Journal of the Air & Waste Management Association

ISSN: 1047-3289 (Print) (Online) Journal homepage: http://www.tandfonline.com/loi/uawm20

A Fuel-Based Assessment of Off-Road Diesel

Engine Emissions
Andrew J. Kean , Robert F. Sawyer & Robert A. Harley
To cite this article: Andrew J. Kean , Robert F. Sawyer & Robert A. Harley (2000) A Fuel-Based
Assessment of Off-Road Diesel Engine Emissions, Journal of the Air & Waste Management
Association, 50:11, 1929-1939, DOI: 10.1080/10473289.2000.10464233
To link to this article: http://dx.doi.org/10.1080/10473289.2000.10464233

Published online: 27 Dec 2011.

Submit your article to this journal

Article views: 634

View related articles

Citing articles: 31 View citing articles

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=uawm20
Download by: [186.68.243.47]

Date: 21 April 2016, At: 08:03

TECHNICAL PAPER

Sawyer,
ISSN 1047-3289 J.Kean,
Air & Waste
Manage. and
Assoc. Harley
50:1929-1939
Copyright 2000 Air & Waste Management Association

A Fuel-Based Assessment of Off-Road Diesel Engine Emissions

Andrew J. Kean and Robert F. Sawyer
Department of Mechanical Engineering, University of California, Berkeley

Downloaded by [186.68.243.47] at 08:03 21 April 2016

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.

Volume 50 November 2000

agriculture, construction, logging, and mining equipment,

but not locomotives or marine vessels) was responsible
for 10% of mobile source NOx emissions nationally,
whereas on-road diesel vehicles contributed 33%.
INTRODUCTION
While on-road vehicles such as cars and trucks have long
been recognized as important mobile sources of air pollution, less attention has been paid to controlling off-road
mobile source emissions. The importance of off-road
sources is expected to grow as on-road vehicle emissions
are controlled. Off-road engine applications include aircraft, locomotives, marine vessels, recreational vehicles,
and equipment used for agriculture, construction, logging,
mining, lawn and garden maintenance, and so on. The
present study focuses on diesel engines, as these are the
dominant off-road mobile source of nitrogen oxides (NOx)
and exhaust particulate matter less than 10 m in diameter (PM10). Emissions from aircraft (primarily powered
by kerosene-type jet fuel and gasoline) and residential lawn
and garden equipment (primarily powered by gasoline)
are not considered here. All other off-road engine applications include at least some use of diesel fuel.
Off-road mobile sources are defined by the U.S. Environmental Protection Agency (EPA) as engines used offroad that are moved at least once within a 12-month
period.1 An engines mobility is important because regulatory agencies generally separate the emissions of air
pollutants into two main source categories: stationary
sources and mobile sources. Making this distinction for
many off-road sources has proved difficult because engines used in generators, pumps, compressors, and welders may be either mobile or stationary. This means that
two of the same engines being used for similar purposes
may be regulated differently.
Total U.S. emissions during 1996 from off-road diesel
engines including locomotives and diesel marine sources
have been estimated by EPA to be 3.6 109 kg NOx and
3.2 108 kg exhaust PM10.2 Throughout this investigation,
Journal of the Air & Waste Management Association 1929

Downloaded by [186.68.243.47] at 08:03 21 April 2016

Kean, Sawyer, and Harley

NOx emissions are reported as NO2 equivalents (i.e., a molecular weight of 46 g/mol is assumed), even though most
of the NOx is expected to be emitted as NO. Nationally,
off-road diesel engines have been estimated by EPA to account for 34% of mobile source NOx emissions and 48%
of mobile source exhaust PM10 emissions. Off-road diesel
equipment used in agriculture, construction, industrial,
commercial, mining, and logging sectors contribute more
to these totals than locomotives and marine vessels do.
With locomotive and marine vessel emissions excluded,
off-road diesel equipment accounts for 25% of mobile
source NOx emissions and 42% of mobile source exhaust
PM10 emissions, according to EPA.2
Emissions estimates for these sources are based primarily on a methodology described in the Nonroad Engine and Vehicle Emissions Study,3 a major assessment of
off-road engine activity and emissions at the national and
local levels, including emissions from over 80 different
types of equipment. Emissions, E, were estimated by EPA
for most off-road engine applications as follows:
E = N P LF A EF

(1)

where N is the engine population (number of engines in

use); P is the average rated power of these engines; LF is
the load factor, which indicates the typical operating load
of the engines relative to their average rated power; A is
activity (in average annual hours of use per engine); and
EF is an emission factor expressed in units of work done
by the engine (e.g., g/kW-hr or g/bhp-hr). EPA is currently
developing a computer model known as NONROAD to
estimate emissions from off-road vehicles, again relying
on eq 1 as the framework.4 Emissions from off-road engines in Europe have been estimated using a similar approach. It was found that off-road diesel engine emissions
accounted for 23% of mobile source NOx emissions and
40% of mobile source exhaust PM10 emissions in the European Union in 1990.5
The large number of off-road diesel engines and applications makes it difficult to quantify the factors appearing in eq 1 accurately. Unlike on-road vehicles, there
is no registration database for most off-road engines. This
makes it difficult to estimate engine populations. Load
factors depend on specific engine applications, which are
numerous and varied. Typically, engine power, population, activity, and load factors for each application are
estimated from engine manufacturer sales surveys and
surveys of users of these engines.
It is also possible to estimate emissions from off-road
diesel engines using a top-down approach that is based
on fuel consumption. In this approach, emission factors
are normalized by fuel consumption, and engine activity is estimated using the total quantity of diesel fuel
1930 Journal of the Air & Waste Management Association

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

Downloaded by [186.68.243.47] at 08:03 21 April 2016

Kean, Sawyer, and Harley

alternate source data such as known regional distillate fuel
oil production at refineries, accounting for imports, exports,
and changes in reserves. The adjustment factors used by
EIA for this purpose were 0.98 for East Coast, 0.98 for Central, 0.85 for South Central, 1.30 for Rocky Mountain, and
0.66 for West Coast regions (these regions are defined in
Table 1). The same adjustment factor was applied to all offroad fuel uses except electric utilities, for all states within a
given region.
A list of distillate fuel end use categories with typical
off-road equipment that falls into each end use is presented in Table 2. National distillate fuel oil sales are broken down into these end uses, as shown in Table 3. The
fractions of these sales that are used by mobile off-road
diesel engines also are included in Table 3.
Distillate fuel oil is a general classification for one
of the petroleum fractions produced by conventional refinery distillation operations. There are three types of
distillate fuel oil: No. 1, No. 2, and No. 4, with No. 1
having the lowest distillation temperatures and No. 4
the highest. Most of the diesel fuel used in the United
States is a form of No. 2 distillate fuel oil, although No. 1
and No. 4 diesel fuels are also used. Further description
of the different types of distillate fuel oil is provided in
Table 4.

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.

Distillate Sales (10 L)

Off-Road Diesel Fuel (10 L)

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

Journal of the Air & Waste Management Association 1931

Kean, Sawyer, and Harley

Table 2. Off-road diesel fuel end-use categories and associated engine applications.
End-Use Category

Examples of Engine Applications

Commercial
Industrial
Farm
Oil company
Railroad
Vessel bunkering

Welders and pressure washers

Forklifts, aerial lifts, and mining equipmenta
Tractors and combines
Bore/drill rigs and other oil field equipment
Locomotives and railway maintenance equipment
Recreational boats, tugboats, ferries, and
ocean-going vesselsb
Military trucks and other equipment
Cranes, paving, and earth-moving equipment
Logging equipment

Military
Construction
Other
a

Downloaded by [186.68.243.47] at 08:03 21 April 2016

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.

by off-road engines. All fuel sales to railroads were assumed

to be used in locomotives, and all vessel-bunkering sales
were assumed to be used by main propulsion engines in
marine vessels.
The fuel sales survey data used in this study are subject to sampling errors because fuel sales are determined
using a sample of fuel vendors rather than an all-inclusive census. Estimates of these sampling errors, stated as
coefficients of variation, have been calculated by EIA. Here,
the COV is defined as the standard deviation of the estimated nonresidential retail distillate fuel sales volume
divided by the total nonresidential retail distillate fuel
volume. In 1996, four of five regions of the United States
had COVs in fuel sales of 2.2% or less. The remaining
region (South Central) had a COV of 13%, due to large
sampling errors in Texas. The U.S. average COV was 3%.
Therefore, uncertainties in fuel sales arising from statistical sampling of fuel vendors appear to be small at the
national and regional levels compared with other uncertainties involved in this method.
Emission Factors
Unlike on-road vehicles, where emissions most often are
reported in mass per distance traveled, emission factors for
off-road engines typically are reported in mass per unit of
brake work output from the engine. In most cases, emission factors are developed using an engine dynamometer
test with a duty-cycle consisting of a set of steady-state
modes that are time-weighted to represent in-use conditions.12 Due to engine size, locomotive and marine engine
emissions are sometimes measured in-use. By dividing
the measured emission factor by brake-specific fuel consumption determined during the same test, the emissions
can be normalized to fuel consumption. The question of
1932 Journal of the Air & Waste Management Association

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

Kean, Sawyer, and Harley

Downloaded by [186.68.243.47] at 08:03 21 April 2016

Table 3. National 1996 distillate fuel salesa and estimated fractions that are diesel and used in off-road engines.

are not typically used for dedicated

domestic shipping. Marine diesel enEnd Use
Distillate Type
Distillate Sales
% of Sales Included
gines are used on vessels ranging from
(109 L)
As Off-Road Diesel
recreational boats to ocean-going
tankers. Because of the wide range of
Residential
No. 1 distillate
0.7
0
engines used, marine diesel engines
No. 2 distillate
24.7
0
are often subdivided based on engine
Commercial
No. 1 distillate
0.5
0
speed, with slower, larger engines genNo. 2 distillate
11.2
15
b
erally being more powerful.
No. 2 fuel oil
5.7
0
b
Another difficulty in estimating
Low-sulfur diesel
3.8
6
b
emissions from marine engines is deHigh-sulfur diesel
1.7
100
termining what type of fuel they use.
No. 4 distillate
1.3
0
Industrial
No. 1 distillate
0.2
0
Very large marine diesel engines typiNo. 2 distillate
7.5
49
cally use less expensive residual fuel
No. 2 fuel oilb
1.7
0
oil or a combination of diesel fuel and
Low-sulfur dieselb
2.0
6
residual fuel oil to reduce their fuelHigh-sulfur dieselb
3.8
100
ing costs.19,21 Therefore, only emisNo. 4 distillate
0.2
0
sions from combustion of diesel fuel
Farm
Diesel
12.2
100
in main propulsion marine engines
Other distillate
0.4
0
are addressed in this investigation.
c
Electric utility
Unspecified
2.6
0
Residual fuel contains significantly
Oil company
Unspecified
2.4
50
more impurities and is more viscous
Railroad
Unspecified
12.4
100
than diesel fuel, and consequently the
Vessel bunkering
Unspecified
8.3
100
emissions characteristics are different.
On-highway
Diesel
102
0
Military
Diesel
1.2
100
Both NOx and PM10 emissions are
Other distillate
0.2
0
higher from the combustion of reConstruction
Unspecified
6.2
100
sidual fuel versus that of diesel fuel.20,21
Otherd
Unspecified
1.6
100
A final difficulty in determinTotal distillate sales
196
25
ing emission factors is that very few
emissions tests have been performed
a
Source of the fuel sales data is EIA;10 bNo. 2 fuel oil, low-sulfur diesel, and high sulfur diesel are all subcategories of No. 2
on marine engines. Emissions data
distillate; cType of distillate fuel was not specified on fuel sales survey form; dFuel sales for use in logging equipment is
from primarily 18 marine diesel enincluded here.
gines have been used by EPA to estimate baseline emissions from all
marine diesel engines.20 Information regarding the brakeline-haul and passenger locomotives are available in the
7,15,17,18
literature.
specific fuel consumption for these engines was not proIn total, data are available from emissions
tests on ~25 locomotive engines. Emission factors for lovided, so to develop fuel-based emission factors, fuel
consumption data from similar land-based engines were
comotives all fall in the range of 60100 g NOx per kg fuel
used.13,20,21 This is appropriate because the emissions test
consumed and 1.42.4 g PM10 per kg fuel. A fleet-averaged emission factor, presented in Table 5, was determined
duty cycles of marine and land-based diesel engines have
by weighting EPA emission factors16 for each locomotive
been shown to be similar.20 Marine diesel emission facapplication by the distribution of fuel use mentioned
tors ranged from 30 to 100 g NOx per kg fuel and 0.7 to
above; the resulting emission factors are therefore heavily
2.1 g PM10 per kg fuel. Small, high-speed engines had lower
influenced by values used to represent line-haul locomoNOx emission factors and large engines had higher NOx
tive emissions.
emission factors, but no such trend was present for PM10
emissions.
Marine Vessels. It is more difficult to estimate fleet-average
Lloyds Register,22 a worldwide ship classification soemission factors for marine diesel engines than for locociety, has performed emissions tests on 50 vessels, with
motives for several reasons. First, the types of engines used
results similar to EPA baseline emissions.21 Lloyds has esand number of applications are significantly larger in the
timated fleet-average emissions factors to be 57 g NOx
marine sector. Most locomotive engines have a rated power
per kg fuel and 1.2 g PM10 per kg fuel for medium-speed
between 1100 and 3000 kW, whereas marine diesel engines
marine diesel engines and 87 g NOx per kg fuel and 7.6 g
range from <100 to 30,000 kW, though the largest engines
PM per kg fuel for low-speed marine diesel engines.21,22
Volume 50 November 2000

Journal of the Air & Waste Management Association 1933

Kean, Sawyer, and Harley

Table 4. Classification of distillate fuels.10
Distillate Type

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

Downloaded by [186.68.243.47] at 08:03 21 April 2016

Subcategory

in Table 5 were used for all marine engines

burning diesel fuel.

Diesel fuel

Construction and Mining

Logging
Locomotives7,15-18 (fleet average)
Line-Haul
Switch
Marine vessels (fleet average) 13,20-22
High-speed
Medium-speed
Slow-speede

Used in high-speed engines subject to wide variations

in speed and load. Includes type C-B diesel fuel used in
city buses.
Light distillate intended for use in vaporizing pot-type
burners.
Used in high-speed engines generally operated under
uniform speed and load conditions. Includes type R-R
used by railroad locomotives and type T-T used by diesel trucks. Low-sulfur fuel contains <0.05% sulfur by weight.
Used in atomizing-type burners for domestic space heating
and moderate capacity commercial/industrial-type burners.
Used in low- and medium-speed diesel engines such
as marine engines.a
Used in large burner installations not equipped with fuel
preheating facilities, especially industrial plants. It is a
blend of distillate and residual fuel oils.

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

Downloaded by [186.68.243.47] at 08:03 21 April 2016

Kean, Sawyer, and Harley

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

Journal of the Air & Waste Management Association 1935

Downloaded by [186.68.243.47] at 08:03 21 April 2016

Kean, Sawyer, and Harley

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

Downloaded by [186.68.243.47] at 08:03 21 April 2016

Kean, Sawyer, and Harley

fuel-based estimate for PM10. It is not clear why EPA estimates are lower for NOx but higher for PM10. Corbett
and Fischbeck6,21 have presented detailed analyses of
emissions from marine engines and reported similar findings when their estimates were compared with previously
published EPA emission estimates. One source of uncertainty in these comparisons arises due to the international nature of shipping via ocean-going commercial
vessels. In this case, fuel can be consumed far away from
the point of sale. However, ocean-going vessels with large
low-speed engines typically burn heavier residual fuels,
not diesel fuel, in their propulsion engines.20 By focusing only on diesel fuel use in the marine sector, we emphasize emissions from river and lake, harbor, and
near-shore vessels, which tend to be smaller and have
medium- or high-speed diesel engines. Readers seeking
a more detailed analysis of marine vessel emissions are
referred to the work of Corbett and Fischbeck.6,21
Overall Mobile Source Inventory
In contrast to the results for off-road engines, based on a
similar fuel-based analysis, we found that on-road heavyduty diesel engine NOx emissions were approximately twice
those of past EPA inventory estimates. On-road measurements of diesel engine emissions in tunnels and by remote
sensing have been reviewed by Kirchstetter
et al.,23 who found an emission factor of 42 5 g NOx per kg
fuel to be near the middle of the range of values reported.
Using this emission factor in combination with the taxable on-road diesel fuel sales from EIA, we estimated national emissions from on-road heavy-duty diesel trucks in

1996 to be 3.6 109 kg NOx, whereas EPA estimated NOx

emissions from these vehicles to be 1.8 109 kg.2
A breakdown of EPAs national mobile source NOx
emission inventory2 is presented in Figure 3a. Note that
aircraft emissions include only those associated with landing and takeoff operations, idling, taxiing, and climbing
below 910 m (3000 feet) altitude. In Figure 3b, fuel-based
emissions estimates derived in the present study for offroad diesel equipment, locomotives, marine vessels, and
on-road heavy-duty vehicles have been used in place of
EPA estimates. As seen by comparing Figures 3a and 3b,
NOx emissions from off-road diesel equipment are less
important than current inventories suggest, whereas emissions from on-road diesel engines appear to be underestimated. Also, a fuel-based analysis showed that locomotives
were responsible for 34% of NOx emissions from off-road
diesel sources, compared with 23% when EPA estimates
are used. Despite large differences in emissions estimates
for both on-road and off-road diesel engines, overall mobile source NOx emissions calculated in the present study
agree with previous inventories. The breakdown between
on-road and off-road contributions is still important because different emission standards, fuels, and timetables
for controlling emissions apply in these sectors; this breakdown is important when determining the most effective
control strategies to be pursued.
CONCLUSIONS
A fuel-based approach was used to estimate emissions of
NOx and PM10 from U.S. off-road diesel engines. Fuel sales
data can be obtained more readily than information on

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

Journal of the Air & Waste Management Association 1937

Downloaded by [186.68.243.47] at 08:03 21 April 2016

Kean, Sawyer, and Harley

populations, average load factors, and annual hours of
use, which is required in current methods of estimating
emissions. For 1996, U.S. emissions by all off-road diesel
engines using this fuel-based method were estimated to
be 2.3 109 kg of NOx and 1.5 108 kg of PM10. Emissions
from off-road diesel equipment were 1.2 109 kg of NOx
and 1.2 108 kg of PM10. Past estimates of emissions by
these engines were higher by a factor of 2.3 for both NOx
and exhaust PM10. These differences appear to be the result of overstated engine activity in the off-road diesel
equipment sector; similar emission factors were used in
both methods. Emissions from locomotives estimated in
this study agree with EPA estimates, while unexplained
differences in marine vessel emissions remain.
EPA estimates suggest off-road diesel equipment was
responsible for 25% of national mobile source NOx emissions in 1996. The estimate was reduced to 10% when fuelbased estimates of on- and off-road engine emissions were
used instead. For all applications of off-road diesel engines,
emissions estimates will improve as further engine activity
and emissions test data are obtained. Based on currently
available off-road engine emissions data, control of off-road
diesel equipment emissions may not lead to air quality
benefits as large as have been projected. The use of diesel
fuel in off-road engines is nevertheless a significant source
of NOx and PM10, and controlling these emissions will therefore be important to improving air quality.
ACKNOWLEDGMENTS
This material was based upon work supported by the National Science Foundation under Grant No. BES-9623385.
We thank Alice Lippert of EIA, and Greg Janssen, Joseph
Somers, and Alan Stout of EPA for their helpful comments
and suggestions.

9.
10.
11.
12.

13.
14.
15.
16.
17.
18.
19.
20.
21.

22.

23.

Dreher, D.B.; Harley, R.A. A Fuel-Based Inventory for Heavy-Duty

Diesel Truck Emissions; J. Air & Waste Manage. Assoc. 1998, 48, 352358.
Fuel Oil and Kerosene Sales 1996; DOE/EIA-0535(96); Energy Information Administration, U.S. Department of Energy: Washington DC,
1997. http://www.eia.doe.gov, accessed June 1998.
Lippert, A. Energy Information Administration, U.S. Department of
Energy, Washington, DC. Personal communication, 1998.
Pollack, A.K.; Lindhjem, C.E. NONROAD MOBILE Emissions Modeling;
ENVIRON International Corporation, Novato, CA, and Office of
Mobile Sources, U.S. Environmental Protection Agency: Ann Arbor,
MI, 1997.
Documentation of Input Factors for the New Off-Road Mobile Source Emissions Inventory Model; Report to the California Air Resources Board by
Energy and Environmental Analysis, Inc.: Arlington, VA, 1997.
Exhaust Emission Factors for Nonroad Engine ModelingCompressionIgnition; Report No. NR-009a; Office of Mobile Sources, U.S. Environmental Protection Agency: Ann Arbor, MI, 1998.
Fritz, S.G.; Cataldi, G.R. Gaseous and Particulate Emissions from Diesel Locomotive Engines; Transactions of the ASME 1991, 113, 370376.
Emission Factors for Locomotives, Technical Highlights; Report No.
EPA420-F-97-051; Office of Mobile Sources, U.S. Environmental Protection Agency: Ann Arbor, MI, 1997.
Fritz, S.G. Exhaust Emissions from Two Intercity Passenger Locomotives; Transactions of the ASME 1994, 116, 774-783.
Fritz, S.G.; Markworth, V.O.; Cataldi, G.R. Exhaust Emissions from Inuse Locomotives; 95-ICE-4, The American Society of Mechanical Engineers: New York, 1995.
Woodward, J.B. Low Speed Marine Diesel; Robert Kreiger Publishing
Company: Malabar, FL, 1988.
Draft Regulatory Impact Analysis: Control of Emissions from Compression-Ignition Marine Engines; Office of Mobile Sources, U.S. Environmental Protection Agency: Ann Arbor, MI, 1998.
Corbett, J.J.; Fischbeck, P.S. Commercial Marine Emissions Inventory for
EPA Category 2 and 3 Compression Ignition Marine Engines in United
States Continental and Inland Waterways; Prepared for U.S. Environmental Protection Agency, Order No. 8A-0516-NATX; Carnegie Mellon
University: Pittsburgh, PA, 1998.
Carlton, J.S.; Danton, S.D.; Gawen, R.W.; Lavender, K.A.; Mathieson,
N.M.; Newell, A.G.; Reynolds, G.L.; Webster, A.D.; Wills, C.M.R.;
Wright, A.A. Marine Exhaust Emissions Research Programme; Lloyds
Register Engineering Services: London, 1995.
Kirchstetter, T.W.; Harley, R.A.; Kreisberg, N.M.; Stolzenburg, M.R.;
Hering, S.V. On-Road Measurement of Fine Particles and Nitrogen
Oxide Emissions from Light- and Heavy-Duty Motor Vehicles; Atmos.
Environ. 1999, 33, 2955-2968.

REFERENCES
1.
2.
3.
4.
5.
6.
7.
8.

Nonroad Engine Population Estimates; Report No. NR-006a; Office of

Mobile Sources, U.S. Environmental Protection Agency: Ann Arbor,
MI, 1998.
National Air Pollutant Emission Trends Update, 1970-1997; Report No.
EPA 454/E-98-007; Office of Air Quality Planning and Standards, U.S.
Environmental Protection Agency: Research Triangle Park, NC, 1998.
Nonroad Engine and Vehicle Emission Study; Report No. EPA-21A-2001;
Office of Mobile Sources, U.S. Environmental Protection Agency: Ann
Arbor, MI, 1991.
Average Life, Annual Activity, and Load Factor Values for Nonroad Engine
Emissions Modeling; Report No. NR-005a; Office of Mobile Sources,
U.S. Environmental Protection Agency: Ann Arbor, MI, 1998.
Samaras, Z.; Zierock, K.H. Off-Road Vehicles: A Comparison of Emissions with Those from Road Transport; Sci. Total Environ. 1995, 169,
249-255.
Corbett, J.J.; Fischbeck, P.S. Emissions from Ships; Science 1997, 278,
823-824.
Locomotive Emission Standards, Regulatory Support Document; Office of
Mobile Sources, U.S. Environmental Protection Agency: Ann Arbor,
MI, 1997.
Singer, B.C.; Harley, R.A. A Fuel-Based Inventory of Motor Vehicle
Exhaust Emissions in the Los Angeles Area during Summer 1997;
Atmos. Environ. 2000, 34, 1783-1795.

1938 Journal of the Air & Waste Management Association

About the Authors

Andrew Kean is a graduate student in the Department of
Mechanical Engineering at the University of California at
Berkeley. He holds an M.S. degree in Mechanical Engineering from UC Berkeley and a B.E. degree in Mechanical Engineering from Cooper Union. Robert Sawyer and Robert
Harley are faculty members in the College of Engineering at
UC Berkeley. Please direct correspondence to R. Harley,
Department of Civil and Environmental Engineering, 631
Davis Hall, University of California, Berkeley, CA 947201710.

Volume 50 November 2000

Kean, Sawyer, and Harley

APPENDIX
Table 1A. Diesel fuel by end use and total distillate fuel oil sales (109 L).

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

Downloaded by [186.68.243.47] at 08:03 21 April 2016

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.

Volume 50 November 2000

Journal of the Air & Waste Management Association 1939