TECHNICAL PAPER
Sawyer,
ISSN 1047-3289 J.Kean,
Air & Waste
Manage. and
Assoc. Harley
50:1929-1939
Copyright 2000 Air & Waste Management Association
Robert A. Harley
Department of Civil and Environmental Engineering, University of California, Berkeley
ABSTRACT
The use of diesel engines in off-road applications is a significant source of nitrogen oxides (NOx) and particulate
matter (PM10). Such off-road applications include railroad
locomotives, marine vessels, and equipment used for agriculture, construction, logging, and mining. Emissions
from these sources are only beginning to be controlled.
Due to the large number of these engines and their wide
range of applications, total activity and emissions from
these sources are uncertain. A method for estimating the
emissions from off-road diesel engines based on the quantity of diesel fuel consumed is presented. Emission factors are normalized by fuel consumption, and total activity
is estimated by the total fuel consumed.
Total exhaust emissions from off-road diesel equipment (excluding locomotives and marine vessels) in the
United States during 1996 have been estimated to be
1.2 109 kg NOx and 1.2 108 kg PM10. Emissions estimates published by the U.S. Environmental Protection
Agency are 2.3 times higher for both NOx and exhaust
PM10 emissions than estimates based directly on fuel consumption. These emissions estimates disagree mainly due
to differences in activity estimates, rather than to differences in the emission factors. All current emission inventories for off-road engines are uncertain because of the
limited in-use emissions testing that has been performed
on these engines. Regional- and state-level breakdowns
in diesel fuel consumption by off-road mobile sources are
also presented. Taken together with on-road measurements of diesel engine emissions, results of this study suggest that in 1996, off-road diesel equipment (including
IMPLICATIONS
The contribution of off-road diesel equipment to total
emissions of NOx and PM may be lower than suggested
by current emission inventories. As a consequence, control of these sources may not lead to air quality benefits
that are as large as expected.
(1)
consumed. This method is presently used to estimate emissions from marine vessels6 and locomotives,7 and has been
demonstrated in estimating CO and VOC emissions from
on-road motor vehicles,8 as well as NOx and black carbon
particle emissions from heavy-duty diesel trucks.9 An important advantage of the fuel-based approach is that fuel
consumption data can be obtained more readily than information on engine populations, load factors, and annual hours of use.
The objectives of this study were to describe a fuelbased method for determining emissions from off-road
diesel engines, to use this method to estimate emissions
of NOx and PM10 from off-road diesel engines at national
and regional levels, and to compare the fuel-based assessment of emissions with previous inventory estimates.
METHOD
A fuel-based emission inventory for off-road diesel engines
was obtained by multiplying the diesel fuel consumed by
off-road engines by emission factors that are normalized
by fuel consumption (i.e., mass of pollutant emitted per
unit of fuel consumed).
Off-Road Diesel Engine Activity
Previous fuel-based emission inventories for on-road vehicles8,9 have relied on highway fuel tax data to determine
the quantity of fuel used. Diesel fuel used off-road is not
subject to highway taxes, so in this study, diesel fuel use
was determined from an annual survey of companies that
sell distillate fuels to end users. The survey, the Annual
Fuel Oil and Kerosene Sales Report,10 is conducted by the
Energy Information Administration (EIA), part of the U.S.
Department of Energy. The sample used for this survey was
determined from prior EIA survey results in combination
with a survey of ~30,000 companies that sell petroleum
products. Companies in the EIA survey included all refiners and gas plant operators, those companies doing business in four or more states, and those companies that
account for 5% or more of the distillate fuel sold in a particular end-use category in a state. Statistical procedures
were used to determine the number of companies to be
surveyed to achieve a 5% coefficient of variation (COV)
for distillate fuel oil sales at the state level. Overall, ~4700
companies were included in the survey; the response rate
was 93% for 1996. Missing data were imputed by EIA; these
accounted for sales fractions ranging from 0.07% for oil
company use to 15.8% for off-highway use of distillate fuel.
The survey data were reviewed by EIA both manually and
with an automated computer program to detect missing
and outlying data. In addition, preliminary results were
processed through a series of validation procedures to identify and fix potential misreporting of data. Where possible,
survey results were compared and adjusted by EIA with
Volume 50 November 2000
Table 1a. National and regional distillate fuel sales and off-road diesel fuel sales
by end-use category, 1996.
a
Region
b
East Coast
c
Central
d
South Central
e
Rocky Mountains
f
West Coast
Total U.S.
74
59
31
9.0
22
196
9.9
18
12
3.6
5.5
49
Current national regulations limit the sulfur content of diesel fuel used on-road to less than 0.05% by
mass. Use of low-sulfur diesel fuel for off-road purposes
is allowed, but use of high-sulfur diesel fuel is not allowed in on-road engines. Diesel fuel sold for on-road
use is subject to a highway tax, whereas diesel fuel intended for off-road use is not taxed. In some states, onroad diesel fuel use by government vehicles, school buses,
and transit vehicles is not taxed; the on-highway diesel
sales provided by EIA10 include only taxable fuel sales.
Sales of tax-exempt, low-sulfur diesel fuel for government
vehicles, school buses, and so on are included by EIA in
the commercial end-use category.11 For both commercial and industrial end-use categories, total No. 2 distillate fuel sales presented in Table 3 were broken down by
EIA into low-sulfur diesel, high-sulfur diesel, and No. 2
fuel oil. Consequently, for this study, off-road diesel use
in the commercial and industrial end-use categories was
assumed to equal the high-sulfur diesel fuel sales. California is a special case where all on-road and off-road
diesel fuel is required to have low sulfur content; in that
case, all of the tax-exempt low-sulfur fuel was assumed
to be used off-road, as the detail required to separate onroad and off-road diesel fuel use in the commercial sector was not available.
For other end uses, the distillate sales survey data were
not broken down into distillate type. In the case of farm
and military distillate sales, diesel sales were still separated from the other types of distillate fuels. In those cases,
all of the diesel fuel sold was assumed to be used by offroad diesel engines. For the remaining end uses, diesel
fuel was not separated from other distillate fuels, and the
following assumptions were made based on advice from
EIA:11 no use of distillate fuels in the residential or electric
utility sectors occurs in off-road diesel engines; 50% of oil
company distillate fuel oil is diesel (all of which is assumed to be used off-road); and all railroad, vessel bunkering, construction, and other distillate is diesel fuel used
Table 1b. Distribution of off-road diesel fuel sales among end uses (%).
Region
East Coastb
Centralc
South Centrald
Rocky Mountainse
West Coastf
Total U.S.
Commercial
Industrial
Farm
Oil Company
Railroad
Vessel Bunkering
Military
Construction
Other
7.7
3.1
1.7
1.5
7.4
4.0
12.4
7.6
4.7
3.2
12.3
8.0
13.5
34.4
22.1
27.9
19.5
25.0
0.2
0.5
7.5
1.8
1.7
2.4
21.7
30.2
17.2
47.4
19.2
25.3
19.6
10.2
29.4
<0.1
16.4
16.8
1.6
0.4
4.9
0.3
7.1
2.5
19.3
11.5
9.2
14.5
11.7
12.7
4.2
2.2
3.2
3.5
4.8
3.2
Includes all diesel fuel sold for on-road and off-road use, as well as fuel oils No. 1, No. 2, and No. 4; bPetroleum Administration for Defense (PAD) District I: CT, DE, DC, FL, GA, ME,
MD, MA, NH, NJ, NY, NC, PA, RI, SC, VT, VA, and WV; cPAD District II: IL, IN, IA, KS, KY, MN, MI, MO, NE, ND, OH, OK, SD, TN, and WI; dPAD District III: AL, AR, LA, MS, NM, and
TX; ePAD District IV: CO, ID, MT, UT, and WY; fPAD District V: AK, AZ, CA, HI, NV, OR, and WA.
Volume 50 November 2000
Commercial
Industrial
Farm
Oil company
Railroad
Vessel bunkering
Military
Construction
Other
a
EIA includes mining equipment in the industrial category, whereas EPA groups mining and construction equipment together; bMost recreational boats are gasoline powered; large ocean-going vessels primarily use residual fuel oil. Engine emissions from
the use of gasoline and residual fuel oil are not included in the present study.
deterioration in emission factors is not addressed here because the impact of increasing age on emissions from uncontrolled diesel engines is reported to be negligible for
NOx emissions and small for PM10 emissions.3,7
Off-Road Diesel Equipment. Most of the emission factors
for diesel engines used in the construction, agriculture,
mining, and logging sectors that have been reported are
for steady-state tests on new engines and have been provided by engine manufacturers.3 Correction factors typically are applied to these emission factors by EPA to adjust
for non-steady-state operating conditions.3 Others report
that most off-road equipment operates at quasi-steadystate conditions and argue that correction factors are not
needed.13 Information that can be used to develop fleetaveraged emission factors for these engines is contained
in EPAs NONROAD model and its associated documentation.14 The NONROAD model estimates fuel consumption
and emissions of NOx and PM10 by off-road engines. In
NONROAD, the range of emission factors for engines that
contribute significantly to overall fuel consumption is 38
65 g NOx per kg fuel and 2.810.2 g PM10 per kg fuel. PM10
emissions from these engines are high compared with
other off-road diesel engines and show significant variability. Table 5 presents fleet-averaged emission factors for
off-road equipment, as well as the emission factors for
major types of off-road equipment.
Locomotives. The development of fleet-averaged emission
factors for locomotives is easier than for other types of
off-road engines because there are effectively only two
basic engines that comprise the entire U.S. locomotive
fleet.15 These engines are used for three primary applications: line-hauling, switching, and passenger transportation. The duty cycles, and consequently the emission
factors, vary depending on the application. EPA has estimated that ~85% of the total locomotive fuel is consumed
by national freight line-haul locomotives, 7% by national
freight switching locomotives, 5% by local and regional
freight locomotives, and 3% by passenger locomotives.16
Little is known about the duty cycles and emissions characteristics of local and regional freight locomotives,7 but
it is assumed here that they are most similar to switch
locomotive emissions. Local and regional freight locomotives, like switch locomotives, are typically older and have
lower power duty-cycles compared with national freight
line-haul locomotives.7 Consequently, fuel used by local
and regional freight locomotives has been grouped with
fuel used by switch locomotives.
Average emission factors for each locomotive engine
application have been presented by EPA.16 Typically, switch
locomotives have higher fuel-based emission factors than
line-haul locomotives do. Additional emission factors for
Volume 50 November 2000
Table 3. National 1996 distillate fuel salesa and estimated fractions that are diesel and used in off-road engines.
Description
RESULTS
The fuel-based methodology discussed above
was applied to the United States at the national
and regional levels. Total diesel fuel used by
Fuel oil
off-road engines is shown in Table 1. The ratio
of diesel fuel used off-road relative to the total
No. 2
Diesel fuel
volume of distillate fuels sold ranged from
13 to 40% at the regional level and was 25% at
the national level in 1996. At the national level,
Fuel oil
2.8 1010 L diesel fuel was sold for use in offroad equipment, 1.2 1010 L was sold for railNo. 4
Diesel fuel
road use, and 8.2 109 L was sold for marine
vessel use. For comparison, ~1.0 1011 L diesel
Fuel oil
fuel was consumed by highway vehicles.10
National emissions estimates for off-road
diesel equipment, railroad locomotives, and
marine vessels based on fuel sales are shown
a
Most ocean-going marine engines use residual fuel oil. Residual fuel oil is also used for electric power
in Figure 1a for NOx and Figure 1b for PM10.
generation, commercial space heating, and various industrial purposes.
For comparison, EPA Emissions Trends Report
estimates2 are also provided in Figure 1. EPAs
nationwide estimate of emissions from off-road diesel
Low-speed diesel engines typically are used by oceanequipment is 2.3 times the fuel-based estimate for both
going vessels burning residual fuel oil.20 High-speed dieNOx and exhaust PM10 emissions. In contrast, the ratios
sel engine emissions data are limited, leading, in a previous
21
of EPA emission estimates to the fuel-based estimates preinvestigation, to the use of medium-speed engine emissented here are 1.1 for NOx and 1.2 for PM10 from locosion factors to estimate emissions from high-speed engines. In the present study, the emission factors presented
motives, and 0.34 for NOx and 1.8 for PM10 from diesel
marine engines.
Figure 2 presents emissions estimates from off-road
Table 5. Diesel engine exhaust emission factors.a
diesel equipment for five regions of the United States,
using the emission factors described above. Emissions at
Engine Application
NO (g/kg)
PM (g/kg)
x
10
the state level can be estimated using fuel consumption
by off-road engines presented in the appendix in combiOn-road diesel trucksb,c
42 5
2.5 0.2
nation with emission factors given in Table 5. Emissions
d
Off-road equipment (fleet average)
48 6
5.1 2.3
from locomotives and marine vessels were not modeled
Commercial
45
5.9
at the regional level due to uncertainties in the regional
Industrial
45
6.2
breakdown of fuel consumption by these engines.
Farm
53
3.8
No. 1
Subcategory
Diesel fuel
46
43
75 9
74
81
55 13
40
57
87
5.5
4.7
1.9 0.3
1.8
2.1
1.3 0.6
1.2
1.2
7.6
A diesel fuel density of 0.85 kg/L is assumed for all cases; bOn-road measurements
from 1997 in a San Francisco Bay area highway tunnel;23 cExhaust emission factor for
particulate matter is PM for on-road diesel trucks; dOff-road equipment emission
2.5
factors were obtained from EPAs NONROAD model;14 eSlow-speed emission factors
were measured on engines consuming residual fuel oil.
1934 Journal of the Air & Waste Management Association
DISCUSSION
As shown in Figure 1, significant differences exist between
the emissions estimates for off-road diesel equipment and
diesel marine engines, whereas emissions estimates for
locomotives are in reasonable agreement.
Off-Road Diesel Equipment
EPA estimates2 of both NOx and PM10 emissions from offroad diesel equipment are more than twice the values presented here based on national diesel fuel consumption, even
though emission factors used to derive both inventory estimates were similar. This means that differences in activity for these engines are the primary reason for differences
Volume 50 November 2000
Figure 1. National off-road diesel engine emissions of (a) NOx and (b) PM10 for 1996 from the present study with comparisons to EPA Emissions
Trends Report estimates.2
Volume 50 November 2000
Figure 2. Regional breakdown of off-road diesel equipment emissions. The regions are defined in the footnotes to Table 1.
in the estimated emissions. EPAs NONROAD model, currently under development, estimates national diesel fuel
consumption by off-road diesel equipment to be 6.3 1010 L
in 1996. This is 2.2 times the value derived from EIA fuel
sales surveys,10 suggesting that the total activity of off-road
diesel equipment is substantially overestimated by EPAs
use of eq 1 with the information currently available. Even
if it was assumed that all tax-exempt diesel fuel, regardless
of sulfur content, was used off-road, the ratio of fuel use by
off-road diesel equipment as estimated by EPAs NONROAD
model to the value derived from fuel sales surveys would
still be 1.9. The differences between EPA and fuel-based
diesel engine activity estimates may be larger than stated
here, because some distillate fuel oil other than diesel fuel
has been included in the present studys estimates, as has
diesel fuel used by stationary off-road engines (e.g., stationary generators, pumps, compressors, and welders).
Another factor complicating the assessment of engine activity is the potential for crossover in fuel sold for on-road
versus off-road use. Entities operating the same engines both
on- and off-road may choose to use taxable diesel fuel at
all times to simplify fuel supply issues, and fuel that is exempt from highway taxes at times may be used illegally in
on-road engines.
Figure 2 shows the regional breakdown of off-road
diesel engine equipment emissions. Emission factors may
1936 Journal of the Air & Waste Management Association
vary depending on the mix of equipment types and engines used in a given region, but for these large regions,
the variation in the emission factors is likely to be small.
EPAs NONROAD model also suggests that fleet-averaged
emission factors do not vary significantly by region.
Locomotives
All of the emissions estimates for locomotive engines were
in good agreement. This was expected, because EPA uses a
fuel-based approach to estimate emissions from these engines, instead of relying on the methodology presented in
eq 1. These EPA emission estimates are based on fuel consumption data obtained from the Association of American
Railroads,7 which indicate that 1.36 1010 L diesel fuel was
used by locomotives. This is 10% higher than the estimated
1.24 1010 L diesel fuel used by railroads during 1996 as
given by EIA,10 supporting the assumption made earlier that
virtually all fuel sold to railroads is used in locomotives.
Given uncertainties resulting from the small number of
engines tested and uncertainties in fuel sales, emissions
estimates for locomotives presented here are effectively the
same as EPA estimates.
Marine Vessels
For diesel marine vessels, EPA emission estimates2 are 0.34
times our fuel-based estimate for NOx and 1.8 times our
Volume 50 November 2000
Figure 3. Breakdown of mobile source NOx emissions by engine type and application, showing (a) current official inventory estimates for 19962 and
(b) fuel-based emission estimates developed in the present study. Total NOx emissions = 1.1 1010 kg in both cases.
Volume 50 November 2000
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
REFERENCES
1.
2.
3.
4.
5.
6.
7.
8.
Construction
Total Off-Road
d
Equipment
Total Distillate
e
Sales
1.34
0.01
0.02
0.0
0.28
0.38
0.01
0.09
0.01
0.00
0.02
0.08
0.11
0.10
0.00
0.07
0.01
0.13
0.01
1.91
0.03
0.03
0.001
0.42
0.19
0.05
0.09
0.09
0.02
0.10
0.18
0.21
0.20
0.01
0.08
0.01
0.18
0.03
5.82
0.05
0.07
0.01
0.95
0.86
0.12
0.31
0.16
0.04
0.18
0.31
0.68
0.54
0.03
0.32
0.04
0.72
0.43
74.1
3.58
0.61
0.38
6.25
6.63
2.42
3.49
5.53
1.27
5.66
11.5
5.31
9.88
0.97
2.47
0.93
5.73
1.48
34.5
3.35
3.92
1.74
1.34
2.22
2.92
1.83
2.90
1.27
0.51
4.83
2.05
0.54
2.86
2.26
6.08
0.76
0.34
0.74
0.61
0.12
0.22
0.62
0.22
0.65
0.39
0.33
0.28
0.28
0.22
0.31
2.03
0.24
0.09
0.12
0.10
0.14
0.16
0.17
0.14
0.06
0.02
0.25
0.11
0.03
0.20
0.19
10.5
1.20
0.60
0.91
0.76
0.97
0.56
0.90
0.52
0.74
0.44
0.88
0.50
0.32
0.61
0.63
59.2
6.06
5.69
3.21
2.69
4.49
4.74
3.92
4.42
2.71
1.35
7.19
3.25
1.06
4.38
4.07
South Central
Alabama
Arkansas
Louisiana
Mississippi
New Mexico
Texas
17.1
2.45
1.88
1.92
1.51
1.19
8.17
2.70
0.12
0.35
0.20
0.30
0.06
1.67
1.12
0.22
0.05
0.35
0.06
0.06
0.39
6.52
0.68
0.48
1.32
0.53
0.20
3.30
31.23
3.68
2.64
6.29
2.31
1.58
14.7
Rocky Mountains
Colorado
Idaho
Montana
Utah
Wyoming
4.14
1.03
0.71
0.53
0.97
0.90
1.00
0.33
0.23
0.32
0.07
0.05
0.52
0.20
0.06
0.11
0.08
0.07
1.89
0.59
0.44
0.46
0.23
0.18
8.97
2.36
1.51
1.56
1.58
1.98
West Coast
Alaska
Arizona
California
Hawaii
Nevada
Oregon
Washington
14.6
0.13
2.06
8.01
0.15
0.93
1.46
1.81
1.08
0.00
0.07
0.72
0.02
0.02
0.10
0.16
0.65
0.04
0.09
0.32
0.02
0.04
0.04
0.11
3.57
0.26
0.37
1.77
0.10
0.32
0.28
0.47
22.3
1.40
2.59
10.7
0.77
1.51
2.24
3.02
United States
102
12.2
6.23
28.3
196
Region
On-Road
Farm
East Coast
Connecticut
Delaware
District of Columbia
Florida
Georgia
Maine
Maryland
Massachusetts
New Hampshire
New Jersey
New York
North Carolina
Pennsylvania
Rhode Island
South Carolina
Vermont
Virginia
West Virginia
31.7
0.79
0.22
0.08
3.64
5.09
0.51
1.35
1.19
0.23
2.02
3.50
3.01
4.43
0.17
1.73
0.36
2.87
0.50
Central
Illinois
Indiana
Iowa
Kansas
Kentucky
Michigan
Minnesota
Missouri
Nebraska
North Dakota
Ohio
Oklahoma
South Dakota
Tennessee
Wisconsin
a
Diesel fuel that was taxed for use on-road; bDiesel fuel used by farm equipment; cDiesel fuel used by construction equipment; dDiesel fuel used by all off-road diesel equipment, including farm and construction
e
10
equipment; Total distillate fuel oil sales.