Full text of DOI:10.1080/10473289.2002.10470755

This article was downloaded by: [University of Illinois Chicago]
On: 11 November 2014, At: 11:27
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK
Journal of the Air & Waste Management Association
Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/uawm20
Factors Affecting Heavy-Duty Diesel Vehicle Emissions
Nigel N. Clark a , Justin M. Kern a , Christopher M. Atkinson a & Ralph D. Nine a
a Mechanical and Aerospace Engineering Department , West Virginia University ,
Morgantown , USA
Published online: 27 Dec 2011.
To cite this article: Nigel N. Clark , Justin M. Kern , Christopher M. Atkinson & Ralph D. Nine (2002) Factors
Affecting Heavy-Duty Diesel Vehicle Emissions, Journal of the Air & Waste Management Association, 52:1, 84-94, DOI:
10.1080/10473289.2002.10470755
To link to this article: http://dx.doi.org/10.1080/10473289.2002.10470755
PLEASE SCROLL DOWN FOR ARTICLE
Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained
in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the
Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and
should be independently verified with primary sources of information. Taylor and Francis shall not be liable for
any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever
or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of
the Content.
This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions

Clark et al.
TECHNICAL PAPER
ISSN 1047-3289 J. Air & Waste Manage. Assoc. 52:84-94
Copyright 2002 Air & Waste Management Association
Factors Affecting Heavy-Duty Diesel Vehicle Emissions
Nigel N. Clark, Justin M. Kern, Christopher M. Atkinson, and Ralph D. Nine
Mechanical and Aerospace Engineering Department, West Virginia University, Morgantown
ABSTRACT
and legislative pressures to monitor and abate diesel ve-
Societal and governmental pressures to reduce diesel ex-
haust emissions are reflected in the existing and projected
future heavy-duty certification standards of these emis-
sions. Various factors affect the amount of emissions pro-
duced by a heterogeneous charge diesel engine in any
given situation, but these are poorly quantified in the
existing literature. The parameters that most heavily af-
fect the emissions from compression ignition engine-pow-
ered vehicles include vehicle class and weight, driving
cycle, vehicle vocation, fuel type, engine exhaust
aftertreatment, vehicle age, and the terrain traveled. In
addition, engine control effects (such as injection timing
strategies) on measured emissions can be significant.
Knowing the effect of each aspect of engine and vehicle
operation on the emissions from diesel engines is useful
in determining methods for reducing these emissions and
in assessing the need for improvement in inventory mod-
els. The effects of each of these aspects have been quanti-
fied in this paper to provide an estimate of the impact
each one has on the emissions of diesel engines.
hicle emissions are rising. Also, the activity of heavy-duty
vehicles is projected to continue to increase over the next
decade. Sales of class 8 [more than 14,969 kg (33,000 lb)
gross weight] vehicles in 1999 in the United States ex-
ceeded 250,000 units. Presently, the heavy-duty diesel
emissions inventory is based on emissions factors devel-
oped from certification data gained using a stationary
engine dynamometer, and there is no sophisticated ac-
counting for the application of that engine in the vehicle
or the nature of vehicle behavior. This paper reviews chas-
sis dynamometer data, existing emissions data in the lit-
erature, and modeling to examine those factors that most
affect diesel vehicle emissions, which are usually expressed
in units of emissions mass per unit distance traveled (e.g.,
g/km). The main parameters that affect the emissions from
compression ignition engines include vehicle class and
weight, driving cycle, vehicle vocation, fuel type, engine
exhaust aftertreatment, vehicle age, and the terrain trav-
eled. In addition, the effects of engine controls, such as
injection timing strategies, on measured emissions must
be addressed.
INTRODUCTION
Prediction of heavy-duty diesel vehicle emissions inven-
tory is substantially less mature than the prediction of
gasoline passenger car emissions. However, diesel vehicles
are now receiving attention, because they are acknowl-
edged to be significant contributors to the atmospheric
inventory of particulate matter (PM) and NOx. Societal
VEHICLE CLASS AND WEIGHT
The effect of vehicle class on emissions is significant, yet
there is little detailed information available on this fac-
tor. Vehicle classes are defined by several entities, includ-
ing the American Automotive Manufacturers Association
(AAMA), and are usually based on the gross vehicle weight
rating (GVWR) as shown in Table 1. The GVWR is the
maximum weight a vehicle is allowed to achieve, includ-
ing the vehicle, driver, payload, and fuel. Unfortunately,
registration weight and GVWR often do not agree.
For heavy-duty vehicles, emission regulations are im-
posed on the engine regardless of the class of vehicle or
specific use of the vehicle in which the engine may be
installed. The federal testing procedure prescribed in the
Code of Federal Regulations, Title 40, Part 86, Subpart N1
is a transient test used to establish engine certification to
emissions standards that are, thus, based solely on the
engine performance. Conversely, the light-duty certifica-
tion test employs a chassis dynamometer and is affected
by road-load power and vehicle weight. All AAMA class 1
IMPLICATIONS
Heavy-duty vehicle contributions to the emissions inven-
tory are poorly documented and presently are based on
engine certification emissions data. This paper clarifies
the effect of various factors in truck and bus usage that
can have an effect on diesel emissions and cause the
reality to stray from the idealized model. In this way, mod-
elers can identify shortcomings in the inventory and be-
gin to address improved models. In particular, the inves-
tigation highlights the effects of vehicle use, load, age,
and maintenance on emissions, with emphasis on the pro-
duction of NOx and particulate matter (PM).
84 Journal of the Air & Waste Management Association
Volume 52 January 2002

Clark et al.
Table 1. Vehicle classes as defined by the AAMA.
weight differs by a factor of 5. This indicates that the fuel
consumption is not directly proportional to weight but
that it does increase as the vehicle weight increases. It is
well documented2 that, for a given engine meeting a given
emissions standard, emissions of NOx may be related
closely to CO2 emissions, so that the higher fuel consump-
tion implies higher emissions levels of NOx. Figure 1 shows
the ratio of emissions from the class 2 and class 8 model
trucks as speed varies on different grades.
Class
GVWR (kg)
GVWR (lb)
1
2
3
4
5
6
7
8
2721 and less
2722–4536
6000 and less
6001–10,000
4537–6350
10,001–14,000
14,001–16,000
16,001–19,500
19,501–26,000
26,001–33,000
33,001 and more
6351–7257
7258–8845
8846–11,793
11,794–14,969
14,970 and more
It is evident that, if two trucks employ the same en-
gine, all else being equal, the heavier vehicle will demand
higher energy (as axle-kilowatt-hour or akW-hr). Energy
at the rear wheels, given units of akW-hr, differs from
engine energy, given units of bkW-hr (brake-kilowatt-
hour), by the factor of drivetrain efficiency. Auxiliary and
fan engine loads also increase the brake-to-axle power
ratio. Such variation would be accounted for if emissions
variations were linear with power, as is NOx2, and if dif-
ferences in the demanded energy were appropriately mod-
eled. By this argument, even if the emissions in g/akW-hr
(or g/bkW-hr) were similar for the two vehicles, the emis-
sions in g/km would vary by a factor of the ratio of the
akW-hr/km used by each vehicle.
and some class 2 trucks may be emissions-certified using
the light-duty automotive approach.
The effect of vehicle class on emissions will first be
addressed from an analytical point of view. To compare
two vehicles of different truck classes, a theoretical model
of a vehicle operating at a constant speed was employed.
A simple road-load relation considering aerodynamic drag,
tire rolling resistance, and grade is shown in eq 1.
P = 1/ ρ C AV3 + µMgV + MgV sin θ
(1)
2
a
d
Heavy-duty vehicle engines are certified to emissions
levels in units of g/bhp-hr (equivalently g/bkW-hr), so if
two vehicles have an engine certified to the same stan-
dard, then their emissions in g/km will be influenced solely
by the ratio of bkW-hr/km. In this case, one would argue
that in regular service, a light–heavy-duty pickup truck (at
15 L/100 km economy) would emit at ~40% of the rate, in
g/km, of a tractor-trailer (at 78 L/100 km economy). This is
the argument embodied in the present inventory process,
but it is flawed if the emissions rates are nonlinear with
respect to power demand (as CO and PM are known to be)
or if “off-cycle” operation induces NOx emissions rates that
differ from certification rates, in units of g/bhp-hr.
where P is power required to maintain a steady speed, ρa
is density of air, Cd is aerodynamic drag coefficient of the
vehicle, A is frontal area of the vehicle, V is speed at which
the vehicle is traveling, µ is tire rolling resistance coeffi-
cient, M is mass of the vehicle, g is acceleration due to
gravity, and θ is angle of inclination of the road grade.
The three main factors that cause a vehicle to demand
engine power are vehicle speed, weight, and the incline
traveled. Note that this is for a steady-state (constant
speed) case only. As the required power increases, the
amount of fuel burned to produce that power also in-
creases, and the rate of regulated emissions produced will
generally increase. [Note, however, that brake-specific
emissions levels of some constituents, such as hydrocar-
bons (HC), may be high at low power ratings.] This im-
plies that emissions will directly vary with truck class. The
higher truck classes are heavier and, thus, produce more
regulated emissions. This implication will be further in-
vestigated in the following discussion.
A comparison of a pickup truck [3629 kg (8000 lb)
GVWR, class 2] and a tractor truck [18,143 kg (40,000 lb)
GVWR, class 8] powered by the same engine shows the
trend of higher emissions through fuel consumption. An
estimate of fuel economy was made from the energy the
vehicle required for a particular section of travel. Typical
fuel economies are 13.1 L/100 km (18 mi/gal) and 33.6
L/100 km (7 mi/gal) for the pickup truck and the tractor
truck, respectively. The fuel economy differs between these
two trucks by a factor of 2.5, while the corresponding
Figure 1. Ratio of class 2/class 8 NOx emissions for varying grades
and speeds.
Volume 52 January 2002
Journal of the Air & Waste Management Association 85

Clark et al.
For a comparison of truck classes from test data, the
the vehicle), the CO and PM (in units of g/km) were both
considerably higher than for the lighter test weights. How-
ever, NOx (in units of g/km) was relatively insensitive over
the range of test weights.
emissions of two different heavy vehicles with the same
engine were compared, noting that these two trucks have
a different vocation, transmission, and horsepower rat-
ing. These vehicles had different engine power ratings
of 224 kW (300 hp) for the tractor truck and 207 kW
(277 hp) for the transit bus. Each vehicle was tested on a
different test cycle; however, the two different cycles are
the most similar test vehicles available from the West
Virginia University (WVU) Transportable Laboratory3
database when comparing transit bus data to truck data.
The bus was tested on the central business district (CBD)
Cycle,4 and the tractor truck was tested on the Truck-
CBD Cycle (also called the Modified CBD Cycle).5 The
Truck-CBD Cycle has slower acceleration ramps so that
a vehicle with a lower power-to-weight ratio and with
an unsynchronized manual transmission (tractor truck)
can follow the scheduled speed.
Research by Graboski et al.7 for the Northern Front
Range Air Quality Study (NFRAQS) reported emissions
testing on 21 different heavy-duty vehicles using the
WVU truck (i.e., 5-Peak) Cycle,5 U.S. Environmental Pro-
tection Agency (EPA) Urban Dynamometer Driving
Schedule for Heavy-Duty Vehicles (UDDS or Test-D),8 and
the CBD Cycle as described in SAE J1376.4 Results of
NFRAQS included comparisons to the GVWR of the ve-
hicles against the emission results. The conclusion by
Graboski et al.7 was that a heavier vehicle uses more fuel
and, thus, produces more exhaust gas on a g/km basis. It
was also noted that, as a vehicle following a cycle used
more fuel, higher emissions were produced in units of
g/km. For example, the CBD Cycle yielded the highest
fuel consumption and also the highest emissions as com-
pared with the other cycles.
The tractor truck exhibited lower emissions of NOx,
HC, and PM of 26, 8.2, and 30%, respectively. The total
emissions of CO were higher for the tractor truck by 12%.
It is evident from these data that conclusions based on
vehicle class alone are not reliable and that vocation (as
mimicked by the test cycle) and transmission type must
be considered. The results are opposite of the expected
lower emissions from the less powerful bus engine.
Testing performed using a bus from the Flint Mass
Transit Authority has also been used for various compari-
sons.6 Testing was performed on this bus using several
different driving test cycles run consecutively. The bus
was outfitted with a Detroit Diesel Series 50 engine coupled
to a five-speed automatic transmission. The engine was a
four-cylinder unit, having 8.5 L of displacement rated at
205 kW (275 bhp) operating on No. 2 diesel. Although
these are data from just one bus, many other similarly
equipped buses were tested at this Flint site and showed
consistent bus-to-bus correlation of emissions data for
operation on the CBD Cycle.
TEST CYCLE EMISSIONS COMPARISON
The data from the trucks tested in the NFRAQS were evalu-
ated by Coburn,9 who concluded that more vehicles need
to be tested to obtain reliable and precise estimates of
average PM emissions. Also, the level of PM emissions
measured depends on the driving cycle used to test the
vehicle. The conclusions from Graboski et al.7 indicate
the trend of the CBD Cycle producing the highest emis-
sions and the WVU 5-Peak Cycle producing the lowest
emissions with the heavy-duty UDDS (Test-D) between
them. This trend was attributed to the CBD Cycle being
the most aggressive cycle with more acceleration ramps
and more sustained high acceleration than the other
cycles. One particular heavy-duty diesel-powered vehicle
that was tested on all three cycles was a telephone truck
of 8845 kg (19,500 lb) curb weight and 36,287 kg (80,000
lb) GVWR. It was a 1983 model year vehicle powered by a
Cummins NTC400 with the odometer showing 80,876
mi (equivalently, 130,157 km). For the NFRAQS telephone
truck data, NOx ranged from 11.7 to 19.4 g/km for the
three cycles used, which is a wider relative variation than
for the WVU bus data discussed below.
The data collected from the Flint Mass Transit Au-
thority bus contained one portion in which the test weight
of the bus was varied while tested on the CBD Cycle. Table
2 shows the emissions results from this testing. For the
test weight set at 17,237 kg (38,000 lb) (max GVWR for
A sequence of tests performed by the WVU Trans-
portable Laboratory will be considered to evaluate the ef-
fect of driving test cycles on the emissions produced. From
the WVU data on the Flint bus, it is evident that the units
in which the emissions are expressed are significant. The
comprehensive data are shown in Table 3. For example,
considering NOx, the NY Bus Cycle is highest of all the
cycles in g/km but lowest in average g/sec. Also, the CBD
Cycle and WVU 5-Peak Cycle yield similar emissions in
g/km but emissions differing by a factor of 2 in g/akW-hr.
Table 2. Emission results from varied test weights for the Flint bus driven on the
CBD cycle.
Test Weight (kg)
NOx (g/km)
CO (g/km)
17,237
19.1
14,889
20.0
12,542
17.8
4.29
2.80
2.83
HC (g/km)
0.09
0.09
0.09
PM (g/km)
0.21
0.14
0.13
86 Journal of the Air & Waste Management Association
Volume 52 January 2002

Clark et al.
Table 3. Flint bus emissions for various test cycles at a test weight of 14,889 kg.
certification, (NOx + HC) are regulated as a
sum. Both the NFRAQS and the WVU data
showed HC cycle-to-cycle variations of a fac-
tor of 2, when the units were in g/km, exclud-
ing the NY Bus Cycle.
Of specific interest is the comparison of
the WVU data for the bus using the WVU
5-Peak Cycle and the WVU 5-Mile Route. Both
cycles are similar except that the WVU 5-Peak
Cycle does not demand full power from the
bus engine upon acceleration. PM values in
g/km for the full power operation (route) are
slightly more than twice as high as for the
cycle. This confirms the sensitivity of CO and
PM emissions to engine loading, in contrast
to the relative stability of the NOx emissions
in units of g/km.
A portion of testing at WVU involved test-
ing a single truck on various different test
cycles.10 The results for NOx in g/km are shown
in Figure 2. The vehicle was a 1995 GMC box
truck with a Caterpillar 3116 engine rated at 127
kW (170 hp). The fuel used was D2 diesel, and
the vehicle has a GVWR of 9980 kg (22,000 lb).
It is concluded that the test cycle used has a
profound effect on PM emissions and a signifi-
cant effect on NOx emissions.
Cycle
g/km
g/cycle
Avg. g/sec
g/akW-hr
g/gCO2
g/L Fuel
NOx
CBD
20.0
17.8
15.4
43.5
16.7
62.0
143.0
123.8
44.1
0.1080
0.1682
0.1375
0.0735
0.1398
CO
15.5
26.0
22.0
28.4
18.2
0.01145
0.02238
0.01868
0.01301
0.01536
30.7
60.1
50.1
34.6
41.1
WVU 5-peak
WVU 5-mile
NY-Bus
Test-D
148.5
CBD
2.80
0.81
1.55
27.5
3.85
13.94
6.50
0.0243
0.0076
0.0139
0.0465
0.0323
HC
3.50
1.18
2.23
18.0
4.21
0.0026
0.0010
0.0019
0.0082
0.0036
6.89
2.73
5.07
21.9
9.52
WVU 5-peak
WVU 5-mile
NY-Bus
12.53
27.91
34.35
Test-D
CBD
0.087
0.28
0.35
0.30
0.38
0.33
4.93E–04
4.12E–04
3.34E–04
6.41E–04
3.13E–04
PM
0.071
0.063
0.054
0.248
0.040
5.22E–05
5.48E–05
4.54E–05
1.13E–04
3.44E–05
0.140
0.148
0.122
0.301
0.092
WVU 5-peak 0.044
WVU 5-mile 0.037
NY-Bus
Test-D
0.379
0.037
CBD
0.137
0.69
0.40
0.85
0.84
2.05
1.20E–03
4.71E–04
9.46E–04
1.40E–03
1.93E–03
0.172
0.072
0.152
0.540
0.252
1.27E–04
6.26E–05
1.29E–04
2.47E–04
2.12E–04
0.341
0.169
0.346
0.658
0.568
WVU 5-peak 0.050
WVU 5-mile 0.106
NY-Bus
Test-D
0.827
0.230
This is to be expected because the vital ratios of the cycles,
such as akW-hr/km, vary widely. In currency of g/km,
the WVU bus data show that virtually all cycles yield
NOx in the range of 15.4–20.0 g/km, with the NY Bus
Cycle an outlier at 43.5 g/km. This is because the NY Bus
Cycle covers a short distance over its duration relative
to other cycles. One may conclude that NOx data, in
g/km, remain fairly consistent for most cycles in current
use, provided that the cycle does not contain excessive
idle or low power operation.
VEHICLE VOCATIONS AND LOCAL
DRIVING ACTIVITY
The particular vocation or specific use of a vehicle can
have an effect on the emissions produced. The transients
and cruising behavior of each vehicle vocation, along with
the load carried, can be reproduced in testing by chang-
ing the testing weight and the driving cycle. A driving
For diesel vehicles, variations in both CO and PM
are acknowledged to be higher than those for NOx, all
else being equal. This is borne out by the data of both
the NFRAQS and WVU studies. For the Flint bus, ex-
cluding the NY Bus Cycle as an outlier, emissions of
CO in g/km varied by a factor of 5 over the four cycles
used, and the NFRAQS data yielded a similar ratio. For
PM in both studies, the range was a factor of 3–4, in
g/km. One must conclude that the cycle chosen has a
profound effect on PM and CO levels, if they are ex-
pressed in g/km. The WVU data showed that choice of
units in g/akW-hr did not improve cycle-to-cycle agree-
ment. Hydrocarbon emissions from diesel engines are
customarily low and are of less interest in inventories
than NOx and PM emissions, although for 2002/2004
Figure 2. NOx emissions from various test schedules on one truck.10
The truck was a 1995 GMC tested at 9980 kg (22,000 lb) vehicle weight.
The vehicle was equipped with a Caterpillar 3116 engine rated at 170
hp and an automatic transmission. The test fuel was No. 2 diesel.
Volume 52 January 2002
Journal of the Air & Waste Management Association 87

Clark et al.
Table 4. Combined survey data from Roadway and Overnite tractors.
cycle for a bus should produce the characteris-
tics of a bus route (frequent stops, high accel-
eration, and low average speeds) and produce
representative values of the exhaust emissions.
This would also be true for a particular truck
vocation; either a local delivery, long haul, or
shipping yard route would be used where appli-
cable. For example, the axle energy per distance
used by a vehicle following different test cycles
can indicate the difference in vocation. The same
bus following the WVU 5-Mile Route and the
NY Bus Cycle used 0.70 akW-hr/km (axle kilo-
Microtrip Type
Distance
(km)
Average Speed
(km/hr)
Average
Average
Acceleration
(km/hr/sec)
Deceleration
(km/hr/sec)
Interstate
Suburban
City
319
246
42
52.5
28.6
16.4
10.1
25.7
0.98
1.32
1.21
1.17
1.29
–1.43
–1.93
–1.80
–1.40
–1.91
Yard
8.8
288
Suburban and city
watt-hour per kilometer) and 1.53 akW-hr/km for each
cycle, respectively. The difference in the axle energy per
distance has a direct effect on the emissions produced in
mass per distance. If this vehicle were equipped with an
engine that consistently produced 6.7 g/bkW-hr of NOx
(8.38 g/akW-hr assuming 80% overall drivetrain effi-
ciency), then the WVU 5-Mile Route would produce 5.8
g/km of NOx and the NY Bus Cycle would produce 12.8
g/km. Likewise, the PM emissions would be 0.12 g/km for
the WVU 5-Mile Route and 0.25 g/km for the NY Bus
Cycle, although in reality PM cannot be taken as energy-
specific with reliability.
vocation is difficult to tackle but is also covered, in part,
by the discussion of test cycles above. It is evident that a
line haul tractor may be expected to emit at lower levels
in g/km than would an inner-city refuse truck, because
long idle periods and stop-and-go operation will increase
emissions in g/km.
Local driving activity also affects heavy vehicle emis-
sions but is difficult to quantify. For example, driving in a
road system that requires frequent stops is likely to raise
emissions. This is very close to the definition of vehicle
vocations and also has an impact on the discussion on
test cycles. Local driving habits will also affect the vehicle
emissions due to driver-to-driver variations. The effect that
these factors have on vehicle emissions is comparable with
the effect of different driving cycles that mimic the par-
ticular driving patterns or vehicle uses.
Comparing other vocations of a line haul tractor at
steady cruise and a refuse truck operating on the New York
Garbage Truck (NYGT) Cycle, similar results are obtained.
The energy requirement for a 36,287 kg (80,000 lb) trac-
tor trailer traversing flat terrain at 97 km/hr was calcu-
lated to be 1.5 akW-hr/km. Conversely, a refuse truck
following the NYGT Cycle would use 1.1 akW-hr/km.
The analysis of vehicle weight using the road-load
equation disregards the fact that the vehicle has to accel-
erate to the assumed steady-state condition. Under accel-
eration, it is assumed that a heavy vehicle is customarily
using the maximum power available from its engine, thus
producing the maximum amount of exhaust gas and typi-
cally high rates of NOx and PM. So then, over a typical
day of use for any vehicle, one that stops and then accel-
erates more often will produce higher distance-specific
emissions, providing all else is held constant. This effect
is from the differences in the use of the vehicle, also called
the vehicle’s vocation.
Research at WVU, funded by National Renewable
Energy Laboratory, and leading to the development of the
City-Suburban Heavy Vehicle Route,11 included recording
data from two delivery companys’ tractor trucks as the
drivers performed their respective tasks. Table 4 shows the
data collected from this survey. It is evident that the pri-
mary difference between operation in yard, city, subur-
ban, and interstate service lies in the average speed,
whereas typical accelerations are similar, most likely requir-
ing full engine power. The issue of the effect of vehicle
FUEL DIFFERENCES
Fuels other than conventional diesel can provide a means
of reducing heavy-duty engine emissions. Fuels such as
compressed natural gas, liquefied petroleum gas, and vari-
ous alcohols have been used but require engine modifica-
tions for operation. However, using a reformulated diesel
or a diesel equivalent fuel that does not require engine
modifications can produce significant reductions in engine
emissions while avoiding the expense of vehicle modifica-
tions. Complete fuel reformulation would affect all heavy-
duty diesel vehicles and has the potential to reduce NOx
and PM significantly.12 Diesel fuel additives have also been
used for reduction of emissions as shown by Lange et al.13
and Green et al.14 The most common additive has been a
cetane number enhancer. The results show that a fuel with
a higher cetane number ignites earlier and, thus, may use
less fuel for the same power output. In some cases, earlier
ignition may increase NOx emissions, while in others, re-
duction of the premix fuel burn can reduce NOx. Brown15
states that the most beneficial and cost-effective solution
to reduce exhaust emissions is high-quality, fully reformu-
lated diesel combined with exhaust aftertreatment.
In the study by Graboski et al.,16 the transient emis-
sions from D2 diesel and biodiesel blends in a DDC Series
88 Journal of the Air & Waste Management Association
Volume 52 January 2002

Clark et al.
60 engine were investigated. The fuels tested were a refer-
ence diesel and 20, 35, and 65% biodiesel (methyl soy
ester) blends in the reference diesel, as well as 100%
biodiesel. This testing was performed on an engine dyna-
mometer with NOx, CO, HC, and PM recorded in g/bkW-
hr. The biodiesel showed a strong trend of reduction in
CO, HC, and PM, but an increase in NOx emissions.
Recent emissions characterization by WVU included
a fuel from Malaysia produced using a Fischer-Tropsch
(F-T) process. This fuel is produced from natural gas and
has similar physical properties to diesel fuel but usually
with zero sulfur and very low aromatic levels. Transient
engine tests were performed at WVU using the F-T fuel
along with federal No. 2 and California No. 2 diesel. Table
5 shows the averaged results of three tests using these fu-
els on a Navistar 444 engine using the Federal Test Proce-
dure engine cycle. Both the California diesel and the F-T
fuel showed a substantial decrease of each exhaust gas
component over U.S. federal diesel fuel. The F-T fuel pro-
duced the greatest gain of 65% reduction in HC. From
this comparison, one can conclude that fuel type has the
potential to provide a substantial reduction in emissions.
Clark et al.17 recently compared a variety of diesel fuels
and diesel fuel substitutes. Testing for this study included
available diesel fuels, biodiesel blends (e.g., BD20 is 20%
biodiesel), fuels from the Fischer-Tropsch process (termed
F-T and MG), and a blend of the MG fuel with isobutanol.
The results of the testing were then compared with simi-
lar studies by Graboski et al.16 and Schaberg et al.18 The
PM results are shown in Figure 3. The fuels tested by
Schaberg et al.18 were diesel fuel from the Sasol slurry phase
distillate process and blends of the two and are termed
2D, B1, B2, B3, N, K, and C. The comparison showed a
maximum decrease of 66% in PM emissions for the fuels
tested relative to No. 2 diesel.
Figure 3. Normalized PM emissions results comparison for different
fuel types reproduced from Clark et al.17 Data are from studies by
Schaberg et al.,18 Graboski et al.,16 and Clark et al.17 For each study,
the low-sulfur diesel results were set at 100%. The Clark et al.17 BD20
results appear anomalous, but no corrective explanation can be found.
promising results. An oxidation catalyst is similar in prin-
ciple to the oxidizing section of a gasoline engine cata-
lytic converter. It works by oxidizing the gaseous HC and
CO in the exhaust to produce water vapor and CO2.19
While this type of aftertreatment is effective in reducing
HC and CO, these are not the specific diesel exhaust pol-
lutants that are desirable to eliminate. The quest for a lean
burn NOx reduction catalyst for diesel engines remains
an unattained grail, although systems employing HC re-
ductants and urea are now under investigation.
A particulate trap or filter is a device in which the PM is
collected or filtered out of the exhaust and is regenerated by
some external means. This is usually accomplished by heat-
ing the trap to burn the PM, using a fuel additive that causes
regeneration, or employing a filter surface incorporating a
catalyst. These methods are sometimes not effective in re-
generating at low loads and low exhaust temperatures.20
The most recent type of exhaust aftertreatment devel-
oped is the CRT. This type of system combines an oxida-
tion catalyst and a particulate trap filter that reduces both
gaseous and particulate emissions. The exhaust gases first
pass through the catalyst to oxidize CO and HC and also
convert the majority of NOx to NO2, which is then used to
oxidize the PM in the particulate trap.19 This aftertreatment
method continuously regenerates with no fuel ad-
EXHAUST AFTERTREATMENT
Although only a small number of heavy-duty vehicles are
equipped with any aftertreatment device, the effect on
emissions can be substantial. The three primary types of
diesel exhaust aftertreatment are diesel oxidation cata-
lyst, particulate traps, and continuously regenerating traps
(CRT). These have been tested in various studies and show
Table 5. Engine emissions comparison of various fuels.
ditives or heater control system. A drawback of this
system is that it requires low-sulfur diesel, because
Fuel
D2
CA D2
Difference (%)a
F-T
Difference (%)a
combustion of sulfur degrades the catalytic reactions
in the system. Continuously regenerating particu-
late traps have the potential to reduce PM by a fac-
tor of 3.5, and reduction in NOx is ~10%.19,21
NOx (g/bkW-hr)
CO (g/bkW-hr)
HC (g/bkW-hr)
PM (g/bkW-hr)
7.04
1.89
0.31
0.16
6.38
1.23
0.20
0.15
–9.3
–35
–35
–8.3
5.99
1.02
0.11
0.13
–15
–46
–65
–17
The WVU Transportable Laboratory has per-
formed testing on vehicles with and without cata-
lytic converters. A refuse truck was tested in 1995
aUsing D2 diesel as baseline.
Volume 52 January 2002
Journal of the Air & Waste Management Association 89

Clark et al.
with a catalytic converter manufactured by Donaldson. The
vehicle was a 1992 model year with 90,952 km (56,515 mi)
on a Cummins LTA-10 engine supplying power through a
four-speed automatic transmission. This engine was rated
at 194 kW (260 hp) operating on D2 diesel. The truck was
driven on the WVU 5-Peak Cycle, and Table 6 shows the
results of this comparison. Each of the exhaust gas con-
stituents was lowered by use of the catalytic converter. The
total PM was lowered the most at 24%, NOx was lowered
~9%, and the CO was lowered the least at 8.3%. One may
conclude that catalytic converters are successful in reduc-
ing HC and PM. Substantial reductions of only CO, HC,
and perhaps fine PM can be expected on this basis.
model year in which the vehicle was made. From the EPA
emissions standards that are used to certify engines, a 1998
model year, heavy-duty engine (used in a bus) would pro-
duce less NOx by a factor of 2.5 and less PM by a factor of
12 relative to a 1988 model year engine. This is true only
if the vehicle produces emissions that correlate directly
to the emissions certification standards.
The effect of the technology level of the engine on
vehicle emissions was determined by comparing test data
from two different vehicles with the same model engine.
These vehicles were tested by the WVU Transportable
Laboratory in 1994. The engine was a Detroit Diesel Cor-
poration 6V-92TA burning D2 diesel. This engine has six
cylinders and a displacement of 9.05 L and produces 207
kW (277 hp). Table 7 shows the summary of vehicle in-
formation and the measured emissions from each vehicle.
The newer engine was made five years after the older
engine and produced emissions that, for NOx and CO, were
lower than those of the older engine. The largest reduction
offered by the newer engine was in CO and was 71%. The
total HC production was 11% higher on the newer engine,
but the PM was reduced by 45%. The HC and CO changes
most likely imply a substantial reduction in elemental car-
bon in the PM. This simple comparison shows that for the
majority of the exhaust gas constituents, reduction has
occurred in the five-year time period. Interestingly though,
the fuel economy of the newer vehicle experienced a de-
crease of 10%, most likely due in part to the retarding of
the timing to meet NOx emissions requirements.
VEHICLE AGE
There are two separate factors of vehicle age that affect
the emissions produced. First, it is assumed that, as a ve-
hicle ages and accumulates high mileage, the engine will
slowly wear and produce higher emissions even though
diesel engine deterioration is recognized to be slow for
purposes of certification. This would imply that, for ex-
ample, a 20-year-old vehicle would produce higher emis-
sions after 20 years of use than it did when it was new.
The second factor is the change in technology. The chang-
ing technology implies that the engines produced today
are different than older ones and must meet more strin-
gent emissions standards. There are few data available for
an age comparison of the same truck that was tested new
and after some certain useful life.
There is documentation by EPA on vehicle deteriora-
tion factors for heavy-duty vehicles. These deterioration
factors are supplied by manufacturers and may be either
additive or multiplicative in modifying the baseline emis-
sions from a new engine. Generally, diesel engines are
reported to deteriorate little over the first 466,710 km
(290,000 mi) of use (the useable limit previously set by
EPA). However, engines now last 800,000–1,600,000 km
until the first rebuild. No data can be found to clarify
deterioration in the final stages before the rebuild.
The age of the vehicle has a significant effect on emis-
sions when pertaining to the technology in the particular
Testing at WVU has included testing the same ve-
hicle annually. Testing from the Bi-State Development
Agency in St. Louis, MO, included a transit bus powered
by a DDC 6V-92TA engine operating on D2 diesel. Table 8
shows the emissions over a span of 2 years as the bus
Table 7. Vehicle specifications and emissions for vehicle age comparison.
Vehicle
1
2
Type
Transit bus
1988
Transit bus
1993
Model Year
Transmission
Test Weight
Test Cycle
Test Date
Odometer
3-speed auto.
15,297 kg
CBD
4-speed auto.
15,048 kg
CBD
Table 6. Emissions results for comparison of exhaust aftertreatment.
June 4, 1994 March 16, 1994
Constituent Without Catalytic With Catalytic
Converter (g/km) Converter (g/km)
Difference (%)a
287,747 km
171,794 km
Differencea
–35 %
NOx (g/km)
CO (g/km)
HC (g/km)
PM (g/km)
23.7
13.8
1.99
1.92
15.5
4.00
2.20
1.06
84.0
NOx
CO
14.4
1.88
1.08
0.39
3.31
13.1
1.72
0.88
0.29
3.31
–9.1
–8.3
–18
–24
–0.2
–71 %
+11 %
HC
–45 %
PM
Fuel Consumption (L/100 km) 75.6
+11 %
Fuel economy
aUsing 1988 bus as baseline.
aWithout converter used for baseline.
90 Journal of the Air & Waste Management Association
Volume 52 January 2002

Clark et al.
Table 8. Emissions from one vehicle as mileage accumulated.
terrain grade has been used in the evaluation
of a hybrid fuel cell bus,24 this is not typically
applied to diesel vehicles. For a comparison
of emissions produced from a vehicle travel-
ing varying terrain, the theoretical power re-
quirement can be determined. The power can
then be related to the emissions rate for a par-
ticular vehicle from experimental brake-spe-
cific emissions data. A simple, theoretical
road-load relation considering aerodynamic
Odometer
Reading
(km)
NOx
(g/km)
CO (g/km)
HC (g/km)
PM (g/km)
Fuel
Test Date
Consumption
(L/100 km)
219,741
288,946
370,785
20.4
30.5
28.2
8.70
4.60
4.66
1.06
1.43
1.31
0.329
0.447
0.392
60.0
62.1
63.6
6/7/94
3/20/95
4/17/96
accumulated mileage (from 1994 to 1996). From these
data, the only trend observed was that the fuel economy
decreased as the mileage increased. The emissions show
no definite trend of increasing or decreasing.
The NFRAQS testing compared the collected emissions
data by emissions model year and mileage since last en-
gine rebuild. A trend of reduced PM emissions was re-
corded by Graboski et al.7 as model year increased;
however, no trend of NOx reduction was apparent from
the chassis testing. The conclusions show that there was
no change in emissions as the vehicle mileage since last
rebuild varied, which supports the low deterioration fac-
tors currently in use.
The emissions certifications that an engine must meet
are specified by year of manufacture. Although diesel en-
gines produce far less CO and HC than standards allow,
PM and NOx are usually close to the limit. Small devia-
tions from these levels may have occurred due to emis-
sions credit banking, but it is evident that levels of NOx
and PM have been forced to decline through use of im-
proved technologies over the years. For example, higher-
pressure injection has emerged as a tool to reduce PM. A
recent development of technology toward reducing die-
sel engine emissions is the use of exhaust gas recircula-
tion (EGR). EGR provides an effective means of NOx
emissions reduction. There are many other methods of
reducing emissions that would be considered technology
advances, such as combustion chamber design and in-
troduction of variable geometry turbochargers. More
stringent emissions standards (such as the year 2007
EPA standards) suggest that emissions of NOx and PM
will be reduced in newer model engines, but existing
data bear out only the substantial reduction in PM in
actual vehicle use.7,22,23
drag, tire rolling resistance, and grade is shown in eq 1 in
the previous discussion of vehicle class and weight.
Ramamurthy et al.25 plotted the relation between axle
power and NOx emissions rates for some typical diesel
vehicles. This type of data may be used to project the NOx
emissions from vehicles under different use. Using the
required power calculated from the road-load equation,
the NOx emissions rate can be predicted from the regres-
sion equation of the data and is shown in Figure 4. This is
only the lower on-cycle portion of all the data from one
test and is fairly consistent. However, all of the operating
points of the test sequence do not produce emissions that
fall along this line. When all of the points are considered,
there is a bifurcation present in the data associated with
off-cycle operation. The full data set can be seen in Figure
5 and is discussed in the next section dealing with injec-
tion timing variances, the cause of this bifurcation.
Effect of terrain has been estimated using a simple
model. For this analysis, the incline grade was limited to
7% (~4º above horizontal). The required power for a model
vehicle was determined from the road-load equation for
constant grades. The model vehicle was a class 8 tractor-
trailer, and steady-state cruising was used for the model-
ing. To climb a 7% grade, the model vehicle would use
205 kW to maintain a steady-state speed of 48 km/hr. Al-
most the same power (207 kW) is required for this vehicle
to maintain a speed of 77 km/hr when climbing a 3.5%
TERRAIN TRAVELED
The effects of the terrain traveled by a vehicle are referred
to as grade effects. The chassis testing performed on the
WVU Transportable Chassis Laboratory uses power absorb-
ers and inertial flywheels to provide a load to the vehicle
based on a road-load equation.3 For this equation, it is
assumed that there are no hills, and the load is calculated
for perfectly flat, level terrain. Although testing that includes
Figure 4. Smoothed axle power vs. shifted NOx as used for terrain
modeling.
Volume 52 January 2002
Journal of the Air & Waste Management Association 91

Clark et al.
Many present-day electronically controlled engines
do not embody timing throughout their operating range
that reflects the timing employed during the engine cer-
tification test. Although this practice has been curtailed
for the years 1999 and onward, a large portion of the fleet
now has engines with off-cycle timing strategies. Devia-
tions in timing during off-cycle operation may lead to
emissions of NOx that are higher than those that would
occur during the certification test at the same engine speed
and load. In some cases, available data support a binary
timing map, with a high and a low NOx emissions rate.
Because history effects may determine which of the two
timing choices is in effect, it is not always possible to pre-
dict unambiguously the NOx emissions rate given the
engine torque and speed.
Figure 5. Smoothed axle power versus shifted NOx showing
bifurcation of data. The reader is encouraged to compare the slope of
the 8.38 g/akW-hr (6.25 g/ahp-hr) line with the best-fit line of the low
NOx mode in Figure 4.
Figure 5 presents a plot of chassis-based NOx emis-
sions versus power output at the rear axle for a late-model
diesel truck. The lower NOx data set, when plotted versus
axle power, corresponds well to the line of 8.38 g NOx/
akW-hr. A certification rate of 6.71 g/bkW-hr, coupled with
an assumed overall drivetrain efficiency of 80%, yields an
8.38 g/akW-hr value. The higher NOx data set represents
the off-cycle operating points.
All present-day truck emissions values used for in-
ventory prediction rely on the certification data, but Fig-
ure 5 shows that certification data will underestimate NOx
emissions in off-cycle operation. For example, in Figure
5, the whole cycle required 7.41 akW-hr of energy from
the truck and yielded 110.9 g NOx. This corresponds to
an actual emissions rate of 15.0 g/akW-hr for this cycle,
in comparison to the 8.38 g/akW-hr value (approximately)
that might be expected. The real NOx value in this case
was 1.8 times higher than the expected value.
grade. The value of the required power is the same for
each case and, when applied to the linear regression of
the NOx emissions, the emissions rate is the same. This
shows that for ascending a grade, because NOx emissions
are often linear with power, a simple prediction can be
made. An oscillating terrain simulation (ascending and
descending grades) would be informative if such factors
as vehicle braking and driver shifting patterns were known.
Also, knowledge about the emissions produced from the
vehicle when the engine is operating in a power absorb-
ing or motoring mode is unavailable. Similar modeling
for CO, PM, and HC would be useful, but nonlinearities
make the results less certain.
A simpler method is to assume that a vehicle emits
levels that correspond to the emissions standards. Us-
ing the axle power values from the model above, PM
levels would vary from 11 to 35 g/hr considering a con-
stant travel speed on level ground and on a 3.5% grade.
This assumes an overall transmission efficiency of 80%
and shows a difference by a factor of 3 for PM. This
same procedure can be used for the other regulated emis-
sions, and the accuracy of each depends on the ability
of the vehicle to produce emissions that correspond to
the certification levels.
Between the range of 60 and 97 kW, there are two
noticeably different sets of data points, namely high NOx
and low NOx modes. A least-squares line was fit to each
set of data in this range and evaluated at the mid-point
(78 kW). The results show that in high NOx mode, 0.44
g/sec of emissions were produced, and in low NOx mode,
0.18 g/sec were produced. These two modes differ by a
factor of 2.4 at the 78-kW point of operation, where the
low NOx mode corresponds well with certification data.
Bifurcations in timing cause accurate emission predictions
to be unreasonably difficult. The authors have observed
that the choice of injection timing may be triggered by
modest variations in driving style over the same chassis
cycle, so that significantly different NOx emissions may
arise for fairly similar truck behavior.
INJECTION TIMING VARIANCES
Emissions of NOx and PM are known to be affected
strongly by the timing of the in-cylinder fuel injection in
diesel engines. Indeed, it is common to present a hyper-
bolic NOx-PM tradeoff curve for an engine, with more
advanced timing at the same speed and load leading to
higher NOx and lower PM. Within a reasonable operating
range, there is also a tradeoff between NOx and efficiency,
with advanced timing leading to a higher NOx and higher
thermal efficiency.
Timing variations influence the overall emissions in-
ventory in two ways. First, the timing variations cause
the actual NOx inventory to be higher than predicted based
on certification data, and second, the timing variations
cause the actual PM inventory to be lower than predicted
92 Journal of the Air & Waste Management Association
Volume 52 January 2002

Clark et al.
Figure 6. Effect of various factors on PM emissions. Each bar represents a specific comparison discussed in the text and cannot be taken to
represent all cases encountered.
Figure 7. Effect of various factors on NOx emissions. Each bar represents a specific comparison discussed in the text and cannot be taken to
represent all cases encountered.
Volume 52 January 2002
Journal of the Air & Waste Management Association 93

Clark et al.
8. EPA Urban Dynamometer Driving Schedule for Heavy-Duty Engines.
Code of Federal Regulations, Part 86, Subpart M, Title 40, 1996.
9. Coburn, T. Statistical Analysis of Particulate Matter Emissions from Light-
Duty and Heavy-Duty Diesel Vehicles; NREL/TP-540-224942; National
Renewable Energy Laboratory: Golden, CO, 1998.
10. Nine, R.D.; Clark, N.N.; Norton, P. Effect of Emissions on Multiple Driv-
ing Test Schedules Performed on Two Heavy Duty-Vehicles; SAE 2000-01-
2818; Society of Automotive Engineers: Warrendale, PA, 2000.
11. Clark, N.N.; Daley, J.J.; Nine, R.D.; Atkinson, C.M. Application of the
New City-Suburban Heavy Vehicle Route (CSHVR) to Truck Emissions
Characterization; SAE 1999-01-1467; Society of Automotive Engineers:
Warrendale, PA, 1999.
12. Diesel Exhaust: A Critical Analysis of Emissions, Exposure, and Health
Effects; Health Effects Institute: Cambridge, MA, 1995.
13. Lange, W.W.; Cooke, J.A.; Gadd, P.; Zürner, H.J.; Schlögl, H.; Richter,
K. Influence of Fuel Properties on Exhaust Emissions form Advanced Heavy
Duty Engines Considering the Effect of Natural and Additive Enhanced
Cetane Number; SAE 972894; Society of Automotive Engineers:
Warrendale, PA, 1997.
based on certification data. Timing variations in electroni-
cally controlled diesel engines present the single greatest
obstacle to present-day mobile source emissions inventory
prediction. This factor also can corrupt conclusions while
comparing other factors. The advanced technology applied
to heavy-duty vehicles would be expected to lower emis-
sions to comply with recent regulations. However, vehicles
with advanced electronic controls may emit in practice at
higher levels than certification data would suggest.
CONCLUSIONS
For each of the factors in the previous sections, a relative
comparison was made to estimate the effect of that par-
ticular factor on the emissions produced. The analysis was
completed using comparisons of measured data and ana-
lytical modeling. Figures 6 and 7 graphically represent
the results of these comparisons, and each bar depicts the
amount a factor could change the emissions of PM and
NOx. These two emissions species are of the most interest
in compression ignition engines because the production
of HC and CO from diesel engines are typically well be-
low the standards. The largest effect on emissions was
exemplified by the driving cycle that is used to test the
vehicle. The test data comparisons showed that the PM
emissions could vary by a factor of 15 and NOx emissions
could vary by a factor of 3 when measured using different
chassis dynamometer test schedules. This reinforces the
fact that the test schedule must be correctly matched to
the vehicle and accurately mimic real-world use. The in-
jection timing variances, which lead to off-cycle opera-
tion, also affect the measured emissions. The data
comparisons show that injection timing variances can in-
crease NOx emissions by a factor of 2 depending on oper-
ating conditions. The extent to which off-cycle emissions
affect the measured emissions is difficult to predict, be-
cause the frequency and duration of off-cycle operation
are obscure.
14. Green, G.J.; Henly, T.J.; Starr, M.E.; Assanis, D.N.; Syrimis, M.;
Kanafani, F. Fuel Economy and Power Benefits of Cetane-Improved Fuels
in Heavy-Duty Diesel Engines; SAE 972900; Society of Automotive En-
gineers: Warrendale, PA, 1997.
15. Brown, S.J. An Experimental Investigation of Affordable Alternative
Fuel and Exhaust Aftertreatment Options for Urban Buses. In Trucks
and Buses, the Results of Real-World Emissions Testing, Millbrook Proving
Ground Seminar Proceedings; Millbrook Proving Ground Ltd.: Millbrook,
Bedford, United Kingdom, 1997.
16. Graboski, M.S.; Ross, J.D.; McCormick, R.L. Transient Emissions from
No. 2 Diesel and Biodiesel Blends in a DDC Series 60 Engine; SAE 961166;
Society of Automotive Engineers: Warrendale, PA, 1996.
17. Clark, N.N.; Atkinson, C.M.; Thompson, G.J.; Nine, R.D. Transient
Emissions Comparisons of Alternative Compression Ignition Fuels; SAE
1999-01-1117; Society of Automotive Engineers: Warrendale, PA, 1999.
18. Schaberg, P.W.; Myburgh, I.S.; Botha, J.J.; Roets, P.N.; Viljoen, C.L.;
Dancuart, L.P.; Starr, M.E. Diesel Exhaust Emissions Using Sasol Slurry
Phase Distillate Process Fuels; SAE 972898; Society of Automotive En-
gineers: Warrendale, PA, 1997.
19. Galey, M.H. Catalyst and CRT Technologies for Control of Heavy Ve-
hicle Exhaust Emissions. In Trucks and Buses, the Results of Real-World
Emissions Testing, Millbrook Proving Ground Seminar Proceedings; Millbrook
Proving Ground Ltd.: Millbrook, Bedford, United Kingdom, 1997.
20. Bach, E.; Zikoridse, G.; Sandig, R.; Lemaire, J.; Mustel, W.; Naschke,
W.; Bestenreiner, G.; Brück, R. Combination of Different Regeneration
Methods for Diesel Particulate Traps; SAE 980541; Society of Automo-
tive Engineers: Warrendale, PA, 1998.
21. Hawker, P.; Hüthwohl, G.; Henn, J.; Koch, W.; Lüders, H.; Lüers, B.;
Stommel, P. Effect of a Continuously Regenerating Diesel Particulate Fil-
ter on Non-Regulated Emissions and Particulate Size Distribution; SAE
980189; Society of Automotive Engineers: Warrendale, PA, 1998.
22. Clark, N.N.; Prucz, J.C.; Gautam, M.; Lyons, D.W. The West Virginia
University Heavy-Duty Vehicle Emissions Database as a Resource for In-
ventory and Comparative Studies; SAE 2000-01-2854; Society of Auto-
motive Engineers: Warrendale, PA, 2000.
23. Yanowitz, J.; McCormick, R.L.; Graboski, M.S. In Use Emissions from
Heavy-Duty Diesel Vehicles; Environ. Sci. Technol. 2000, 34 (5), 729-
740.
24. Wimmer, R.R.; Clark, N.N.; McKain, D.L.; Fletcher, J. Emissions Test-
ing of a Hybrid Fuel Cell Bus; SAE 980680; Society of Automotive Engi-
neers: Warrendale, PA, 1998.
25. Ramamurthy, R.; Clark, N.N.; Atkinson, C.M.; Lyons, D.W. Models for
Predicting Transient Heavy-Duty Vehicle Emissions; SAE 982652; Society
of Automotive Engineers: Warrendale, PA, 1998.
REFERENCES
1. Protection of the Environment. Code of Federal Regulations, Part 86,
Subpart N, Title 40, 1996.
2. Ramamurthy, R.; Clark, N.N. Environ. Sci. Technol. 1999, 33, 55-62.
3. Lyons, D.W.; Bata, R.M.; Wang, W.G.; Clark, N.N.; Palmer, G.M.; Howell,
A.D.; Loth, J.L.; Long, T.R. Design and Construction of a Transportable
Heavy-Duty Vehicle Emission Testing Laboratory; SAE Special Publication
P-256; Society of Automotive Engineers: Warrendale, PA, 1991.
4. Fuel Economy Measurement Test (Engineering Type) for Trucks and Buses,
SAE Handbook, Vol. 4; SAE J1376; Society of Automotive Engineers:
Warrendale, PA, 1993.
5. Clark, N.N.; McKain, D.L.; Messer, J.T.; Lyons, D.W. Chassis Test Cycles
for Assessing Emissions from Heavy-Duty Trucks; SAE 941946; Society
of Automotive Engineers: Warrendale, PA, 1994.
6. Clark, N.N.; Lyons, D.W.; Bata, R.M.; Gautam, M.; Wang, W.G.;
Norton, P.; Chandler, K. Natural Gas and Diesel Transit Bus Emissions:
Review and Recent Data; SAE 973203; Society of Automotive Engineers:
Warrendale, PA, 1997.
About the Authors
Justin M. Kern participated in this research as part of his
M.S. study in the department of Mechanical and Aerospace
Engineering (MAE) at West Virginia University (WVU), and has
recently joined the staff of Argonne National Laboratory,
Argonne, IL. Nigel N. Clark (corresponding author; e-mail:
nclark@wvu.edu) is the George Berry Chair of Engineering
and Ralph D. Nine is a program coordinator, both at WVU
MAE. At the time of writing, Christopher Atkinson was as-
sociate professor of MAE at WVU. He is now chief engineer
at Calico Systems, Morgantown, WV.
7. Graboski, M.S.; McCormick, R.L.; Yanowitz, J.; Ryan, L. Heavy-Duty
Diesel Vehicle Testing for the Northern Front Range Air Quality Study;
Colorado School of Mines: Fort Collins, CO, 1998.
94 Journal of the Air & Waste Management Association
Volume 52 January 2002