weighted levels are based on sales
of engine horsepower during the past five years for domestic applica-
tions. Sales of horsepower to standby services were excluded from this
computation, since engines sold for standby applications will be exempted
from standards of performance (see Chapter 9).
The sales-weighted averages for diesel, dual-fuel, and natural gas
engines are presented in Figures C-39(a-f), which also show each
manufacturer's CO and HCt data. The weighted average uncontrolled CO
level for diesel engines is 2.9 g/hp-hr; for dual-fuel units, 2.7 g/hp-
hr; and for natural gas engines, 7.7 g/hp-hr. The weighted average
uncontrolled HCt level for diesel engines is 0.3 g/hp-hr; for dual-fuel
units, 2.8 g/hp-hr; and for natural gas engines, 1.8 g/hp-hr.
C-196
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------An error occurred while trying to OCR this image.
-------
REFERENCES FOR APPENDIX C
(1) Storment, 0. 0. and K. J. Springer. Assessment of Control Techniques
Report AR-8§4 $11*19™ LoCOrn°t1ve En9ines- Southwest Research Institute
(2) DEMA Exhaust Emission Measurement Procedure for Low and Medium Speed
CleveUnd?°Oh1oti0S E5§1nei974D1eSe1 Eng1ne Manufacturer* Association.
r«nne' Fc ?' a?drKl V-.Be19htol (Cooper-Bessemer). Effect of Operating
Conditions on Exhaust Gas Emissions of Diesel, Gas Diesel, and Spark Ignited
Stationary Engines. Unpublished Cooper-Bessemer Co. Report Presented
at eas Compression Shortcourse, Norman, Oklahoma, Fall 1973.
(4) McGowin, C. R., F. S. Schaub, and R. I. Hubbard. Emission Control
rnn3iHn^°nJrLl??"Dtroke^park;G!s,Enf1ne b* Modification of Operating
Conditions, AGA/IGT Proceedings 2nd Conference on Natural Gas Research
and Technology, Atlanta, Georgia. 1972.
(5) Hanley, G. P., Marketing and Technical Data on Reciprocating Engines
for Stationary Application. General Motors Corp. January 24, 1974.
f-' ?' and/;.V' Bei9htol. Effect of Operating Conditions
iS?Sb«iF' c' ^°^er"?ecsemer)' M6^0^ of Reduction of NOX Emissions
a9°WSPd DeS1E U'S' *"
(Acurex/
M. P. Thompson (White Superior) to D. R. Goodwin (EPA),
pJlLrf5!?*!* *',C; Em1"ion Reduction Study on a Carbureted Natural Gas
Fueled Industrial Engine. White Superior Division. White Motor Corporation.
C-203
-------
(Presented at Annual ASME Meeting, Diesel and Gas Power Division, New
Orleans, April 1975).
(12) Letters from N. S. Cox (Waukesha) to G. R. Offen (Acurex/Aerotherm)
November 20, 1974 and June 16, 1975.
(13) Caterpillar Tractor Co. Data submitted to Acurex/Aerotherm, August
27, 1974.
(14) Brisson, B.s et al., A New Diesel Combustion Chamber -- The Variable
Throat Chamber. SAE Paper 730167. 1973.
(15) Letter from M. P. Thompson (White Superior) to D. R. Goodwin (EPA),
August 3, 1976.
(16) Letter from R. Sheppard (Ingersoll-Rand) to D. R. Goodwin (EPA),
June 17 and September 22, 1976.
(17) Letter from A. R. Fleischer (DeLaval) to D. R. Goodwin (EPA),
July 30, 1976.
(18) Letter from R. D. Smalley (White Alco) to D. R. Goodwin (EPA),
July 29, 1976.
(19) Letter from W. E. Snyder (Waukesha) to D. R. Goodwin (EPA),
July 20, 1976.
(20) Letter from F. S. Schaub (Cooper) to D. R. Goodwin (EPA), August
4, 1976.
(21) Letter from C. L. Newton to D. R. Goodwin (EPA), September 10, 1976.
(22) Letter from R. D. Henderson to Fred Porter (EPA), March 17, 1977.
(23) "Revised Heavy Duty Engine Regulations for 1979 and Later Model
Years," Federal Register, Volume 41, No. 101, May 24, 1976.
(24) Coordinating Research Council (CRC), "Effect of Humidity of Air
Intake on Nitric Oxide Formation in Diesel Exhaust," CRC Report 447,
December 1971.
(25) Krause, S. R., "Effect of Engine Intake-Air Humidity, Temperature,
and Pressure on Exhaust Emissions," SAE Paper 710835.
(26) Ibid.
(27) "Revised Heavy Duty Engine Regulations for 1979 and Later Model
Years," Op_. C1t.
(28) Krause, Op,. Cit.
(29) Ibid,
C-204
-------
(30) Ibid.
(31) Ib1_d.
(32) Ib1_d.
(33) Ibid.
E"9lni
of
(36) State of California, A1r Resources Board, 1973, "California Exhaust
Emission Standards for Gasoline Powered Motor Vehicles," Proposed, November
13, 1970.
(J7ii E2Vi1ip°I?int!1 Protect1on ASency (40 CRF Part 85, Subpart A), "Control
of Air Pollution from New Motor Vehicles and New Motor Vehicles Engine,"
(Notice of Proposed Rule Making), February 1971.
(38) Scott, Op_. C1t.
.1 P' 5;^'Humii!1cLE!!fects on En91ne N1tric Ox1de Emissions
at Steady-State Conditions," SAE Paper 700467.
(40) Krause, 0£. C1t.
(41) ibid.
(42) Ibid.
(43) Brown, et al., 0£. C1t.
(44) Ibid.
(45) Scott, 0£. Clt.
fiFny^!; "El!l1ssi;nACont'"01 of a Stationary Two-Stroke Spark-
Co^plny Rlport%972" n Operat1n9 Conditions," Shell Development
R' Fle1scher (0eLava1 Turb1ne Inc«) ^ D. R. Goodwin
(48) Coordinating Research Council, Op_. Cit.
-r' S: 5" Merr1on, D. F., and Green, G. L. "Effect of Inlet
Air Humidity and Temperature on Diesel Exhaust Emissions," SAE Paper 730213.
"Ren1ser^eavy Duty Eng1ne Regulations for 1979 and Later Model
. up. Clt.
C-205
-------
(51) Krause, Merrlon, and Green, Op_. C1t.
(52) Coordinating Research Council, OJD. C1t.
(53) Krause, Merrlon, and Green, 0£. C1t.
(54) Coordinating Research Council, Oj>. C1t.
(55) Krause, Merrlon, and Green, Op_. Cit.
(56) Touchton, G. L. and N. R. Diebeluis, "A Correlation of Nitrogen
Oxides Emissions with the Gas Turbine Operating Parameters," ASME Paper
76-GT-14.
(57) Shaw, H., "Effect of Water on Nitric Oxide Production in Gas Turbines
Combustors," ASME Paper 75-GT-70.
(58) Blazowski, W. S. and D. E. Walsh, "Prediction of Aircraft Gas Turbine
NOx Emission Dependence on Engine Operating Parameters and Ambient Conditions,"
AIAA Paper 73-1275.
(59) Vaught, J. M. "The Effect of Inlet Temperature and Pressure on an
Industrial Turbine Engine Exhaust Emission," ASME Paper 75-WA/GT-ll.
(60) Hung, W. S. Y., "An Experimentally Verified NOx Emission Model for
Gas Turbine Combustors," ASME Paper 75-GT-71.
(61) Marzeski, J. W. and W. S. Blazowski, "Ambient Temperature and Pressure
Corrections for Aircraft Gas Turbine Idle Pollutant Emissions," ASME Paper
76-GT-130.
(62) Hung, Cj£. Cit.
(63) Touchton and Diebeluis, £p_. Cit.
(64) Ibid.
(65) Krause, Merrion, and Green, Op_. Cit.
(66) Gas Turbine International, May-June 1974.
(67) Ibid.
(68) Krause, Merrion, and Green, Op_. Cit.
(69) Private communication between S. B. Youngblood (Acurex/Aerothenn)
and J. McDermon (EPA), October 1976.
(70) Ibid.
(71) Krause, Merrion, and Green, Op_. Cit.
C-206
-------
^I2l ?u> T; ??d Kl J> McAulay» "Predicting Diesel Engine Performance
at Various Ambient Conditions," SAE Paper 730148. romance
ancM ^ges*"18' F' A" Combust1on Theory. Add1 son-Wesley, Chapters 3
(74) Wilson, R P., Jr., E. B. Mu1r and F. A. Pell1cc1ott1, "Emission
Study of a Single-Cylinder Diesel Engine," SAE Paper 740123.
(75) Williams, Op_. Cit.
R< W< Sheppard dngersoll-Rand) to D. R. Goodwin (EPA),
oMeSJSHM^^nS^n'I^n?0"";1] £CRC) "Co°Perative Evaluation of Techniques
for Measuring NO and CO in Diesel Exhaust: Phase III," CRC Report, 1972.
" SAE Recommended
(79) Perez, J M., L. C. Broering, and J. H. Johnson, "Cooperative Evaluation
of Techniques for Measuring NO and CO (Phase IV Tests)," SAE Paper 75204
nSi n?oord.1;at1n? Research Council (CRC) "Cooperative Study of Heavy
Duty Diesel Emission Measurement Methods," CRC Report 487, July 1976.
(83) Perez, Broering, and Johnson, Op_. C1t.
(84) TRW Systems Group, "A Study of Mandatory Engine Maintenance for
Reducing Vehicle Exhaust Emissions, Volume VI: A Comparison of Oxides
of NUrogen Measurements," CRC APRAC CAPE 13-68-12, July 1972.
(85) Krause, Merrion, and Green, £p_. Cit.
(86) TRW Systems Group, Op_. Cit.
(87) Tuttle, J. H., R. A. Shisler, and A. M. Mellor, "Nitrogen Dioxide
Formation in Gas Turbine Engines: Measurements and Measurements Methods "
Purdue University Report No. PURDU-CL-73-06, December 1973 |viei:noas»
(88) Ibid.
"Instrumentation for the Determination
Emi'ssions'" Volume I.
C-207
-------
(90) Maahs, H. G., "Interference of Oxygen, Carbon Dioxide, and Water
Vapor on the Analysis for Oxides of Nitrogen by Chemilumlnescence.
(91) Matthews, R. D., R. F. Sawyer, and R. W. Schefer, "Interference in
Chemilumlnescent Measurement of NO and N02 Emissions from Combustion Systems,
Draft of Paper Submitted to Combustion and Flame, Fall 1976.
(92) Slgsby, J. E., Jr., et al., "Chemlluminescent Method for Analysis
of Nitrogen Compounds in Mobile Source Emissions," Environmental Science
and Technology, Volume 7, No. 1, January 1973.
(93) Coordinating Research Council (CRC), "Cooperative Study of Heavy
Duty Diesel Emission Measurement Methods," CRC Report 487, July 1976.
(94) Tuttle, Shlsler, and Mellor, Cjp, Cit.
(95) TRW Systems Group, Op_, C1t.
(96) Ibid.
(97) Clemens, W. B. (EPA/Ann Arbor), "Impact on Existing Standards Due
to Proposed Instrumentation Changes in the Heavy Duty Federal Test Procedure,
Interoffice Memo, September 18, 1974.
(98) Perez, Broering, and Johnson, Op_. Cit.
(99) Krause, Merrion, and Green, Op_. Cit.
(100) Op_. Cit., Reference 93.
(101) Personal comunication between S. B. Youngblood (Acurex/Aerotherm)
and W. B. Clemmens (EPA/Ann Arbor), July 22, October 19, and November 1,
1976.
(102) Ibid.
(103) Schaub, F. S. and K. V. Beightol, "NO* Emission Reduction Methods
for Large Bore Diesel and Natural Gas Engines," ASME Paper 71-WA/DGP-2.
(104) OJK Cit.
(105) TRW Systems Group, Op_. Cit.
(106) [bid.
(107) 0_p_. Cit., Reference 101.
(108) Ibid.
(109) Ibid.
(110) "Revised Heavy Duty Engine Regulations for 1979 and Later Model
Years," Op_. Cit.
C-208
-------
(Ill) Diesel Engine Manufacturers Association (DEMA), "DEMA Exhaust
Measurement Procedure for Low and Medium Speed Internal Cornbustio! ! E
°C°' and N° tn De1Se1 ExhaUSt'" SAE
] !?0t0r? Corporation Statement on Draft SSEIS for Stationary
Internal Combustion Engines," May 14, 1976. ««-'unary
(114) Op_. Cit. . Reference 93.
(115) °R- £21.. Reference 101.
(116) TRW Systems Group, Op_. Cit.
(117) Tuttle, Shiser, and Mellor, Op_. Cit.
(118) Schaub and Beightol, Op_. Cit.
(119) Clerrtmens, Op_. Cit.
(120) Ibid.
(121) 0£. Cit.. Reference 101.
(122) Ibid.
(123) TRW Systems Group, Op_. Cit.
a"d °' R' Goodw1n -
(127) Schaub and Beightol, OJD. Cit.
(128) 0_£. Cit.. Reference 126.
0rM» R?SeauC!; Counci1 (CRC)' "Cooperative Evaluation of
1975 Measur1n9 H^rocarbons in Diesel Exhaust," CRC Report 471,
C-209
-------
GLOSSARY
aftercooler
air-cooled
air injection
air-to-fuel ratio
annu.a!1zed cost
baseload
BDC
blower scavenging
bore
bottom dead center
brake horsepower
(bhp)
- a heat exchanger used to cool inlet air that has
passed through a turbocharger. Also referred to as
an Intercooler (see Figure 4-32).
-- a method of engine cooling, often used on small
engines. A1r 1s sucked or forced over the engine (by
an engine-driven fan) transferring heat from the
engine block. F1ns are generally used to conduct heat
away from the combustion chambers and increase the
area available for heat transfer.
-- an emission control technique in which air is pumped
(injected) into the exhaust manifold to complete the
reaction of any unburned fuel and CO which is still
present.
- a ratio of the mass flowrate (g/hr) of air into an
engine to the mass flowrate of fuel (g/hr).
-- initial costs, or the sum of initial and operating
costs, expressed as an annual charge.
— continuous operation, generally considered to be 8000
hr/yr.
-- see bottom dead center.
— a method of charging the cylinder of an engine with
air in which a low-pressure blower driven by the
engine forces air into the cylinder.
— the diameter of the cylinder of an engine.
-- position of the piston when it is at the bottom of the
cylinder; corresponds to maximum available gas volume.
- the power delivered by the engine shaft at the output
end. The name is derived from the fact that it was
first measured by the power consumed in a brake
attached to the output shaft. Brake horsepower equals
G-l
-------
brake mean
effective pressure
(bmep)
brake specific
carburetor
catalytic converter --
cetane number
chemiluminescent
analyzer
compression
ignition (CI)
compression ratio
(00
connecting rod
continuous rating
cooling system
total power delivered by the pistons less losses and
power used to drive auxiliaries (fan, pumps, etc.).
the hypothetical constant pressure that would have
to be exerted on the head of the piston during the
entire stroke to generate the same torque as is
actually generated by the engine. Hence, frequently
used synonymously with torque.
emissions expressed on the basis of power output,
i.e., emissions g/hp-hr.
a device on a spark ignition engine which controls the
flowrate of air and fuel and mixes them in the proper
proportions for combustion.
a device which uses a catalyst to promote a reaction
that alters the chemical composition of the gas
passing through it. Oxidation type converters change
CO and HC to C02 and H2° using a precious metal
catalyst, such as platinum. Reduction catalysts are
used to reduce NO to N2 and 02-
a reference number for compression ignition engine
fuels. Higher numbers indicate better ignitability
and better antiknock characteristics.
— a
device used to measure NOx emissions,
— one of two methods of initiating combustion in the
engine cylinder. In CI engines, the air charge is
introduced into the cylinder and compressed, thereby
raising its temperature above the auto-ignition
temperature of the fuel (temperature at which the fuel
ignites spontaneously). Fuel is then injected into
this hot compressed air and ignites spontaneously.
All diesel and dual fuel engines are compression
ignited.
-- ratio of the volume in the combustion chamber when
the piston is at the bottom of the stroke to that when
the piston is at the top.
— a rod which connects the piston to the crankshaft and
permits the reciprocating motion of the piston to be
transferred to rotary motion by the crankshaft.
-- see rating.
— the system by which combustion heat is removed from
engine block. This system may consist of a water
jacket through which a liquid is circulated to remove
G-2
-------
cost effectiveness —
crank case blowby
crankshaft
crude
cylinder Uner
den1tr1f1cation
derating
desulfurlzatlon
digester gas
direct Injection
displacement
distillate
dribble
dry
heat from the engine and a radiator, cooling tower, or
heat recovery system to remove the heat from the
liquid before It 1s redrculated through the water
Jacket. It may also consist of fins to conduct heat'
away from the combustion chambers and a fan to move
air past these fins to remove the heat from them (see
"air-cooled"),
ratio of the cost to reduce a pollutant from an engine
to the amount of pollutant removed ($/kg),
unburned fuel, combustion gases, and lubricating oil
which escape from the cylinder past the piston rings
into the crankcase and are then vented.
the shaft that receives engine power from the
reciprocating motion of the pistons and delivers 1t as
rotary motion.
-- unrefined oil.
-- a steel Uner Inserted Into the cylinder. It can be
removed and replaced after excessive wear.
-- the removal of nitrogen from fuel during refining.
Some removal occurs concurrently with sulfur removal.
-- a control technique which limits the maximum load of
an engine to less then the design value.
-- the removal of sulfur from fuel. This occurs during
the refining of crude on Into light distillates and
gasoline.
— fuel gas formed from sewage sludge. Primary
constituents are CH4 (methane), C02, and Hg.
-- the Injection of fuel directly into the cylinder.
-- the volume a piston sweeps out moving from the bottom
of the cylinder (bottom dead center, BDC) to the top
of the cylinder (top dead center, TDC).
— a product of the refining of crude oil. The most
common distillates are No. 2 dlesel oil and gasoline.
~ the loss of fuel from the injector tip after fuel
Injection.
~ refers to gas measurements made under conditions where
water has been removed from the gas before the
measure 1s made. Also gas measurements which have been
6-3
-------
durability
exhaust gas
red roil at1 on
(EGR)
exhaust manifold
field gathering
flame ionization
detector (FID)
fossil steam
four-stroke
fuel additive
fuel-bound
nitrogen
fuel pump
governor
heavy duty mobile
higher heating
value
mathematically adjusted to be equivalent to measure-
ments made on water-free gas.
-- the ability of an engine to operate throughout a
normal service life without excessive wear or failure
of engine components.
« an emission control remote technique whereby a portion
of the exhaust gases are retained 1n the cylinder
(Internal EGR) or are routed back to the engine Intake
(external EGR) to displace some of the Inlet air.
- an internally ducted casting which receives exhaust
products from the cylinders and transfers these
products to the exhaust systems (see Figure 4-47).
-- collection of oil or natural gas at the surface of a
well (the well-head) into the pipes (feeder lines)
which carry it to the major pipelines.
-- an analytical instrument used to measure HC emissions.
-- steam for electric utility power generation produced
from the combustion of coal, oil, or natural gas.
- a type of engine which requires four traverses of the
piston in the cylinder (two revolutions of the
crankshaft) per power stroke (i.e., to complete a
cycle).
-- a substance added to fuel, usually to reduce smoke
from an engine.
- nitrogen contained in the fuel rather than in the air.
- device which pumps fuel to the fuel injection system
or carburetor.
— device which controls the amount of fuel supplied to
the cylinder according to the load demand on the
engine, e.g., a governor is used to maintain a given
speed (rpm) under varying load in electric generator
applications.
-- a term referring to vehicles over 6000-lbs gross
vehicle weight (total weight when vehicle is fully
loaded) subject to federal emission standards.
— the heat produced by the complete combustion of a
unit quantity of fuel (at standard conditions of
temperature, pressure, and humidity) such that the
G-4
-------
phase
products of comb"stion is in the liquid
horsepower
indirect injection —
the time rate of doing work. One U.S. horsepower is
equal to 33,000 foot-pounds per minute. (One US
horsepower equals 1.014 metric horsepower.)
initial cost
injection rate
injection timing
injector
injector rack
intake manifold
intercooler
intermittent rating
jacket water
knock
lean
mph "son chamber, or
antechamber, where combustion commences in an oxygen
deficient environment before expanding into the
cylinder for completion of the combustion in the
presence of excess oxygen (see Figure 4-42).
the purchase price of the engine, including all
auxiliaries necessary for its operation (e.g.,
etc.P' afterC°°1er> Carting motor,
1ntroduced into
cylinder
the time measured in degrees of crankshaft rotation,
that it takes for fuel to be admitted into the
cylinder. Start of injection (in degrees before TOO
is an important parameter in emissions control.
Figure ^injects fuel int° the cylinder (see
a mechanical linkage controlled by the governor which
determines the pressure of the fuel supplied by the
fuel pump to the fuel injector. y
internally ducted casting that distributes incoming
an ^ m'XtUre 1nt° the ™
-- see aftercooler.
~ see rating.
- engine cooling water that circulates from a cooling
tower or radiator through cavities (or iacketO in th*
engine block. j«.*ets>y in tne
-- premature ignition of fuel in cylinder of an engine.
mixture which contains
burn all th f 1 " lv""'I
-------
low sac nozzle
injector
lower heating value —
major overhaul
manifold air
temperature cooling
(MAT)
maximum rating
minor overhaul
a fuel injector with a minimum of volume beneath
the injector nozzle, thus reducina injector dribble
and HC emissions (see Figure 4-55).
the heat produced by the complete combustion of a unit
quantity of fuel (at standard conditions of
temperature pressure, and humidity such that the
Serin the products of combustion is in the vapor
phase.
- an overhaul which generally includes removal and
replacement (or reconditioning) of the cylinder
liners, pistons, rings, valves, fue pump, and
crankshaft. The engine is frequently removed from the
installation to perform this overhaul.
- a control technique based on lowering the intake air
temperature. Usually accomplished with an intercooler
(see Figure 4-32).
-- see rating.
- an overhaul which generaly includes replacing
rings, valves, injectors or spark p ugs, and
occasionally pistons. This overhaul is> usual y
performed without removing the engine from its
installation and at more frequent intervals than major
overhauls.
naturally aspirated ~
nondispersive
infrared analyzer
(NDIR)
nonpropulsive
nuclear steam
a type of air charging whereby the f'st™
into the cylinder as it travels to the bottom of the
cjl?nder. No? turbocharged, supercharged, or blower
scavenged.
— an instrument used to measure CO and C02*
an application where the engine is never used to move
The structure or device on which it is mounted.
steam (usually for electric generation) produced by
heat from nuclear reactions.
operational change --
original equipment
manufacturer (OEM)
a change in engine operation which requires only an
adjustment of given operating conditions but no
hardware additions - e.g., ignition timing.
a firm which, in this case, buys an engine from an
engine Manufacturer and incorporates It into a product
of which the engine is only a component -- e.g.. J
portable compressor, which is usually assembled by an
6-6
-------
overhaul
oxidation
paramagnetic
analyzer
peak cylinder
temperature
pilot oil
piston crown
piston ring
ports
precombustion
chamber
physical change
quenching
rating
Worn arts 1n an
usually
ariHiHnn J s document» oxidation refers to the
addit on of oxygen to a molecule by a chemical
reaction, as when CO combines with oxygen to form CO*
(The more general definition is the loi of electron!'
by a reactant in a chemical reaction.) eie"rons
oxygen content of
-- the^maximum temperature in the cylinder during combus-
- a small amount (<10 percent by heating value) of
.trX. ofU6a duna?.C^eVLaginTtoend^tethteheC°^sreeoSuSrn
-- the upper surface of the piston.
"" p'iston/V^sJjnTin'gstlal th^cif0 a-9r°°Ve in th6
filling the gap between the pistol and^he^iUer.^
"" fSl^ent'e^the^ Cf1-irfder liner though which air and
-- see indirect injection.
°PP°Sed
the cooling below the combustion point of fuel which
has impinged on the walls of the cylinder.
brake horsepower output of the engine. Reported
either as continuous (power that Ingine cSTdeflver
continuously), intermittant (power that engine can
6-7
-------
reactivity
(photochemical)
reduction
residence time
residual
retard
rich
rpm
sac
scrubber
smoke
spark ignition (SI)
spark timing
squish lip
potential of a hydrocarbon compound to react with
other species in the atmosphere to produce smog.
. when used in this document, reduction refers to the
removal of oxygen from a molecule by a chemical
reaction, as when NO is reduced to N2 and 09 In the
presence of a catalyst. (The more general definition
is the gain of electrons by a reactant in a chemical
reaction.)
- the time interval after ignition during which the
air-fuel mixture can burn at elevated temperature and
pressure.
- a heavy viscous oil, often containing large amounts of
sulfur (>1 percent by volume), which remains after
distillation to produce other fuels.
- a NOX control technique wherein ignition of the fuel
is delayed by delaying the spark [SI engines) or the
start of fuel injection (CI engines) by several
degrees of crankshaft rotation.
- refers to an air and fuel mixture which contains less
oxygen than is stoichiometrically necessary to
completely burn all the fuel.
— rotations per minute; a measure of the engine
crankshaft rotational speed.
- the small volume below the nozzle of an injector (see
Figure 4-55).
- a device which removes a pollutant from an exhaust gas
stream through absorption of the pollutant by the
scrubber liquid.
-- visible emissions from an engine exhaust.
- one of two methods of initiating combustion in the
engine cylinders (see also compression ig;1*10"'- ™
SI engines, an electrical spark is generated across a
aao between two electrodes at the tip of the spark
plSg to ignite the fuel-air mixture .All gasoline and
natural gas fueled engines are spark ignited.
- the degrees of crankshaft rotation before top dead
center (TDC) at which the spark commences.
- refers to a cavity shape in the piston head which
gfnlrates squish, or radially outward motion of the
air-fuel mixture. Combustion is initiated in this
cavity.
6-8
-------
standby
stationary
sto1chiometr1c
stratified charge
stroke
swirl
TDC
thermal efficiency
thermal fixation
thermal reactor
top dead center
torque
turbocharger
a limited usage service, typically 200 hr/yr or less
(e.g., emergency electrical generators)
an application in which the engine is never
propel the structure or
the X&^l
movement of the piston between thesTpoints
"
— see top dead center.
s: sr»sss
gases, thereby reducing HC and CO emissiSn
position of the piston when it is at top of the
cylinder; corresponds to minimum available gas volume.
"" ^ncrla06 th1Ch US6S 3 centr1fV9al compressor to
f'lCoourc OT LilP l nf^r^m ^no JSTV* TL***
»* «VHH BK • •"•••••>* I I I WWII I I ( 'U U1I I rlfrJ
r^firTi fiV*ttor»^\w»^*. .•_.._ i i 3 •» • i • I I 1C
two-stroke
pss ?; f9fs,±s SV^^T^ o
crankshaft) per power stroke (i.e.. £ compute
of
G-9
-------
valve
valve camshaft
valve overlap
volatility
wet
a device which opens into the cylinder to admit air or
air and fuel mixture or to exhaust combustion
products.
a shaft driven off the crankshaft having eccentric
lobes which open the intake and exhaust valves at the
proper time.
the interval, in degrees of crankshaft rotation,
during which the intake and exhaust valve are both
open.
. a measure of the ability of a fuel to evaporate at a
given temperature.
- refers to the existence of water vapor in an exhaust
gas sample from water of combustion and water
contained in the intake air.
6-10
-------
fl7 REPORT NOT
, EPA-450/2-78-125a
|4. TITLE AND SUBTITLE
Stationary Internal
VT*4MS<{«UkWa. ^ . „ _ _ _ . *
REPORT DATA
the reverse before completing)
J3. RECIPIENT'S ACCESSION NO.'
Stated:
'. PERFORMING ORGANIZATION NAME AND ADDRESS'
Standards Development Branch
Emission Standards and Engineering Division
Research Triangle Park, N. C. 27711
12. SPONSORING AGENCY NAME AND ADDRESS
DAA for Air Quality Planning and Standards
Office of Air, Noise and Radiation
u. S. Environmental Protection Agency
15. SUPPLEMENTARY NOTES
TI_ £ _ . . . .
and
5. REPORT DATE
8. PERFORMING ORGANIZATION REPORT NO.
J'1'O. PROGRAM ELEMENT NO."
TTTCONTRACT/GRANT NO."
13. TYPE OF REPORT AND PER.OD COVERED
"14. SPONSORING AGENCV CODE
EPA/200/04
environmental
DESCRIPTORS
KEY WORDS AND DOCUMENT ANALYSIS
Internal Combustion enqines Air
IB. DISTRIBUT ION STATEMENT"
/n1,]?601 Availab1e from EPA library
(MO-35), EPA, Research Triangle Park
TERMS
Oxides Water injection.
1S.|ECURITY OLASS
unrl^ti^ifiorl
zo. SECURITY CLASS (Thispagef
c,m 2220., ,R.V. 4.77) PREV,OUS ED1TION (s OBSOLETE
^U.S.GOVERNMENTPRIhrriNGOFFICE;1979 -6*0-01* 3 9 0 7 REGION NO. 4
. OF PAGES
-------
US, Environmental Protection Agency
Region 5, library (PI-12J)
77 West Jackson Boulevard, 12th Floor
Chicago, II 60604-3590
-------