EP A/600/A-97/093
Comparison of Emission Models with On-Road Heavy-Duty Diesel Modal Data
by
J, Edward Brown
Acurex Environmental Corporation
P.O. Box 13109
Research Triangle Park, NC 27709
and
D. Bruce Harris and Foy G. King
National Risk Management Research Laboratory
U. S. Environmental Protection Agency
Research Triangle Park, NC 27711
INTRODUCTION
As progressively more states and local agencies develop State Implementation Plans (SIPs)
to meet ambient air quality standards, mandated by the Federal Clean Air Act, it is clear that sources
of ozone precursors and particulate matter (PM) will receive considerable attention. For those areas
that are striving to meet ozone and fine particulate air quality goals, it is becoming increasingly
apparent that mobile sources, primarily internal combustion engines, are contributing a substantial
portion of the overall emission inventory. The diesel engine, already recognized as a major source
of ozone precursor (volatile organic compound [VOC] and nitrogen oxide [NOJ) and PM emissions,
has become even more of a focus as reductions from gasoline engines become harder to obtain.
For the benefit of the emissions inventory community and regional air quality modelers,
EPA's Office of Mobile Sources (OMS) has produced a model, the latest version of which is called
MOBILE5b, that predicts mobile source emission factors for ozone precursors and carbon monoxide
(PM exhaust emission factors are estimated with a similar, but separate, model: PARTS). This
model was developed using information on motor vehicle fleet characteristics, usage patterns, and
the regulatory and economic forces that affect them. Among a vehicle's pertinent characteristics
are its weight, fuel type, aerodynamic configuration, and emission control technology. Because
1
-------
many of these parameters vary by model year, the MOBILE model internally calculates emissions
contributions separately by model year, vehicle class, and fuel type. The final outputs from the
MOBILE model are composite emissions factors, usually expressed as grams/vehicle-mile of travel,
in which all of these factors have been weighted together to produce average emission factors for
each pollutant and each vehicle class and fuel type for a specific calendar year.
To develop the MOBILE model, it was necessary to incorporate emissions test data for each
type of vehicle. Among the most readily available sources of emissions data for motor vehicles are
the certification test results. Passenger car and light-duty truck models, prior to being sold in the
United States, must be certified on a chassis dynamometer, using a test cycle that is meant to
simulate the full range of "normal" vehicle operation. This type of test yields an emission
measurement (grams/mile), which is used to calculate average emission factors for development
of emission inventories.(1) Because off-road and heavy-duty vehicles often include engines from
different manufacturers (bringing about an enormous number of potential vehicle/engine
combinations), and because chassis dynamometers are limited in their load capacity, it is the
engines that are certified for these classes of vehicles.
The engine dynamometer test results are expressed in grams per brake horsepower-hour
(g/BHP-hr). In order to convert the results to grams per mile, a series of "conversion factors" (CFs),
in units of BHP-hr/mile, was developed.(2) These CFs are calculated from fuel density, engine
brake specific fuel consumption (BSFC), and fuel economy (FE). Although the CF development
process performs separate calculations for each vehicle class and model year, the data that are
actually incorporated into the MOBILE model is a single "Fleet-Average CF" for each model year.
In response to EPA's desire that MOBILE use accurate CFs reflecting actual on-road emissions,
its Air Pollution Prevention and Control Division (APPCD) has begun a validation program to confirm
or modify those CFs using actual on-road-measured emissions data. (3)
2
-------
BACKGROUND
Developers of emissions inventories require, and MOBILE calculates, emission rates in
grams per vehicle-mile. This value is the product of two primary inputs and several "correction
factors." The primary inputs are the base emission (BE) rates and the conversion factor (CF). The
"correction factors" adjust for conditions that do not correspond to BE measurement conditions. For
heavy-duty dies'ei vehicles (HDDVs), the only correction that is applied is a speed correction, which
accounts for traffic speeds other than 20 MPH (which BE rates represent). The BE values that are
incorporated into the model are meant to represent the nationwide fleet, and come from a
combination of engine dynamometer testing and assumptions about regulatory compliance.
CFs are calculated, for each vehicle class and fuel type, using the formula
CF
1 bhp-hou. ^
mile
p (lb/gallon)
BSFC (lb/bhp-hour) x FE {miles/gallon)
where p is fuel density. Table 1 shows the input parameters, along with the data sources for the
most recent versions of the MOBILE model. A number of sources can provide fuel density values,
with relatively little variation among them; the most recent versions of the MOBILE model use data
from Motor Vehicle Manufacturers Association (MVMA)1 fuel surveys.(4) BSFC values are
interpolated from a combination of engine certification test results (pre-1979) and responses to
direct inquiries of manufacturers. Truck fuel economy statistics come from the U.S. Department of
Commerce's quintennial Truck Inventory and Use Survey (TIUS).
The APPCD validation program currently has on-road data available for two heavy-duty
trucks. The first test tractor was a 1989 Ford CL9000 Cab-over with a 315 hp Cummins NTC-315
engine and approximately 105,000 miles of primarily short-trip, no-load use. The second test tractor
'Current title: American Automobile Manufacturers Association (AAMA)
3
-------
Table 1: Inputs to MOBILE Model Conversion Factor Calculation
Fuel Density
BSFC
Fuel Economy
Carbon content and mass
fraction.data published in
40CFR Part 600 - MOBILE3
Motor Vehicle Manufacturers
Association (MVMA) -
MOBILE4&5
Engine Manufacturers
Association (EMA) -
MOBILE3-5
Individual Engine
Manufacturer Contacts
MOBILE3-5
1972 TIUS -
MOBILE3-5
1982 TIUS -
MGBILE4&5 :
was a 1990 Freightliner with a 325 hp Caterpillar 3176 engine, showing over 550,000 miles on its
odometer; its history and maintenance condition are unknown.(5) The on-road emissions dataset
consists of parametric test data (constant load, grade, and speed/acceieration conditions),
representative urban route data (delivery, interstate bypass, and terminal entry/exit), and on-road
simulations of dynamometer test cycles. This paper will analyze the MOBILE conversion factors,
in the context of how well they represent the on-road test data.
CONVERSION FACTOR (CF) ANALYSIS
The CF methodology relies on 'several assumptions, many of which were made for lack of
better or contradictory data. First, it was assumed that a single CF could represent all trucks of a
given model year. This assumption is being abandoned, beginning with MOBILE6.(6) Second, it
was assumed that brake-specific emissions can be used to effectively model emissions of all
pollutants. The on-road data show that the approach of using a power-demand-based model may
be feasible for accurately estimating NOx emissions.(5) Third, it was assumed that the fuel statistics
that serve as inputs to the CF calculations (fuel density, engine BSFC, and on-road fuel economy)
are representative of suitably realistic and similar conditions, justifying their combination into a
simplified CF. It is the third assumption that will be addressed here.
4
-------
Fuel Density
There is little doubt that the fuel density values used in the MOBILE models are
representative of the overall fuel supply. Figure 1 shows several values referenced in the CF
technical reports (4,7), as well as several density measurements that have been made during the
on-road emissions study. The MOBILE3 value is calculated from carbon mass fraction and carbon
mass per unit fuel volume values published in the "Fuel Economy of Motor Vehicles" section of
40CFR,(8) but the 1983 MVMA fuel survey value is included in the EPA technical report for
comparison.(7) MOBILE4 abandons the CFR values in favor of the average of the 1982-1985
MVMA fuel survey values (7.11 lb/gal). All of the values, including those measured from the on-road
facility's fuel supply, are within ±1% Relative Standard Deviation (RSD).
Fuel Density, lb/gallon
0 2 4 6 8 10
MOBILE3 value from CFR
1982 MVMA fuel survey
1983 MVMA fuel survey
1984 MVMA fuel survey
1985 MVMA fuel survey
Sample collected 12-16-93
Sample collected 3-12-96
Sample collected 1-20-97
Sample collected 9-8-97
Figure 1: Comparison of Diesel Fuel Densities
Brake-Specific Fuel Consumption
Figures 2a and 2b show estimates, derived from on-road measurements, of BSFC for the
first two test trucks, aJong with the most recent MOBILE BSFC of 0.39 lb/bhp»hr (which is applied
to all model years 1987 and later). Other than the obvious upturn at low horsepower, there appear
5
i i i r
//////////////////////w
7.07
7-098
7.131
y/////////////////////////////////////////////////////
7.102
7.114
7.06
7.04
7.05
7.00
-------
Brake Specific Fuel Consumption
1989 Ford CL-9000 with Cummins NTC-31S Engine
1.5
G.
X
m
o
u.
-------
to be no statistically significant trends in the on-road BSFC data. The moderate-to-high-horsepower
BSFC values correspond to the MOBILE line, within the precision of the data. The on-road driving
cycles, by no means outlyers on the figures, do show a modest upturn in BSFC. The Freightliner
data show elevated values under acceleration conditions, but the effect cannot be statistically
quantified because of the overall scatter in that truck's data. It is likely that, under more controlled
test conditions, it may be possible to quantify the effect of transient operation on BSFC in a
statistically defensible manner. Nevertheless, the current BSFC values incorporated into the
MOBILE conversion factors represent these two trucks reasonably well.
Fuel Economy
The engine certification cycle is generally accepted as "representative" of urban driving.
Therefore the use of BSFC values from this test is seldom criticized, at least not for "urban"
conversion factors. What has often been brought up as a "loose end" in the conversion factor
methodology is the use of TIUS fuel economy values, which may not represent urban operation.
(4,7) Fortunately, the on-road facility has furnished a large amount of fuel economy data under a
variety of operating conditions. Figures 3a and 3b show some of the fuel economy measurements;
each bar represents average fuel economy for the specified load (pounds Gross Combined Weight,
GCW), and operating condition (test cycle, route, constant speed, or acceleration). Bars are
pattern-coded to facilitate comparisons.
All but the "delivery route" measurements represent level grade; these measurements are
included for comparison with the on-road transient driving cycles (the top three bars in each graph)
that were run at the same load. These cycles, the West Virginia University (WVU) 5-Peak Driving
Cycle (9), EPA's HDDV Schedule-D (10), and an on-road federal test procedure (FTP) simulation
(9), were developed as in-chassis corollaries to the engine certification test. A fuel economy
comparison shows that those cycles, all of which contain a considerable amount of transient
operation, actually represent urban operation reasonably well.
7
-------
Fuel Economy, MPG
198S Ford CI.-9000 with Cummins NTC-315 Engine
0 2 4 6 8 10
WVU 5-Peak Criving Cycle
HDDV Schedule-D
FTP OrvRoad Simulation
Half-Loaded Delivery Route
73110 Accel, Progressive
49500 Accel. Progressive
25890 Accel. Progressive
73110 Accel, Governed
49500 Accel, Governed
25890 Accel, Governed
73110 GCW, 55 MPH
73110 GCW, 55 MPH
49500 GCW, 55 MPH
49500 GCW, 55 MPH
25890 GCW, 55 MPH
25890 GCW, 55 MPH
73110 GCW, 35 MPH
49500 GCW, 35 MPH
25890 GCW, 35 MPH
73110 GCW, 15 MPH
73110 GCW, 15 MPH
49500 GCW. 15 MPH
49500 GCW, 15 MPH
25890 GCW, 15 MPH
25890 GCW, 15 MPH
V//////////////////////////77Z
V///////////////////////////S//////////S///77777Z
V///////////////////////////////////////////&
V///////////////////////////A
Low Gear
High Gear
Low Gear
Low Gear
High Gear
High Gear
Low Gear
Low Gear
Low Gear
High Gear
High Gear
oooo<>oc«x>coooQW?wcoc>»i High Gear
Figure 3a: Truck 1 Fuel Economy
Fuel Economy, MPG
1990 Freightliner with Caterpillar 3176 Engine
10
15
WVU 5-Peak Driving Cycle
HDDV Schedule-D
FTP On-Road Simulation
Fully-Loaded Delivery Route
75440 Accel, Progressive
51830 Accel. Progressive
28220 Accel. Progressive
75440 Accel. Governed
51830 Accel. Governed
28220 Accel. Governed
75440 GCW. 55 MPH
75440 GCW. 55 MPH
51830 GCW. 55 MPH
51830 GCW. 55 MPH
28220 GCW. 55 MPH
28220 GCW. 55 MPH
75440 GCW, 35 MPH
51830 GCW. 35 MPH
28220 GCW. 35 MPH
75440 GCW. 15 MPH
51830 GCW. 15 MPH
28220 GCW. 15 MPH
7/////////////////////SM
'/////////////////////////A
•//////////////A
'///////////////*
'////////////Z777A
'//////////////A
W////////////A
l°w Gear
High Gear
§§§§§! Low Gear
High Gear
Low Gear
High Gear
Figure 3b: Truck 2 ruel Economy
8
-------
The parametric data show that, for both trucks at all loads, the peak fuel economy was at
the 35 MPH test condition. The drop in fuel economy at slow speeds is a result of the rapid rise in
BSFC at near-minimal power demands (i.e., the engine uses proportionately more fuel to operate
accessories and overcome internal friction as external power demands are decreased). As speeds
increase above 35 MPH, the wind drag component of on-road power demand (nominally
proportional to velocity squared) causes fuel economies to fall sharply. The MOBILE model
incorporates this second-order variation of FE with speed in the form of a "speed correction factor,"
which reaches a minimum value at about 33.8 MPH for HDDVs. The model makes no attempt to
distinguish between "urban" and "interstate" driving modes. In so doing, it neglects the effects of
idle emissions and acceleration modes on urban emission inventories. Not surprisingly, both on-
road test trucks see their lowest fuel economies of all during acceleration test conditions.
Some of the load/speed conditions are represented by more than one bar. In these cases,
the prescribed condition was established with more than one gear ratio. The purpose of this
duplication was to investigate RPM effects on emissions, but there was obviously an effect on fuel
economy, as well. The effect was most pronounced on the Ford truck, with 15 MPH fuel economies
suppressed by 30% during low-gear, high-RPM testing. Nonetheless, since this effect is likely to
vary significantly from truck to truck, no attempt will be made to incorporate any type of "gearing"
correction into the conversion factors.
For a more realistic comparison of fuel economies, we compare our "delivery" route to our
"interstate" route, both of which represent multiple speeds and grades, as dictated by topography
and traffic conditions. Figures 4a and 4b compare the fuel economy (FE) measurements for the two
trucks. As expected, the "interstate" values are consistently higher than those for the "delivery"
route, with load and overall grade (the eastbound interstate has more downhill driving than uphill)
having predictable effects. Each bar represents three test runs of the same route; 90% confidence
intervals are shown. A popular assumption in the modeling community is that long-haul (interstate)
9
-------
Fuel Economy versus Operating Mode
1989 Ford CL-9000 with Cummins NTC-315 Engine
® Delivery Route ¦ Westbound Interstate C3 Hastbound interstate
Figure 4a: Truck 1 Modal Fuel Economy Comparisons
Fuel Economy versus Operating Mode
1950 Freightliner with Caterpillar 3176 Engine
7.3*0.1
4.4s 0.3
Empty Hail Full
WX Delivery Route ¦ Westbound interstate ~ Eastbound Interstate
Figure 4b: Truck 2 Modal Fuel Economy Comparisons
10
-------
trucks are half- to fully-loaded most of the time, while delivery trucks are most often empty to half-
loaded. Tables 2a and 2b cross-compare all of the loads, with the statistical significance of the
differences shown. The interstate fuel economy values are combined into a single, "round-trip"
figure for this comparison. It appears that significant fuel economy differences exist between
delivery and interstate driving for all but one comparison on one of the two trucks.
Table 2a: Significance of FE Differences Table 2b: Significance of-FE Differences
Interstate
Empty
Delivery
Half
Full
Empty
Half
Full
98.76%
98.62%
Overlap
99.76%
99.80%
90.98%
99.94%
99.96%
99.47%
Interstate
Empty
Delivery
Half
Full
Empty '
Half
Full
99.99%
99.97%
95.37%
99.99%
100.00%
99.61%
100.00%
99.97%
89.96%
EMISSIONS MODEL COMPARISON
The MOBILE model calculates heavy-duty vehicle emissions as the product of the base
emission (BE) rate, correction factors, and the conversion factor (CF). Within the MOBILE program,
BE rates are specified as zero-mileage levels (ZMLs) and deterioration rates (DRs). The DR is a
value that is added to ZML for each 10,000 miles accumulated on a vehicle or engine. ZML and DR
are tabulated for each pollutant as a function of vehicle class, mode! year, and fuel type, but the
MOBILE program also includes an option to accept user-supplied values. The speed correction
factor (SCF) is a dimensionless value which is calculated as a function of vehicle speed.
Parameters for calculating SCFs are tabulated as a function of vehicle class and pollutant. CF is
tabulated as a function of model year and fuel type only.
Table 3 shows the ZML, DR, SCF equation, and CF values for HDDVs of the 1989 (Truck
1) and 1990 (Truck 2) model years. The zero DR values indicate that average BE rates are
expected to remain the same (equal to ZML) throughout the life of that model year of trucks. The
-------
SCF equation does not vary by model year. The CF that is used in MOBILE is a fleet average CF,
intended to represent all classes of HDDVs.
Table 3: MOBILESb Input Parameters for Tested Model Years (11)
Model
ZML
DR
SCF equation
CF
Year
g/mile
g/raile
bhp'hr/mile
1989
16.77
0
SCF = EXP(0.676
- 0.Q48*MPH +
0. Q0071*MPH3)
.2.1
1990
11.65
0
SCF = EXP(0.676
0 . 048*MPH +
0 . 00071*MPH2)
2.07
Table 4 shows values that are substituted for MOBILE inputs in order to develop alternate
versions of the MOBILE model. The Certification BE values are based on the actual reported
engine certification test results for the engine family and model year of the trucks' engines.(12,13)
Class CF is the class specific conversion factor which was developed during the most recent CF
analysis, but was not actually incorporated into the model. Delivery CF and Interstate CF are the
modal conversion factors that were calculated from on-road test data for each truck. The calculation
follows the same general outline as the MOBILE CF calculation, except that the inputs are specific
to the trucks. The BSFC input comes from an on-road simulation of the engine certification test.
The FE inputs are specific to the actual routes (delivery or interstate) that they represent, which is
what makes the resulting values "modal" CFs. MOBILE fuel density inputs were used for the modal
CF calculations.
Table 4: Substitute Inputs for Alternative Models
Test Vehicle
Cert BE
Class CF
Delivery CF
Interstate CF
Truck 1
7.88
3.129
3 .278
2.962
Truck 2 _
6.12
3 .129
3 .253
2.562
12
-------
Figures 5a and 5b compare actual "delivery" and Interstate" emissions to various modeling
approaches. The "delivery" route is a 57 mile loop, traversing downtown Raleigh, NC, that includes
three stops at retail malls. The measurements in the figures represent an empty to half-loaded
truck. The "interstate" route is a 75 mile section of 1-40, passing through Raleigh and Durham, NC.
The measurements in the figures represent a half- to fully-loaded truck.
Uncorrected BE rates are included among the built-in inputs to the MOBILE model.
"MOBILE BE" represents those rates, with the speed correction factor (SCF) applied. These are the
emissions that MOBILE would predict, given only the model year, fuel type, operating speeds, and
the pollutant of interest The "Cert BE & MOBILE CP bars show the emissions that MOBILE would
predict if, in addition, it were given the engine certification results for the test engines; MOBILE uses
its buiit-in fleet-average conversion factors to convert engine certification results to grams/miie. .
The "Cert BE & Class CF" bars substitute a class specific CF for the fleet average CF that is built
into MOBILE. The "Cert BE & Modal CF' bars abandon MOBILE CFs and SCFs in favor of mode-
specific CFs that are derived from trends observed in the on-road test data.
In comparing same-truck emissions between delivery and interstate operating modes, there
are primarily two factors affecting the comparison. First is the difference in average speed, which
would tend to increase interstate emissions relative to delivery emissions. This effect is
incorporated into the MOBILE model through its SCF, which is a second order equation that reaches
a minimum at about 33.8 MPH. It is because of this correction that the predicted interstate
emissions are consistently higher that the corresponding predicted delivery emissions for each of
the non-modal models. The second difference between delivery and interstate modes is due to idle
time and acceleration. This effect is not incorporated into the MOBILE model, but shows up in the
measured data as delivery emissions that meet or exceed interstate emissions for the same truck.
The modal model arrives at its correction factors by direct comparison of fuel economy data
between operating modes. Therefore, this model takes both speed and idle/acceleration effects into
13
-------
Comparison of Modeling Approaches
1989 Ford CL-9000 with Cummins NTC-315 Engine
NOx grams/mile
0 10 20 30 40
! 1 J
¦I Measured
~ MOBILE BE
m Cert BE & MOBILE CF
11 Cert BE & Class CF
B3 Cert BE & Modal CF
Interstate
Figure 5a: Truck 1 Comparison of Modeling Approaches
Comparison of Modeling Approaches
1990 Fre.ightliner with Caterpillar 3176 Engine
NOx grams/mile
Deli\«ry
Interstate
¦ Measured
a MOBILE BE
m Cert BE & MOBILE CF
fil Cert BE & Class CF
E3 Cert BE & Modal CF
Figure 5b: Truck 2 Comparison of Modeling Approaches
14
-------
consideration. This explains why it is the only model that consistently predicts emissions within
±10% of the measured value.
CONCLUSIONS AND RECOMMENDATIONS
Of the three inputs to the MOBILE conversion factor calculations, two of them represented
real world data quite well. Fuel density varies very little between any of the references and
measurements. Measured BSFC values for both test trucks are concentrated around the MOBILE
value of 0.39 Ib/bhp*hr. Some of MOBILE'S assumptions regarding fuel economy, however, may
need to be revised. The existence of significant FE differences between "delivery" and "interstate"
operating modes would indicate that two sets of conversion factors are needed to accurately model
emissions from those two modes.
ACKNOWLEDGMENTS
The authors gratefully acknowledge the contributions of the following: Bob Franklin and
Sherman Graham of the North Carolina Truck Driver Training School for providing the test vehicles,
service, and background information; Bill Mitchell and Bill Squire of Acurex Environmental Corp. for
their guidance and hard work in designing and building the test facility; and L.C. Smith and Todd
Murray of the North Carolina Department of Transportation for providing detailed maps of the test
roadways. .
This work was sponsored by the Emissions Characterization and Prevention Branch of EPA-
NRMRL's Air Pollution Prevention and Control Division, with additional funding from the North
American Research Strategy for Tropospheric Ozone (NARSTO).
REFERENCES
1. Federal Register, 40 CFR Part 86, Subpart N, U.S. Government Printing Office, Washington,
DC, 1991.
2, Energy and Environmental Analysis, Inc., "Historical and Projected Emissions Conversion
Factor and Fuel Economy for Heavy-Duty Trucks 1962-2002," prepared for the Motor
Vehicle Manufacturers Association, Detroit, Ml, 1983.
-------
3. Harris, D. Bruce, King, Foy G., and Brown, J. Edward, "Development of On-Road Emission
Factors for Heavy-Duty Vehicles," presented at the Diesel Engine Emissions Reduction
Workshop, San Diego, CA, July 24,1995.
4. Machiele, Paul A., "Heavy-Duty Vehicle Emission Conversion Factors II, 1962-2000," Report
EPA-AA-SDSB-89-01 (NTIS PB89-196349), U.S. EPA, Office of Mobile Sources, Ann Arbor,
Ml, 1988.
5. Brown, J. Edward, Harris, D. Bruce, and King, Foy G., "On-Road NOx Emissions from
Heavy-Duty Diesel Vehicles: A Two-Truck Study," In Review.
6. Conference call between Terry Newell, EPA Office of Mobile Sources, and Louis Browning,
Mike Jackson, and Ed Brown of Acurex Environmental Corporation, July 15,1997.
7. Smith, Mahlon C., "Heavy-Duty Vehicle Emission Conversion Factors 1962-1997," Report
EPA-AA-SDSB-84-1 (NTIS PB85-111375), U.S. EPA, Office of Mobile Sources, Ann Arbor,
Ml, 1984.
8. Federal Register, 40 CFR Part 600, §600.113-78, U.S. Government Printing Office,
Washington, DC, 1983.
9. Clark, Nigel N., et al., "Chassis Test Cycles for Assessing Emissions from Heavy Duty
Trucks," SAE 941946, Society of Automotive Engineers, Warrendale, PA, 1994.
10. Federal Register, 40 CFR Part 86, Appendix I (d), U.S. Government Printing Office,
Washington, DC, 1991.
11. EPA Office of Mobile Sources, MOBILE5b FORTRAN source code, files dated 9-14-96,
downloaded from OAQPS Technology Transfer Network.
12. Environmental Protection Agency, "Control of Air Pollution from New Motor Vehicles and
New Motor Vehicle Engines: Federal Certification Test Results 1989 Model Year," 1989.
13. Environmental Protection Agency, "Control of Air Pollution from New Motor Vehicles and
New Motor Vehicle Engines: Federal Certification Test Results 1990 Model Year," 1990.
METRIC EQUIVALENTS
Readers more familiar with the metric system may use the following equivalents to convert
to those units:
5/9(°F-32) = °C
1 ft = 0.3048 m
1 ft3 = 0.0283 m3
1 gal. = 3.785 I
1 hp = 745.7 W
1 lb = 0.4536 kg
1 mi = 1.609 km
16
-------
ATDH/rar DTB-D OQ/1 TECHNICAL REPORT DATA
IN rt IVLtt Xj U1 r ir ZOff (Please read Instructions on the reverse before complel
t. REPORT NO. 2,
EPA/600/A-97/093
3.
4. TITLE ANO SUBTITLE
Comparison of Emission Models with On-road Heavy-
duty Diesel Modal Data
5. REPORT DATE
6. PERFORMING ORGANIZATION CODE
7. author(s) j grown (Acurex), and D. B. Harris and
F.G.King (EPA)
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Acurex Environmental Corporation
P. 0. Box 13109
Research Triangle Park, North Carolina 27709
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-D4-0005 WA 31
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Air Pollution Prevention and Control Division
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Published paper; 10/95-9/97
14. SPONSORING AGENCY CODE
EPA/600/13
15,supplementary notes ^ PPCD project officer is D. Bruce Harris, Mail Drop 61, 919/541-
7807. Presented at AWMA Conference, Emission Inventory: Planning for the Future,
Research Triangle Park, NC, October 28-29, 1997.
is.abstract paper discusses a validation program aimed at confirming or modifying
conversion factors used in EPA's MOBILE model, using actual on-road-measured
emissions data. For the benefit of the emissions inventory community and regional
air quality modelers, EPA's Office of Mobile Sources has produced a model, the
latest version of which is called MOBILE5b, that predicts mobile source emission
factors for ozone precursors and carbon monoxide (particulate exhaust emission
factors are estimated with a similar, but separate, model, PARTS). This model
was developed using information on motor vehicle fleet characteristics, usage
patterns, and the regulatory and economic forces that effect them. Among a ve-
hicle's pertinent characteristics are its weight, fuel type, aerodynamic configura-
tion, and emission control technology. Because many of these parameters vary by
model year, the MOBILE model internally calculates emissions contributions se-
parately by model year, vehicle class, and fuel type. The final outputs from the
MOBILE model are composite emissions factors, usually expressed as grams/ve-
hicle-mile of travel, in which all of these factors have been weighted together to
produce average emission factors for each pollutant and each vehicle class and fuel
type for a specific calendar year.
17. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Pollution Particles
Mathematical Models
Emission
Diesel Engines
Ozone
Carbon Monoxide
Pollution Control
Mobile Sources
Particulate
13 B
12 A
14 G
21G
07B
18. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
------- |