ANALYSIS OF HEAVY DUTY TRUCK
FUEL EFFICIENCY TO 2001
Prepared for:
ENVIRONMENTAL PROTECTION AGENCY
Energy Policy branch
Washington, D.C.
Prepared by:
ENERGY AND ENVIRONMENTAL ANALYSIS
1655 Norch Fort Myer Drive, Suite 600
Arlington, Virginia 22209
September 1991
-------
TABLE OF CONTENT
1. INTRODUCTION 1-1
2. HEAVY - DUTY TRUCK FLEET CHARACTERISTICS 2-1
2.1 Introduction 2-1
2 2 Data Used In the Analysis ... 2-2
2.3 Characteristics of Heavy-Duty Trucks ... 2-4
3. OPERATIONAL FACTORS TO IMPROVE FUEL EFFICIENCY 3-1
3.1 Introduction 3-1
3.2 Federal Regulations on Weight 3-5
3 3 Analytical Methodology 3-12
3.4 The Effect of Empty Mileage 3-18
3.5 Fuel Productivity Under An Uncapped Bridge Formula B .... 3-25
3.6 Summary of Results 3-38
4. TECHNOLOGICAL IMPROVEMENTS TO TRUCKS 4-1
4 1 Overview . 4-1
4 2 Fuel Consumption Sensitivity 4-2
4 3 Weight Reduction ... 4-9
4.4- Aerodynamic Drag .... 4-13
4 5 Rolling Resistance Reduction 4-15
4.6 Improvements to the Engine 4-18
4.7 Turbocompound Diesel Engines 4-29
4.8 Drivetrain Optimization .... .... , . 4-31
4.9 Electronic Control 4-33
4.10 Other Improvements 4-35
4 LI Total Improvement in Fuel Efficiency . . . ... 4-36
APPENDIX A: STATISTICAL RESULTS OF THE REGRESSION MODELS
TO ESTIMATE GPM
-------
LIST OF TABLES
Table 2-1 Distribution of Trucks By Engine Type and GVW Class .... 2-5
Table 2-2 Sales and Diesel Penetrations By Class . .... . 2-5
Table 2-3 Penetration Rates of Fuel Economy Options By GVW Class and
Engine Type 2-10
TabLe 2-4 Gasoline Trucks by Use and Body Style Percent By GVW Class 2-12
Table 2-5 Diesel Trucks by Use and Body Style Percent by GVW Class . 2-13
Table 2-6 Percent of Trucks by Area of Operation By GVW Class and
Fuel Type . 2-14
Table 3-1 State Size Limits ... 3-8
Table 3-2 Interstate National Network Allowable Gross Weight ... 3-9
Table 3-3 Definitions of TIUS Variables Used In The Analysis . 3-14
Table 3-4 Definitions of EEA Variables 3-16
Table 3-5 Listing of Fitted Models ... 3-19
Table 3-6 Statistical Means of Selected Variables By GVW Class and
Engine Type Combination . . . . . . 3-21
Table 3-7 The Effect of Empty Mileage on Fuel Productivity . . . . 3-23
Table 3-3 Physical and Operational Characteristics of Weight-Limited
5-Axle or Greater Trucks . . . .3-27
Table 3-9 Estimates of Fuel Productivity Gains From Eliminating the
80,000 lb GVW Cap . 3-29
Table 3-10 Typical Dimensions and Volumetric Capacities For 5-,
6-Axle Tractor-Semis, Twin 28s, an LCVs . . . . 3-33
Table 3-11 Physical and Operational Characteristics of Size-Limited
5-Axle or Greater Trucks ... . . 3-35
Table 3-12 Fuel Productivity Estimates for Size-Limited Vehicles
and Potential Gains From Shifts to LCVs . . 3-36
ii
-------
LIST OF TABLES
(Continued)
Table 4-1 Sensitivity of Fuel Economy to Truck Load . . .
Table 4-2 Weight Savings for Specific Aluminum Parts . .
Table 4-3 Comparison of Engine Weights
Table 4-4 Engine Manufacturers Association Brake Specific Fuel
Consumption Heavy-Duty Diesel Engines . .
Table 4-5 BSFC For Selected 1991 Engines (LB/BHP-HR) .
Table 4-6 Improvements to Engine BSFC, 1987-2001 . . .
Table 4-7 Improvements to Class 8B Trucks, 1987 - 2001
Table 4-8 Improvements to Class 6/7 Trucks, 1987 - 2001
Page
4-5
4-10
4-12
4-19
4-20
4-27
4-37
4-38
ill
-------
LIST OF FIGURES
Page
Figure 2-1 Average CID By Vintage and GVW Class Gasoline Trucks . . . 2-7
Figure 2-2 Average Horsepower By Vintage and GVW Class 2-8
Figure 2-3 Average Percentage of Annual Mileage When No Load Was
Carried, By GVW Class 2-15
Figure 2-4 Average Operating Equivalent Weight By GVW Class 2-17
Figure 2-5 Average MPG By Vintage GVW Class 6 Gasoline Trucks ... 2-18
Figure 2-6 Average MPG By Vintage GVW Class 7 Gasoline Trucks ... 2-19
Figure 2-7 Average MPG By Vintage GVW Class 8A Gasoline Trucks .... 2-20
Figure 2-8 Average MPG By Vintage GVW Class 6 Diesel Trucks ... 2-21
Figure 2-9 Average MPG By Vintage GVW Class 7 Diesel Trucks . . 2-22
Figure 2-10 Average MPG By Vintage GVW Class 8A Diesel Trucks .... 2-23
Figure 2-11 Average MPG By Vintage GVW Class 8B Diesel Trucks . . . 2-24
Figure 3-1 Truck Types 3-2
Figure 4-1 Long Haul Central Europe . 4-8
Figure 4-2 Long Haul Central Europe ... .... . . 4-17
Figure 4-3 Effect of Reducing Friction on Engine Efficiency 4-24
iv
-------
1. INTRODUCTION
Heavy-duty trucks are an increasingly important source of carbon dioxide
emissions, since the rate of growth of fuel consumption has been substantially
higher than for other on-highway transportation sectors. Currently, heavy-
duty trucks, defined by EPA as all trucks with a gross vehicle weight (GW)
greater than 8500 lb, account for about 28 percent of total on-highway fuel
consumption. As a result, there is considerable interest regarding this
market segment's future fuel consumption. The EPA's interest lies in
identifying areas where there is scope for the implementation of policies to
reduce fuel consumption without changing the level of service provided by the
trucking industry.
Unlike the light-duty segments of the fleet, there is no standardized measure
of fuel economy for heavy-duty trucks. The only reliable source of
information on truck fuel economy is the Truck Industry and Use Survey (TIUS).
This survey is conducted by census once every five years and the 1987 TIUS,
conducted in calendar year 1988, is the most recent publicly available
version. The TIUS was analyzed to establish historical values of fuel economy
and fuel economy improvement The TIUS also contains a wealth of other data
on truck annual use, scrappage, distribution by weight class and body style,
etc all of which are relevant to a fuel economy study. The analysis of TIUS
data and the results are documented in Section 2
Truck fuel economy is more accurately measured in terms of ton-miles of
payload carried per gallon, which is affected by operational and technological
factors. Operational factors that affect fuel productivity include average
payload weight, empty backhaul, and maximum allowable size and weight. All of
these issues have become important, since the competitiveness of the trucking
industry not only depends on truck technology but also on operational
factors. Two factors - empty backhaul and maximum allowable size/weight are
1-1
-------
analyzed in Section 3. This analysis relies on TIUS data as well as on
information obtained from the American Trucking Association and other
organizations about the operational characteristics of heavy-duty trucks.
Truck technology is also a major force in improving truck fuel productivity.
The advent of stringent new emission standards has created the argument that
future technological benefits to increase fuel efficiency may counterbalance
the negative effects of emission standards and, thus, that an increase in fuel
productivity may not take place. A detailed analysis of all of the
technological improvements likely to occur over the next decade is provided in
Section 4. Data for this Section were developed from interviews with major
U.S. and European manufacturers of trucks and truck engines (European
manufacturers have recently acquired major U.S. truck manufacturers). Even in
the absence of any regulatory incentives, EEA's analysis shows that truck fuel
economy will continue to improve, at least at the historical rate
1-2
-------
2. HEAVY-DUTY TRUCK FLEET CHARACTERISTICS
2.1 INTRODUCTION
The term "heavy-duty" truck encompasses a wide range of weights and operating
characteristics of trucks used in on-highway operations. Trucks with a Gross
Vehicle Weight (GVW) higher than 8,500 lbs are considered by EPA as heavy-
duty. This rating gives the low end of the heavy-duty GVW spectrum, while the
maximum rating allowed for on-highway tractor-trailers is 80,000 lbs.
Industry, on the other hand, classifies the market differently. Until
recently, all trucks with a GVW rating of 10,000 lbs or less were regarded as
light-duty. However, this definition has changed to include trucks with a GVW
rating to 14,000 lbs, as premium versions of models rated above 10,000 lbs
have been introduced in the 10,000 to 14,000 GVW range. In industry terms,
this range is referred to as Class 3, and is a new market where few trucks
were sold, previously Trucks in Class 4 (14,001 to 16,000 lbs) and Class 5
(16,001 to 19,500 lbs) were also limited in sales until recently
Trucks in Class 6 (19,500 to 26,000 lbs) and Class 7 (26,000 to 33,000 lbs)
are typically referred to by industry as medium-heavy-duty trucks. Class 8
trucks include those trucks with GVW ratings between 33,000 and 80,000 lbs.
However, the lower weight range of Class 8 trucks shares many of the
characteristics of medium-heavy-duty trucks To assure that these common
vehicle characteristics are accounted for, EEA found it necessary to
disaggregate Class 8 into two sub-categories. Class 8A includes those trucks
with GVW ratings from 33,001 to 60,000 lbs, and Class 8B refers to those
trucks with GVW ratings from 60,001 to 80,000 lbs. Class 8A trucks are also
considered to be medium-heavy-duty, or 'super-mediums' in industry terms,
while trucks in Class 8B are referred to as heavy-heavy-duty
This section presents fleet characterization data for the medium-heavy-duty
and heavy-heavy-duty truck markets (i.e, Classes 6 to 8B). Such data analysis
is integral in understanding technological and policy effects on fuel
2-1
-------
efficiency. Technological innovations influence truck classes differently,
while policy may only affect a certain subset of the heavy-duty truck fleet.
It is, therefore, important to characterize the physical and operational
characteristics of each GW class independently. Section 2.2 describes data
used in this analysis. Section 2.3 describes physical and operational
characteristics of the heavy-duty truck fleet.
2.2 DATA USED IN THE ANALYSIS
This study uses data from the 1987 Truck Inventory and Use Survey (TIUS) The
TIUS is conducted every five years by the Bureau of Census, and is the only
publicly available survey providing data on the physical and operational
characteristics of the nation's truck population. It is based on a
probability sample of private and commercial trucks registered in each state
during 1987. However, vehicles which are owned by federal, state and local
governments are excluded from the sample universe, as well as ambulances,
buses, motor homes, and farm tractors.
The TIUS data base consists of 104,606 records and approximately 200 variables
that describe the characteristics of each truck in the sample universe. To
assure that the analysis recognizes differences between physical and
operational characteristics across the in-use truck fleet, EEA devised a data
clean-up and accuracy check routine. This routine classified each truck to
its corresponding industry weight class
The TIUS data clean-up process involved three steps. First, some trucks were
re-assigned to different GW classes A preliminary screening of the data
revealed that a number of trucks reported maximum loads well outside the
appropriate range of the TIUS assigned GW category A set of re-
classification rules were developed to re-assign these vehicles to the correct
category. These rules were based on available information regarding a truck's
make, fuel type, fuel economy, engine size, horsepower, number of cylinders,
and maximum loaded weight. A total of 7,212 trucks were assigned to different
GVW categories on the basis of these rules Second, engine and performance
parameters were compared for each truck to the expected range of values for
2-2
-------
the appropriate GVW category. Reported values for make, fuel type, fuel
economy, engine size, horsepower, number of cylinders, and maximum loaded
weight were compared to expected values (or ranges) for the truck's
appropriate GVW category. An 'exception' score was kept for each truck.
Every out-of-the-expected-range value added 1.0 to this score, If a truck did
not report a value for some parameter, the score was incremented by 0.5.
Third, trucks with 'exception' scores of 2.0 or greater were eliminated from
the data set. A total of 6,661 trucks were eliminated on this basis. Trucks
with 'exception' scores of 1.5 or less were accepted in the cleaned data set
There were 18,545 trucks (out of 97,945) in the cleaned data base that had
'exception' scores of 1.5 or less. The majority of these trucks simply
exhibited missing values for engine size, horsepower, or weight.
This clean-up procedure resulted with the following truck distribution by GVW
class
CLASS
GVW
Sample Size
1
6,000 or less
31,367
2
6,001 to 10,000
L0.895
3
10,000 to 14,000
3,334
4
14,001 to 16,000
1,768
5
16,001 to 19,500
1,934
6
19,501 to 26,000
8,927
7
26,001 to 33,000
5,251
8A.
33,001 to 60,000
12,415
8B
60,001 to 80,000
20,521
The small samples for Classes 4 and 5 verify low sales volumes in this market.
Although this market is technically recognized by industry to be part of the
heavy-duty truck market, the fact that only 5% of all heavy-duty trucks are
light-heavy-duty trucks allows for analytical emphasis on the heavier
populations. Therefore, EEA disregards Classes 4 and 5 from the analysis, as
well as light-duty vehicles (i.e , Classes 1, 2, and 3). Finally, 1,533
trucks were found to typically operate beyond 80,000 lbs. These trucks do not
posses consistent physical and operational characteristics, so they were also
disregarded from most of the analysis.
2-3
-------
2.3 CHARACTERISTICS OF HEAVY-DUTY TRUCKS
Class specific physical and operational characteristics are important in
determining the policy and technological options that best improve fuel
efficiency. The physical characteristics of a given truck include, among
other things, the truck's engine type (i.e., gasoline, diesel, LPG, or other),
engine size, and horsepower rating. Table 2-1 demonstrates the distribution
of trucks by engine type and GW class. Over 50X of trucks in Class 6 and
Class 7 are propelled by gasoline. In contrast, only 20.52 of trucks in Class
8A have gasoline engines, while all trucks in Class SB have diesel engines.
This suggests that diesel engine improvements, such as electronic fuel
injection timing control and improved intake and exhaust porting, will have
significant effects on the average fuel efficiency of Class 8A and Class 8B
trucks, but more modest effects on Class 6 or 7 trucks. However, sales data
in Table 2-2 shows that diesel penetration has increased markedly in these
classes. In 1980, 24.4% and 62,2* of new sales in Classes 6 and 7,
respectively, consisted of diesel powered trucks. By 1990, diesel sales
percentages had increased to 71.05% for Class 6 and 81.55% for Class 7, and a
shift away from Class 6 trucks to Class 7 and Class 8A trucks has taken place
Similarly, Class 8A is slowly being fully dieselized, while Class 8B has been
completely dieselized since the mid-1970's, as it exclusively consists of
line-haul trucks.
Trends in engine size and horsepower ratings can help to explain changes in
average fuel efficiencies across GVW classes. Figure 2-1 shows average engine
size (CID) for gasoline trucks by GVW class and vintage. For virtually all
classes, no significant changes have taken place in average CID. The small
dip in Class 8A trucks during model years 1982 to 1984 reflect the fact that
Navistar, the maker of the largest gasoline engines in the early 1980's,
exited that market. Similar data analysis for diesel trucks also showed no
significant engine size trends in any of the GVW classes.
Figure 2-2 shows average horsepower by GVW class and vintage for diesel
trucks There is a contention in industry that during the 1980s diesel engine
horsepower ratings had steadily increased, especially in the heavier truck
2-4
-------
TABLE 2-1
DISTRIBUTION OF TRUCKS BY
ENGINE TYPE AND GVW CLASS
GW Class
Class 6
Class 7
Class 8A
Class 8B
X Gasoline
68.8
52.9
20.5
0.0
X Diesel
30.4
46.4
79.3
100.0
X LPG
0.6
0.5
0 1
0.0
X Other1
0.2
0.2
0.1
0.0
1 Other includes those trucks for which engine type was unknown.
2-5
-------
TABLE 2-2
SALES AND DIESEL PENETRATIONS
BY CLASS
1980
1990
GW Class
Class 5
Class 6
Class 7
Class 8A
Class 8B
Buses
Sales
1,860
51,170
54,360
10,400
93,490
X Diesel
24.4
62.2
74.5
100.0
Sales
I,726
17,687
61,010
II,981
122,181
32,731
X Diesel
71 05
81. 55
99 70
100.0
78.2
2-6
-------
FIGURE 2-1
Average CfD By Vintage and GVW Class
Gasoline Trucks
CID
500 ——
1977 1978 1979 1980 19B1 1982 1983 1984 1985 1986 1937
Model Year
Class 6
Class 7
Class 8A
Model Year 1977 Includes 1977 and
Pre-1977 models
-------
FIGURE 2-2
Average Horsepower By Vintage and
GVW Class (Diesel Trucks)
Horsepower
400
300
200
100
1977
1978
1979 1980
1981 1982 1983 1984 1985 1986 1987
Model Year
Class 6 IHH Class 7 LUciass 8A
Class QB
Model Year 1977 Includes 1977 and
Pre-1977 models
-------
classes However, Figure 2-2 does not support this conjecture. No trends in
average horsepower ratings are apparent in any of the GVW classes. The reader
is cautioned that TIUS data on engine size and horsepower does not specify
each truck's actual engine size or horsepower rating. Rather, the survey asks
the responder to classify the truck into engine size and horsepower ranges
that are provided by TIUS. The data shown in Figures 2-1 and 2-2 describe the
average of the midpoints of the range for each vehicle in a given GVW
category.
Besides engine size and horsepower, other physical attributes impact a truck's
fuel efficiency. For example, fuel economy options like aerodynamic drag
reduction devices, engines with low RPM, high torque rise, turbocharger,
variable fan drives, radial tires, or axle/drive ratios that maximize fuel
economy have important effects on average fuel consumption rates across truck
classes. Table 2-3 presents the penetration rates of these fuel economy
options by GVW class/engine type combination. The penetration of aerodynamic
devices is more prevalent than any other fuel economy option. Newer trucks
are designed with aerodynamic features, and, with the exception of radial
tires, this technology is the most cost efficient to retrofit
However, in order to understand the future market penetration of technologies
that truck buyers can select as options, it is necessary to know the
operational characteristics of trucks by GVW class. While lubricant
improvements, weight reduction, accessory drive improvements, and transmission
improvements are applicable to all trucks, and are usually incorporated into
the standard truck, aerodynamic drag reduction devices, radial tires and speed
control devices are driver (or owner) selected options. For example, radial
tires are not purchased by consumers who operate their trucks in 'rough'
conditions because radial tires are more susceptible to sidewall damage than
bias ply tires Similarly, aerodynamic drag reduction devices are only useful
in trucks that have enclosed vans (dry vans) or trailers, and with tank
trucks On open trucks (such as flatbeds, cattle racks, and dump trucks) drag
reduction devices offer no useful fuel economy improvements TIUS data was
used to estimate the percent of trucks operated in rough and agricultural
2-9
-------
TABLE 2-3
PENETRATION RATES OF FUEL ECONOMY OPTIONS
BY GVW CLASS AND ENGINE TYPE
X With
GVW Class/Engine Aero
Type Devices
Class 6 Gasoline 1.0
Class 6 Diesel 6.3
Class 7 Gasoline 1 4
Class 7 Diesel 6.4
Class 8A Gasoline 0.9
Class 8A Diesel 9.8
Class 8B Diesel 22.5
X With
X With Fuel Variable X With
Efficient Fan Radial
Engines Drives Tires
0.1 0.0 0.1
2.8 0.5 1.3
0.8 0.1 0.3
3.0 1.1 1.5
0.2 0.0 0.0
5 6 1.4 2 7
15.9 4.3 7 3
X With
Fuel Max
Axle
Ratios
0.4
3.7
1 1
3 6
0.5
6.2
17 3
2-10
-------
applications and the percent of trucks that allow the use of drag reduction
devices. The estimate is available by GVW class and fuel type - gasoline in
Table 2-4 and diesel in Table 2-5. EEA has defined 'rough' operation as those
trucks used in construction, forestry, and mining. 'Regular' use trucks, such
as trucks used in wholesale or retail trade, mostly include enclosed vans and
tank trucks.
Area of operation also has a direct impact on a truck's fuel efficiency.
Long-haul trucks that mostly operate on interstate highways at constant speeds
are expected to be more fuel efficient than short-haul trucks that operate at
city cycles, all other things being equal. Table 2-6 characterizes each GVW
class/fuel type combination by the percent of trucks that can be characterized
as local, short-haul, and long-haul. As expected, the percent of Class 8B
trucks operating locally is substantially less than in any other GVW class.
Class 8B includes mostly line-haul vehicles that operate on interstate and
intra-state highways at near constant speeds.
The single most important operational factor influencing a truck's fuel
efficiency is it's average operating weight on a given trip. At any given
moment in time, the operating weight of a truck is defined as the empty weight
of the truck plus the weight of cargo being hauled. However, on any given
trip a typical truck will encounter some empty mileage (i e, when no cargo is
being hauled) and some loaded mileage (i.e., when cargo is being transported).
To estimate the average operating weight of a truck it is necessary to account
for both empty and loaded mileage TIUS variable PNOLOD describes the
approximate percentage of a truck's annual mileage during which no payload was
carried. Figure 2-3 shows average PNOLOD by GVW class/fuel type combination.
Average PNOLOD is surprisingly high in all GVW classes, with gasoline trucks
showing higher rates than diesel trucks. One would expect empty mileage to be
substantially lower in Class 8B, since it largely consists of line-haul
trucks. Line-haul trucks, and other commercial trucks, attempt to minimize
empty mileage because fuel productivity (i.e., ton-miles per gallon of fuel
consumed) is equal to zero when empty operation takes place The fact that
2-11
-------
TABLE 2-4
GASOLINE TRUCKS BY USE AND BODY STYLE
PERCENT BY GVW CLASS
Agricultural Rough Use Regular Use Total
Class 6
Non-Aero 93.9 93 7 61 3 80.7
Aero 6 1 6.3 38.7 19.3
Total* 42 5 17.1 40.4 100.0
Class 7
Non-Aero 94.5 94 2 65.1 82.6
Aero 5 5 5 8 34 9 17 4
Total* 41 1 18 7 40 2 100.0
Class 8A
Non-Aero 96.1 96 9 71 6 89 6
Aero 3.9 3 1 28 4 10 4
Total" 49 3 23.7 27 0 100 0
Horizontal Total reflects % of trucks in a class by type of operation.
Vertical Total reflects % of trucks in a class by body style.
2-12
-------
TABLE 2-5
DIESEL TRUCKS BY USE AND BODY STYLE
PERCENT BY GVW CLASS
Class 6
Non-Aero
Aero
Total*
Class 7
Non-Aero
Aero
Total*
Class 8A.
Non-Aero
Aero
Total*
Class 8B
Non-Aero
Aero
Total*
Agricultural
76. 6
23 4
10 4
71 5
28.5
10 2
79 8
20.2
10.4
72 2
27.8
9 3
Rough Use
88.8
11.2
14.7
93.8
6 2
16 6
94.8
5.2
28.4
92.7
7.3
18.6
Regular Use
43.4
56.6
74.9
46.0
54.0
73.2
44.8
55.2
61.2
37.7
62 3
72 1
Total *
53 6
46.4
100 0
56.6
43.4
100.0
62.7
37.3
100.0
51 2
48 8
100. 0
Horizontal Total reflects % of trucks in a class by type of operation.
Vertical Total reflects % of trucks in a class by body style
2-13
-------
TABLE 2-6
PERCENT OF TRUCKS BY AREA OF OPERATION
BY GVW CLASS AND FUEL TYPE
Area of Operation1.
GVW Class/Engine Tvpe
X Local
X Short-Haul
X Lone-Haul
Class
6
Gasoline
84.4
13.8
1.8
Class
6
Diesel
66.1
27.8
6.1
Class
7
Gasoline
86.2
12.5
1.3
Class
7
Diesel
66.4
29.0
4.7
Class
8A
Gasoline
86.3
11 9
1.8
Class
8A
Diesel
63.4
24 4
12 1
Class
8B
Diesel
25.4
32 9
41.7
1 Local if greatest percentage of annual miles were accrued within a 50
mile radius of home-base. Short-haul if greatest percentage of miles were
accrued between 50 to 200 mile radius of home-base. Long-haul if greatest
percent of miles were accrued beyond a 200 mile radius of home-base.
2-14
-------
FIGURE 2-3
Average Percentage of Annual Mileage
When No Load Was Carried, By GVW Class
Average PNOLOD
50%
-------
PNOLOD rates are so high implies that carriers can benefit greatly by reducing
empty mileage.
Having characterized the average annual mileage that trucks operate without a
load, average operating weight can be estimated using TIUS variables EMWGHT
and AVWGHT. The EMWGHT variable describes a given truck's empty weight, while
AVWGTH describes the empty weight plus weight of cargo of a truck when
carrying a typical payload. Therefore, the average operating weight of a
truck can be defined by the following weighted sum:
[AVWGHT*(1-PNOLOD) + EMWGHT*{PNLOD)].
EEA defines this estimate of average operating weight as equivalent weight
(EQUIVWT) . The equivalent weight of a vehicle will have a considerable impact
on the vehicle's average fuel consumption rate (MPG). Figure 2-4 shows
average EQUIVWT by GVW class/engine type combination.
Model year trends in gasoline fuel economies (MPGs) are plotted by class in
Figures 2-5 through 2-7 for Classes 6, 7 and 8A The error bars give the high
and low ends of the standard deviations of the means An increasing trend in
fuel economy is apparent since 1984 for both Class 6 and Class 7 gasoline
trucks, while Class 8A gasoline trucks show no trend in MPG. Diesel fuel
economy trends are plotted in Figures 2-8 through 2-11. In each GVW class,
diesel HPGs have consistently increased since model year 1977.
2-15
-------
FIGURE 2-4
Average Operating Equivalent Weight
By GVW Class
Average EQUIVWT (x1000 lbs.)
Class 6
Class 7 Class 8A
GVW Class
Class 8B
Gasoline Uli Diesel
-------
FIGURE 2-5
Average MPG By Vintage
GVW Class 6 Gasoline Trucks
1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987
Model Year
Model Year 1977 Includes 1977 and
pre-1977 models
-------
FIGURE 2-6
Average MPG By Vintage
6VW Glass 7 Gasoline Trucks
Model Year
Model Year 1977 Includes 1977 and
pfe-tS77 models
-------
FIGURE 2-7
Average MPG By Vintage
GVW Class 8A Gasoline Trucks
6.50
6.00 -
5.50
to
o
M
P
G
5.00 -
4.50
4.00
1977 1978 1979
1980 1981 1982 1983
Model Year
1984 1985 1986 1987
Model Year 1977 Includes 1977 and
pre-1977 models
-------
FIGURE 2-8
Average MPG By Vintage
GVW Class 6 Diesel Trucks
Model Year
Model Year 1977 Includes 197 7 and
pre-1977 models.
-------
FIGURE 2-9
Average MPG By Vintage
GVW Class 7 Diesel Trucks
Model Year
Model Year 1977 includes 1977 and
pre -19 7 7 model s
-------
FIGURE 2-10
Average MPG By Vintage
GVW Class 8A Diesel Trucks
Model Year
Model Year 1977 includes 1977 and
pre-1977 models
-------
FIGURE 2-11
Average MPG By Vintage
GVW Glass 8B Diesel Trucks
Model Year
Model Year 1977 Includes 1977 and
pre-19 77 models
-------
3. OPERATIONAL FACTORS TO
IMPROVE FUEL EFFICIENCY
3.1 INTRODUCTION
Transportation facilitates the exchange of goods and services. For a given
level of freight transportation demand, many combinations of the capital stock
and modal share (truck, ship, or airplane) can be used to satisfy that level
of demand. Each vehicle type within a mode has an average energy efficiency
(BTUs per vehicle mile), a characteristic load factor (freight ton-mile per
vehicle mile), and intensity of use (vehicle miles traveled per year).
Therefore , given a level of demand, energy conservation in transporting
freight by truck can be accomplished by reducing the energy input through
operational and technological improvements, and/or by increasing the number of
freight ton-miles traveled per unit of energy that is consumed (i.e.,
increasing fuel productivity) This section of the report provides a
quantitative description of the factors that influence fuel productivity
Fuel productivity in the movement of freight by truck largely depends on the
type of truck that is being used. Each vehicle type has physical and
operational characteristics that are unique. Typical vehicle types are shown
in Figure 3-1. Straight trucks are vehicles with the cargo body and tractor
mounted on the same chassis and usually consist of 2-, 3-, or 4-axle
configurations. Two-axle straight trucks are most often used in urban areas
where maneuverability ls important and operate between 19,500 lbs and 33,000
lbs gross vehicle weight (GW). Three and 4-axle straight trucks are
principally used in construction or other 'rough-duty' uses Combination
trucks are vehicles that have a power unit (tractor) that is separate from the
trailer(s), and are generally used for interstate freight movement. Although
combination trucks may have 3 to 9 axles, the most common types are 5-axle
combinations with one 48 ft long semi-trailer and 5-axle double trailer
combinations with two 28 ft trailers (i e , 5-axle Twin 28s). Five-, 6-axle
tractor semitrailers, and 5-axle Twin 28s commonly operate between 60,000 lbs
3-1
-------
Figure 3-1
TRUCK TYPES
STRAIGHT TRUCK
j— 25' ¦ 40' — |
4-AXLE tractop semitrailer
38' • 40'
5-AXLE TRACTOR f UkTBEQ TRACER
J 38 «' 1
fci «k
3-AXJ.E TRACTOR SEMITRAILER
J—-24' ¦ 28' 1
5-AXLE TRACTOR SEMITRAILER
I 40' - 48' 1
5-AXIE TRACTOR TANK TRACER
35' ¦ 40"
TWIN TRAILER OR "DOUBLES"
f -28' 1 I 23'
ROCKY MOUNTAIN DOUBLES
(operated o*it* hi eettam ataWi
45 - 48' 1 1 —23'
TURNPIKE DOUBLES
(operated only in certain statas)
LENGTHS SHOWN ARE TYPICAL; SHORTER OR LONGER LENGTHS
ARE POSSIBLE OEPEHOWG ON CASTERS' NEECS ANO STATE LAWS.
Source: ATA, American Trucking Trends, 1989 3-2
-------
and 80,000 Its GVW. Seven-axle, 8-axle, and 9-axle Twin 28s generally operate
beyond 80,000 lbs GVW (the current legal maximum) under special permit in
accordance with applicable state laws. A double trailer combination with one
or both trailers longer than 28 ft is defined as a Longer Combination Vehicle
(LCV). The definition of LCVs is not consistent within the trucking industry.
Some define 7-axle, 8-axle, and 9-axle Twin 28s as LCVs, since these
configurations hardly ever operate below 80,001 lbs GW Others, including
EEA, consider only those double trailer combinations with one or both trailers
longer than 28 ft operating at weights that exceed 80,000 lbs (under special
permit) as LCVs In this context, the most common types of LCVs are Rocky
Mountain Doubles and Turnpike Doubles. Rocky Mountain Doubles are 7-axle
double trailers commonly having one trailer that is 28 ft long and another
that is 4-8 ft long. Turnpike Doubles are 9-axle double trailers whose
trailers are usually both 48 ft long. Triple trailer combinations (i.e , a
tractor unit with three 28 ft trailers) are also common in some states,
operating under special permit. However, because of their specialized
applications and physical characteristics, an analysis of triples is beyond
the scope of this report.
One way to change fuel productivity is to increase the payload capacity of a
truck Payload capacity is determined by the difference between a vehicle's
empty weight and a vehicle's maximum gross vehicle weight (GVW). A vehicle's
maximum practical GVW is defined by a carrier's needs and by state and federal
regulations on truck size and weight. Truck weight limits constrain GVW by
restricting a vehicle's maximum operating weight, while truck size limits
constrain GVW by restricting a vehicle's volumetric carrying capacity
Liberalizing weight limits to allow trucks to operate at higher GVWs directly
influences fuel productivity. A vehicle's empty weight does not increase m
step with GW, so a more liberal weight limit increases the payload capacity
of the vehicle. Increasing payload capacity compromises fuel economy (as
defined by miles traveled-per-gallon of fuel that is consumed, or MPG) since
the added weight requires more energy input to propel the vehicle into motion
However, in the context of fuel productivity, the potential payload capacity
3-3
-------
benefit is expected to be greater than the loss in fuel economy. Fuel economy
losses will be relatively small because aerodynamic forces do not closely
scale with added vehicle weight and because heavier vehicles have larger
engines that can be more fuel efficient. On the other hand, changes in weight
limits affect not only the operating weights of today's trucks, but also the
types of equipment that motor carriers will operate in the future. If weight
limits are liberalized, some motor carriers will switch to vehicle
configuracions that have higher empty weights. The difference in empty
weights between these new configurations and those used previously are small
relative to the difference in payload capacities. Seven-, eight- and nine-
axle double-trailer trucks with trailing units that are longer than 28 ft -
defined as Longer Combination Vehicles (LCVs) that are currently not allowed
to operate in most states - have empty weights between 30,000 and 40,000 lbs.
These vehicles operate at weights of well over 100,000 lbs, often carrying
payloads of over 70,000 lbs Conventional five-axle tractor semi-trailer
trucks, on which most freight is currently transported, are limited by law to
a maximum GVW of 80,000 lbs These vehicles usually have empty weights of
25,000 lbs and often carry payloads that weigh below 65,000 lbs Even if
changes in the weight limits induce a shift toward vehicles with higher empty
weights, the fuel productivity gain of the added payload capacity can be
significant Increasing payload capacity will also increase the cost
productivity of the vehicle's driver At a given driver cost more payload can
be carried.
Policies that liberalize weight limits are expected to increase fuel
productivity and operational productivity in the movement of freight by truck.
Weight limits are imposed at both the state and federal level However,
current federal regulations supersede many state limits, at least as they
apply to the interstate highway system on which a significant portion of fuel
is consumed.1 Due to this, and because state regulations for non-interstate
roads vary from one state to the other, this report quantifies the effect on
fuel productivity of a change in the federal weight limit. Throughout the
analysis, size limits are assumed to remain unchanged.
3-4
-------
Another way to increase fuel productivity is to decrease the mileage when a
truck is operating empty. Empty operation takes place when a truck carries a
load in one direction but travels empty in the other. During empty mileage
fuel productivity (ton-miles per gallon) is equal to zero. Therefore, to the
extent that carriers can minimize empty backhaul, fuel productivity gains can
be realized. This section also quantifies the influence of empty mileage on
fuel productivity and investigates what can be done from a policy perspective
to minimize empty mileage.
The organization of this section is as follows: Section 3.2 outlines current
federal weight regulatory constraints in the maximization of fuel productivity
and presents the policy change that is investigated in this report, Section
3.3 describes the analytical methodology that was used to develop quantitative
estimates of fuel productivity, Section 3.4 quantifies the effect of empty
mileage on fuel productivity, and Section 3.5 quantifies the effect of a
change in the federal weight limit and the effect of LCV operations.
3.2 FEDERAL REGULATIONS ON WEIGHT
Truck weight regulations, at the federal and state level, are motivated by
concern for protecting pavements and bridges from the effects of heavy loads.
For a given truck, increasing the number of freight ton-miles traveled per
unit of fuel that is consumed implies a heavier operation weight, since the
payload weight increases and empty weight remains constant or does not
proportionately increase with payload The introduction of heavier trucks has
a direct effect on pavement wear and the safety margin for bridges, thus
increasing the cost to a highway agency of maintaining its road network
During the 1970s one of the major issues of concern to the trucking industry
was uniform truck weight regulations. Although previous federal regulation
had increased the allowable weight of trucks on the interstate system, it had
not mandated all states to comply. As a result, six states in the Mississippi
Valley and Montana retained lower limits, and truckers in interstate commerce
passing through these states were forced to operate at the lower limits or
operate illegally 2 In response to the need for uniform regulations the
3-5
-------
Surface Transportation Assistance Act of 1982 (STAA of 1982) was passed into
law. This act resolved the uniform weight issue and expanded the role of
federal regulations to other federally-aided primary roads. Regulations
brought about by the STAA of 1982 are valid today.
Under the STAA of 1982, Congress requires all states to allow on the
interstate highway system and primary roads receiving federal-aid the
following weights: a maximum load of 20,000 lbs on single-axles, a maximum
load of 34,000 lbs on tandem-axles (a tandem-axle is a pair of closely spaced
axles), and an overall gross weight on a group of two or more consecutive
axles produced by application of the Federal Bridge Formula B, provided that
such overall gross weight does not exceed 80,000 lbs. Bridge Formula B is
specified as follows:
W - 50O (LN/(N - 1) + 12N + 36]
where, W =¦ maximum weight in pounds carried on any group of two or more
axles,
L — the distance in feet between the extremes of any two or more
consecutive axles,
N - the number of axles on the vehicle.
The intent of Bridge Formula B is to limit axle weights so that trucks do not
over-stress bridges on interstates and the federal-aid highway network. Under
the Bridge Formula, gross weights are allowed to increase as the number of
axles and the length between axle groups increase. However, the overall gross
weight of a vehicle cannot exceed 80,000 lbs, except for those vehicles
carrying loads which cannot be easily divided and which have been issued
special permits in accordance with applicable state laws.
State laws also determine the length of the trailing units of tractor trailer
combinations. The STAA of 1982 prohibited states from limiting the length of
the semitrailer of a tractor-trailer combination to less than 48 ft or each
trailer of a combination with two trailers to less than 28 ft on the
3-6
-------
interstate highway system.* The federal regulation on length imposes a
minimum criterion on a state's maximum length regulation. States must permit
these lengths, but can set limits on semitrailer and double trailer lengths
beyond 48 ft and 28 ft, respectively (see Table 3-1). These state maximum
length limits are important because in the absence of the 80,000 lbs GVW cap,
Bridge Formula B would allow higher GVW limits depending on the number of
axles and the distance between the extremes of any two or more consecutive
axles, assuming that axle-weight limits are met.
Permissible gross weights under Bridge Formula B are shown in Table 3-2. This
table demonstrates that the maximum allowable GVW for a 2-axle truck is 40,000
lbs providing that the distance between the axles is 10 ft. For a 3-axle
truck the GVW limit under this formula is 60,000 lbs providing that the
distance between the extreme axles is 32 ft. A 4-axle truck combination can
reach an overall weight of 80,000 lbs only when the distance between the
vehicle's extreme axles is not less than 57 ft. However, even if distances
are satisfied, these configurations will not reach the Bridge Formula B GVW
maximums because axle-load limits will rarely be met This means that the
maximum practical GVW for these truck configurations is actually determined by
the axle-weight limits that are discussed above. Under these axle-weight
limits, a 2-axle truck with a 12,000 lbs load on the steering axle and a
20,000 lbs load on the rear single-axle has a practical GVW limit of 32,000
lbs. A 3 - axle truck with two single-axles and 12,000 lbs on the steering axle
has a GVW limit of 52,000 lbs, while a 4-axle truck with one single-axle, one
tandem-axle, and 12,000 lbs on the steering axle is limited to 66,000 lbs GVW.
Likewise, a typical 5-axle tractor-semitrailer with two tandem-axles and
12,000 lbs on the steering axle is limited to a practical GVW of 80,000 lbs by
the axle-weight limits, although Bridge Formula B allows for higher GVWs as
the distance between extreme axles increases.
STAA of 1982 also required states to allow use of combinations consisting
of a tractor and two trailing units on the interstates and the network of primary
roads.3
3-7
-------
TABLE 3-1
STATE SIZE LIMITS
state
WQKT
worn
LENSTM (PT-VI)
Tractor-Samtrailar Combnatma
Twm ComOnatnna
1
Sarmtiaiier
Semitrailer
Semitrailer
Overall
or Trailet on
Twin
m Feeti
Truck
an Intaraiate
Length off
Cambtiraiion
Iniefaiate 4
Combination
(Single
i National
Nttionai
L«nSlh on
National
Ltngth an
Incnaa
in tnen«»
Unil)
Network'
Network'
Otner Roada
Network
Other Roam
Trailer
AiaDama
13-6
102
40-0
53-6
53
NR
28-6
NR
SO
Alaska
14-0
102
40-0
48
45
70
90
75
75
Arizona
14-0
102
40-0
574
53/NR
65
28-6
NR
NR
Arkansas
1^9
102
40-0
53-8
5>fl
75
28
86
NS
California
14-0
102
400
48/53
MR
86
284
75
65
Colorado
1M
102
4(H)
37-4
57-4
70
28-6
70
70
Connecticut
13-6
102
6O-0
48
48
NR
28
NP
60
Delaware
13-6
102
40-0
53
NR
60
29
NP
SO
Oist ot Colum&,a
13-6
102
40-0
48
NP
55
28
NP
55
Florida
13-6
102
4O-0
48/S7-8
48
NR
28
NP
Georgia
13-6
102
eoo
53
48/53
ao/er-e
2B
NP
60
Hawaii
13-6
106
4(H)
48
45
ao
66
85
65
loarto
14-0
102
4(H)
48'52
48
NR
81
61
75
Illinois
13-6
102
42-0
53
53
35
28-6
63
60
Indiana
13-6
102
36-0
53
53
NR
26-6
NR
80
Iowa
13-6
102
40-0
53
NR
30
28-6
60
65
Kansas
14-0
102
42-6
594
5»«
NR
284
NR
65
Kentucky
13-6
102
(M
53
NR
57-9
28
NP
57-9
Louisiana
13-6
102
400
598
NR
85
30
NR
65
Maine
13-6
102
45-0
48
48
65
26-6
NP
65
Maryland
13-6
102
4(H)
48
48
MR
28
NP
55
Masaacftusafl*
13-6
102
40-0
48
45/48
SO
28-6
NP
60
Michigan
13-6
102
4O-0
53
50
58
5tJ
59
Minnesota
(3-9
102
400
53
53
95
28-8
NP
65
Mississiopi
13-6
102
4O-0
53
53
NR
30
NR
NR |
Missouri
13-6
102
4O0
53
NR
60
26
65
65
Montana
14-0
102
4O-0
53
53
NR
26-6
75
75
Nebraska
14-6
102'
4OC
33
- »
W» "
~93
as
85 ""
Nevada
14-0
102
4(H)
53/KR
53/NR
70
2 8-6/Nfi
70
70
New Hsmp3"ire
134
102
400
48
40
NR
28
NP
65
New Jersey
13-6
102
35-0
48
48
NR
28
NP
62
New Mexico
14-0
102
404
57-6
57-6
85
2B-6
65
65
New York
13-6
102
4O0
53
48
05
26-6
NP
60
Norm Carol i"*
13-6
102
400
53
NR
60
28
NP
60
North Dakota
13-6
102
soc
53
53
75 /8B
53
75 /88
75
Ohio
13-4
102
400
53
53
NR
26-6
NR
65
Ok la noma
13-6
102
45-0
59-6
53
NR
29
NR
70
Oregon
14-0
102
4(H)
53
NR
SO
68
60
75
Pennsylvania
13-6
102
40-0
53
NR
90
28-6
NP
60
Rhooe Island
13-6
102
40-0
48-8
48-6
NR
26-6
NP
NS
Soutn Carakfi*
13-6
102
400
53
48
NR
28-8
NP
NR
South OakC*
14-0
102
46-0
53
53
NR
81-8
81-6
80
Tennessee
13-6
102
40-0
50
50
NR
28-6
NP
35
Texas
13-6
102
4S-0
39
59
NR
284
NR
SS
Utah
14-0
102
45-0
48 /53
48
Nfi
81
61 /NR
65
Vermont
13-6
102
eoo
48
45/NR
8fl~
28
NP
80
Virginia
13-6
102
400
53
NR
60'
28-8
NP
BO
Waafungtofl
14-0
102
400
48
48
SO
80
75
West Virgin"4
13-6
102
4O0
53
NR
60
28
NP
60
Wisconsin
13-6
102
40-0
53
NR
SO
28-8
NP
60
Wyoming
14-0
102
600
60
60
NR
80
BO
85
Source: ATA, Size and Weight Limits, July 1990.
3-8
-------
Table 3-2
INTERSTATE/NATIONAL NETWORK ALLOWABLE GROSS WEIGHT
(FEDERAL BRIDGE FORMULA)
GROSS WEIGHT LAW
Slates hove adopted Ihe Federal Budge Formula lor travel on Ihe Interstate and other public
highways alitor by formula (Formula 8) or by ehail (Table B). wilh the exception of Ihe stales
found at Table A Variations may occur due lo rounding language adopted or not adopted by
Ihe respective stale Table B appears ss provided by Ihe Federal Highway Administration
FORMULA B W= 500(LN/N I h 12N I 36)
W maximum weight in pounds carried on any group of two or more amies computed to
nearest 500 pounds
L = distance in feet between Ihe extremes of any group ol two or more consecutive axles
N = number of aides in group under consideration
TABLE B (In 1,000 lbs.)
Dtalanc* In feel
Mai
Imum load In lOOO
S>9
Distance In feet
Mailmam load to lOOO
*0
between the ea-
carried on any oroop of 2
between tfte ei-
carried on any troop of 2
Uiwm of wy ffiouo
Of
tort cofteecotHe ealee
tiemea ot any proup
Of
note cona
ecuttve ealee
of 2 ot more con-
eecuttve ealee
or 1 or more con-
secutive ealet
7
3
4
8
6
r
a
•
7
3
4
S
e
T
•
•
ealee
•altt
ealet
6lt»l
9 ale*
ealee
•ale*
ealee
ealee
•ale*
ealee
ealee
aatee
oateo
oaleo
4
34 0
46
71 6
76 ft
09 9
67 O
010
000
9
34 0
47
73ft
775
010
6/0
OJO
000
e
34 O
46
74 O
760
030
660
030
90 O
r
34 O
49
74 ft
76 8
03 9
006
040
99 0
9 Md teei
340
340
60
76 ft
79 0
04 0
090
049
IOOO
(tit* 0
300
410
•
39 0
419
IO
40 0
43 S
61
B1
76 0
76 ft
600
609
04 9
mo
000
000
000
000
too 9
•Of 0
it
44 0
53
71ft
010
000
070
000
101 0
i*
45 0
600
64
76 0
01 9
000
Of 0
07 0
•019
13
46*
606
55
76 ft
619
07 O
910
07 0
•03 0
14
466
61 6
IS
47 0
610
56
67
79ft
600
63 0
639
07 0
000
010
030
000
000
1039
•040
ie
48 0
616
560
56
640
00 O
940
09 O
•04 9
tt
40 ft
636
666
59
690
006
040
000
109 0
ie
40ft
640
600
60
66 9
0DO
960
too 6
109 9
it
SO 0
640
600
1009
UJ 90
61 O
666
60 6
660
61
960
000
000
•010
61
609
• 1 0
900
tots
•0/ 0
vjd i*
St ft
BOO
61 0
666
63
67 9
010
000
•oto
•0/9
it
SIS
60S
61 6
•7 0
64
000
010
0/0
tns
9000
>3
S30
67 6
616
660
06
669
030
000
100 0
#009
64 0
660
630
666
74 O
»
64 6
MS
63 6
66 O
74 ft
66
67
60 O
900
036
040
000
mo
•000
•04 0
•090
19
55 5
BO S
64 0
90S
760
66
909
06 0
000
9000
if
MO
6O0
66 0
700
76 6
60
91 0
090
100 0
MOO
M
SI 0
606
66 6
710
76 6
610
70
91 9
96 O
101 o
•OOO
19
SI 6
61 S
660
716
It O
62 5
30
66 S
610
66 6
720
11 6
630
71
71
979
030
060
09 0
10# 0
fotro
moo
10/0
31
se o
61 6
67 6
716
76 O
63 9
73
939
00O
rod 0
tor 6
31
60 O
63 6
66 O
730
76 6
64 5
900
74
04 0
006
toso
roo 0
33
64 O
66 6
740
79 0
65 0
009
75
990
60 O
109 0
IOOO
34
64 6
60 0
74ft
600
69 5
Of 0
3ft
65 5
roo
ISO
0O9
66 0
91 9
76
77
96 9
660
006
•OOO
104 0
•06 0
3*
66 0'
70 5
766
010
66 9
910
76
666
tOtO
1000
37
66 6"
710
760
0t 6
07 0
93 O
79
97 9
101 0
106 0
3*
67 6*
71 6
77 O
030
67 9
93 9
60
96 O
10? O
•06 6
3*
660
71 6
77 ft
010
665
94 0
40
666
730
760
636
090
94 9
61
01
90 9
990
103 6
103 0
tor a
•00 0
41
606
73 5
765
040
609
990
63
100 0
104 0
•00 0
41
70 0
74 0
r»o
046
900
99 5
64
104 9
109 0
roo 0
43
70S
76 0
90 O
65 O
90S
900
6S
44
7 1 S
75 6
905
03 5
91 0
96 5
100 0
roe o
45
710
ye o
9*0
<160
9* 5
91 5
66
67
* MOQO«Mrb«ca*i
ar igh«arl
•
f
I
i
e
91
l»OI| tk«h^w0 i*etfe«
•* weigh' •• •©• • redraw*
h»i ol ' rmn«l( fl b«i ralhet >• •nviiafpraiilioe by 'he l*4«ra< bo<
9?
ha* moI mc««i
been adoplad bv nder«Wai iiatat
Source: ATA, Summary of Size and Weight Limits, 1990.
-------
On the other hand, 6-axle tractor-semitrailers and Twin 28s are limited by the
80,000 lbs GVW cap, since without the cap higher GVWs would be allowed by-
Bridge Formula B. A typical 6-axle tractor-semitrailer with a 48 ft trailer,
a kingpin-to-rear-axle distance of 41 ft, a kingpin setting of 1 ft behind the
second axle (giving a distance between extreme non-steering axles of 42 ft),
and 12,000 lbs on the steering axle is allowed to 'gross-out' at 86,000 lbs
under Bridge Formula B. With 3 ft between trailers and assumptions similar to
those previously discussed for axle spacing, the Bridge Formula B GVW limits
for Twin 28s range from 91,500 lbs for 5-axle Twin 28s to 110,000 lbs for 9-
axle Twin 28s - with 7-axle and 8-axle Twin 28s 'grossing-out' at 99,500 lbs
and 104,500 lbs, respectively.4 For these tractor-semitrailer and Twin 28
configurations axle-weight limits are typically satisfied so that maximum
practical GVWs are determined by the 80,000 lbs GVW cap. In practice,
however, 7-axle, 8-axle, and 9-axle Twin 28s seldom adhere to the 80,000 lbs
GVW cap and operate, under special permit, at higher weights that satisfy
Bridge Formula B and axle-weight limits.
LCVs also operate beyond 80,000 lbs under special permit Where not
restricted by a state's length limits, LCVs comply with GVW limits that are
set by Bridge Formula B, axle-weight limits, and additional state specific
requirements concerning horsepower, braking, linkage, driver training, weather
restrictions, and route designation according to the type of LCV.5 Under
Bridge Formula B, a Rocky Mountain Double with 12,000 lbs on the steering
axle, a 48 ft lead trailer, a 28 ft rear trailer, and 3 ft between trailers,
is allowed to 'gross-out' at 108,500 lbs - assuming that the distance between
extreme non-steering axles is at least 74 ft. Turnpike Doubles that have two
48 ft trailers, with 94 ft between extreme non-steering axles and 12,000 lbs
on the steering axle, are allowed to operate at roughly 131,700 lbs under
Bridge Formula B.
The flexibility to choose vehicles and payloads that meet their needs
effectively and at the lowest cost is of great economic value to truck
companies and individual truckers that transport freight to and from markets.
For vehicle configurations that have 5 or more axles the maximum practical
3-10
-------
payload weight is the difference between the vehicle's empty weight and 80,000
lbs (maximum legal operating weight), This GVW limit constrains the
maximization of fuel productivity for many carriers. At this, or any other,
weight limit there is a freight density at which weight capacities are fully
utilized. This is referred to as the optimum density. For finely divisible
commodities, such as liquids, it is possible to load a vehicle to its optimum
capacity. But for many other products, the 80,000 lbs GVW cap results in the
inability to use some portion of the vehicle's potential capacity by
restricting the weight of the items being shipped. To the extent that freight
hauled is more than the optimum density, weighing-out situations occur. The
80,000 lbs capacity is reached before the alternative capacity can be
realized. Such unused capacity is an opportunity cost that motor carriers
must incur. This cost can be metered by losses in fuel productivity.
In this study, EEA investigates the fuel productivity effect of eliminating
the 80,000 lbs GVW cap. Under this scenario, GVW would be controlled by
weight limits set by the application of Bridge Formula B and by current axle-
weight limits. It is assumed that no changes in state length limits take
place. This policy would have no effect on those configurations that are
limited in GVW by current axle-weight limits, such as, straight trucks, 4-axle
tractor-combinations, and 5-axle tractor-combinations. Of course, other
policy options are available regarding weight regulations (and have been
studied extensively), such as redefining the bridge formula or changing axle-
weight limits However, such proposals are more difficult to quantify in a
systematic fashion and are beyond the scope of this study.
When both conventional tractor-semitrailers, 5-axle Twin 28s, and 7- to 9-axle
Twin 28s are limited to 80,000 lbs, the 7- to 9-axle Twin 28s are much less
productive because of their higher empty weights. Weight-limited carriers are
better off operating 5-axle tractor-semitrailers since more payload can be
carried, while size-limited carriers are better off operating 5-axle Twin 28s
that are lighter and more fuel efficient. Under an uncapped Bridge Formula B
scenario, 7-axle, 8-axle, or 9-axle Twin 28 operators would not need special
permit to operate beyond 80,000 lbs GVW A policy that eliminates the 80,000
3-11
-------
lbs cap is, therefore, expected to induce shifts from weight-limited 5-, 6-
axle tractor-semitrailers and 5-axle Twin 28s to these heavier Twin 28s.
Bridge Formula B GVW limits allow 7- to 9-axle Twin 28s to carry more payload
and, thus, be more productive. Such a policy would not, however, affect the
nationwide operation of LCVs. Even without the 80,000 lbs cap, operation of
these vehicles would be restricted by most states' length limits. Given that
this analysis assumes no changes in state size limits, the fuel productivity
benefits of LCVs are recognized by comparing fuel productivity between LCVs
and conventional configurations operating under the uncapped Bridge Formula B
scenario (i.e., 5-, 6-axle tractor-semitrailers and Twin 28s).
As mentioned in the introduction to this section, EEA also analyzes the
effects of empty mileage on fuel productivity. Unlike the elimination of the
80,000 lbs GVW cap, empty mileage affects all freight trucks.
3.3 ANALYTICAL METHODOLOGY
The fuel productivity analysis in this study requires only those trucks used
for the movement of freight. As a result, GVW Classes 1, 2, 3, 4, and 5
(defined in Section 2) were not included in the analysis since vehicles in
Classes 1, 2, and 3 are not predominantly used for this purpose and sales in
Classes 4- and 5 are very low. The remaining classes reflect those trucks
which are recognized by the trucking industry as predominantly commercial
vehicles.
The analysis also excluded from the "active" data base those vehicles that are
propelled by LPG or other fuels. As was shown in Section 2, on average less
than half of a percent of the trucks in each GVW class have LPG or other
engine types Disregarding such trucks was deemed appropriate since no
statistically significant results could be drawn from analyzing them. As a
result, only gasoline and diesel powered trucks are included in the fuel
productivity analysis Classes 6, 7, and 8A consist of both gasoline and
diesel trucks, while Class 8B consists of only diesel trucks.
3-12
-------
Finally, Chose trucks whose greatest percentage of mileage was accrued off-
road were also deleted from the "active" data base. Such vehicles do not
operate on the interstate highway system and have physical and operational
characteristics that are not consistent with commercial trucks. EEA also
investigated the statistical differences between trucks that are predominantly
used for agricultural services, rough use services (i e., construction,
forestry and mining), and regular commercial services (such as, wholesale
trade, retail trade, manufacturing, refining, or processing activities, etc).
The statistical differences were not significant and a distinction in the
analytical procedure was deemed inappropriate.
Fuel productivity is usually defined as ton-miles per gallon, or the number of
freight ton-miles traveled per unit of fuel that is consumed. Therefore, to
determine the effect of lifting the 80,000 lbs GVW cap and the effect of empty
mileage on fuel productivity, it was necessary to devise models that explain
the functional relationship between the rate of fuel consumption (as defined
by GPM, or gallons of fuel consumed per miles traveled) and the variables that
influence this rate These variables should theoretically represent actual
operational and physical characteristics of a given vehicle, or vehicle class,
so that when changes in these variables take place the rate of fuel
consumption for that vehicle will adjust accordingly
Table 3-3 defines relevant TIUS variables used in the analysis. From these
variables fuel productivity can be defined as (AVWGHT - EMWGHT) * WPG. The
difference between AVWGHT and EMWGHT is the typical payload weight of a
vehicle - since AVWGHT is defined as EMWGHT plus weight of cargo This
difference multiplied by MPG equals, by definition, payload-miles (or ton-
miles) traveled per gallon of fuel consumed. However, such an estimate of
fuel productivity does not take into account the effect of empty mileage on
fuel productivity or the effects of a vehicle's physical and operational
characteristics on MPG [such as, the vehicle's engine size (CID) and
horsepower (HRSPWR), the vehicle's age (MODELYR) and area of operation
(AREAOP) , or the presence of aerodynamic devices (AERODN) or a fuel economy
engine (ECOENG)]
3-13
-------
TABLE 3-3
DEFINITIONS OF TIUS VARIABLES
USED IN THE ANALYSIS
VARIABLE
(TIPS)
DEFIHirrCB
MFC
MODEL YR
EMWGHT
AVWGHT
ANNMIL
PNOLOD
AEEAOP
PCAHSZ
PCARWT
CID
HRSPHR
AERODM
ECOENG
The number of miles-par-gallon that the vehicle averaged during 1987
The model year of the vehicle, ranging from 1987 to 1976 Model year 1976 includes all
pre-1976 model years as well
The empty, or tare, weight of the vehicle expressed in tons
The average weight (empty weight plus weight of cargo), expressed m tons, of a vehicle
when carrying a typical payload
The number of miles that the vehicle was driven during 1987
The approximate percentage of the vehicle's annual mileage that that no payload, or
cargo, was carried
The area of operation of the vehicle AREAO? will be take on the following numbers if
the particular condition is met
AREA0P»1 if the vehicle's greatest percentage of miles traveled were off-road
AREAOP^ if the vehicle's greatest percentage of miles traveled were within a 50
mile radius of the vehicle's home base
AREA0P=3 if the vehicle's greatest percentage of miles traveled were withm a 50
to ZOO mile radius of hotnebase
AREAO?^ if the vehicle's greatest percentage of miles traveled were beyond a 200
mile radius from home base
The approximate percentage of annual miles that the vehicle carried payloads that filled
it3 maximum cargo size This variable can be used as an index for the cube-out rate of
the vehicle
The approximate percentage of annual miles that the vehicle carried payloads that filled
its maximum cargo weight This variable can be used as an index for the weigh-out rate
of the vehicle
The size of the vehicle's engine, in cubic inches Rather than an actual number, a
ranged coding scheme is used whereby the responder provides the range in which the
vehicle's engine size falls
The horsepower rating of the vehicle's engine As with CID, rather than an actual
number, a ranged coding scheme is used whereby the responder provides the range in which
the vehicle's horsepower rating falls
AZRCDN-1 if the vehicle has aerodynamic features
EC0EHG=1 if the vehicle has a fuel economy engine with low RIM, high torque rise,
turbocharger, etc
3-14
-------
As a result, EEA used TIUS variables to define new variables that more
accurately explain the factors that determine fuel productivity so that
regression analysis could be employed to quantify the effect of physical and
operational characteristics on a vehicle's fuel consumption rate (GPM). These
EEA variables are defined in Table 3-4.
The variable UTDWGT1 is of special interest to the fuel productivity analysis.
WTDWGT1 is defined as follows:
WTDWGT1 - (.(AVWGHT*FRCLOAD)+(EMWGHT*FRCNLOAD)5/EEAMAX,
where EEAMAX is defined as the high end of the GW range for the vehicle in
question. Therefore, WTDWGT1 describes the actual operating weight of the
vehicle normalized to that vehicle's weight category The numerator of
WTDWGT1 describes the weighted sum of the vehicle's average operating mass by
employing as the weights the loaded to empty mileage ratio and the empty to
loaded mileage ratio. The effect of WTDWGT1 on GPM is expected to be positive
since the greater a mass the more energy input that is required to propel that
mass into motion.
Engine control variables for engine size and horsepower were defined by EEA
using TIUS variables, namely CID and HRSPWR. The purpose of EEA's CIDCTR and
HPCTR variables was to recognize the effect of engine size or horsepower on a
vehicle's fuel consumption rate. The expected relationship being that
vehicles with large engines or high horsepower ratings have higher average
GPMs
A vehicle's age is also expected to directly impact GPM, with older vehicles
being relatively less efficient in the consumption of fuel than newer
vehicles, given recent technological advances in engine and vehicle design
that have improved fuel economy. To recognize this effect, EEA used TIUS's
MODELYR variable to create MDLIDX.
3-15
-------
TABLE 3-4
DEFINITIONS OF EEA VARIABLES
VARIABLE
(EEA)
GPM
LGPM
MDLIDX
FRCLOAD
FRCNLOAD
EQUIWT
EEAMAX
WTDWGT1
LWTWGT1
WIDWGT2
LWTWGT2
WIDWGT3
LWTWGT3
MIDCID
MEANCID
CIDCTR
LCIDCTR
MIDHP
MEANHP
HPCTR
LHPCTR
DLOCAL
DAERO
DECO
DKKlUlTltW
GPM - 1 / MPG
Defined as gallons-per-mile, the inverse of MPG
LGPM = In (GEM)
The natural log of GPM
MDLIDX = MODELYR - 1976
Defined as model year index whose range of possible values is 1 to 11
FRCLOAD - (100 - PNOLOD) / 100.
The fraction of annual mileage that the vehicle did carry a payload
FRCNLOAD = PNOLOD / 100
The fraction of annual mileage that the vehicle did not carry a payload
EQUIVWT = (AVWGHT X FRCLOAD) + (EMWGHT x FRCNLOAD)
The equivalent weight, or average operating weight, of the vehicle accounting for empty
mileage
EEAMAX is defined as the high end of a vehicle's EEAGVW range, expressed in tons
EEAMAX = 13 0 for vehicles in GVW Class 5
EEAMAX = 16 5 for vehicles in GVW Class 6
EEAMAX = 30 0 for vehicles in GVW Class 7
EEAMAX = 40 0 for vehicles in GVW Class 8
EEAMAX ¦ 70 0 for vehicles in GVW Class 9
WTDWGT1 - EQUIVWT / EEAMAX
The equivalent weight of the vehicle, normalized to that vehicles weight class
LWTWGT1 = In (WTDWGT1)
The natural log of WTDWGT1
WTDWGT2 = AVWGHT / EEAMAX
The average weight of the vehicle, not accounting for empty mileage, normalized to that
vehicle's weight class
LWTWGT2 = In (WTDWGT2)
The natural log of WTDWGT2
WTDWGI3 = EMWGHT / EEAMAX
The empty weight of the vehicle, normalized to that vehicle's weight class
LWTWGT3 = Ln (WTDWGT3)
The natural log of WTDWGT3
MIDCID is the midpoint of the reported CID range for a vehicle
MEANCID is the expected value, or mean, of the MIDCIDs for a give GVW clas3
CIDCTR = MIDCID / MEANCID
A vehicle's approximate engine size, normalized to that vehicle's weight class
LCIDCTR = ln (CIDCTR)
The natural log of CIDCTR
MIDHP is the midpoint of the reported HRSPWR range for a vehicle
MEANHP is the expected value, or mean, of the MIDHPs for a given EEAGVW class
HPCTR = MIDHP / MEANHP
A vehicle's approximate horsepower rating, normalized to that vehicle's weight class
LHPCTR = ln (HPCTR)
The natural log of HPCTR
DLOCAL is a dummy variable that equals 1 when AREAOP equals 2 and DLOCAL equals 0 if
AREAOP does not equal 2
DAERO is a duumy variable that equals 1 if AERODN equals 1 and DAERO equals 0 if AERODN
does not equal 1
DECO is a dumny variable that equals 1 if ECOENG equals 1 and DECO equals 0 if ECOENG
does not equal 1
3-16
-------
Finally, dummy variables were created to assess the impact of a vehicle's area
of operation and the presence of fuel economy engines or aerodynamic devices.
EEA used TIUS variables AREAOP, ECOENG, and AERODN to create dummy variables
DLOCAL, DECO, and DAERO.
Various linear and log-linear model specifications, employing different
combinations of the variables listed in Table 3-4, were tested to determine
the most statistically significant model for each GVW class and fuel type
combination. The following models were chosen based on their statistical and
analytical significance:
• For gasoline powered trucks, irrespective of GVW class, the model that
displayed the best statistical significance was;
ln(GPM) - a+b*(MDLIDX) +c*ln(WTDWGT1) +d*ln(C1DCTR.) +e*(DLOCAL)
• For diesel trucks with a GVW classification between 19,500 lbs and
60,000 lbs that model was;
ln(GPM) - a+b*(MDLIDX)+c*ln(WTDWGT1)+d*In(HPCTR)+e* (DLOCAL)+f*(DAERO)
• For diesel trucks in GVW Class 8B (60,000 lbs to 80,000 lbs) that model
was ;
ln( GPM) = a+b*(MDLIDX)+c*ln(WTDWGT1)+d*ln(HPCTR)+e*(DLOCAL)+f*(DAERO) +
g*(DEC0)
The significance of the coefficient of aerodynamic devices on gasoline trucks
was not s tatistically different from zero, and thus the DAERO dummy variable
was excluded from that model The coefficients displayed no statistical
significance because of the small number of gasoline trucks that actually had
aerodynamic devices. For example, of the gasoline vehicles in Class 5 only
1 0% had aerodynamic devices Of all gasoline trucks in Class 6 only 1 4% had
aerodynamic devices, while of those in Class 7 only 0.9% had them (see Section
2). Similarly, other technology variables (such as the presence of radial
tires, variable fan drives and fuel efficient axle or drive ratios) were
tested but also did not display any statistical significance. Finally, note
that the engine control variable that displayed the best statistical
significance for gasoline powered trucks was engine displacement, rather than
horsepower. The lack of variability m horsepower across a GVW class/gasoline
3-17
-------
combination explains this insignificance. The exact opposite was found to be
true for diesel vehicles.
The statistical results of the chosen regression models are presented in
Appendix A. The significance probabilities, or p values (shown in the output
tables as Prob > |t|), indicate that the estimated coefficients are
significant above the 95% significance level. The fitted models are presented
in Table 3-5.
3.4 THE EFFECT OF EMPTY MILEAGE
Empty mileage usually takes place when a truck carries an empty backhaul
(i e , carries a payload in one direction but not in the other). "The 1980
Interstate Commerce Act abolished backhaul restrictions on the federal
interstate highway system. Today any carrier which obtains Interstate
Commerce operating authority - and does not, for example, carry food produce
in one direction and hazardous waste in the other, as stipulated by the recent
"Garbage Bill" - is free to carry backhaul loads and avoid empty mileage when
traveling the interstate roads. However, the ability of a carrier to avoid
empty mileage largely depends on logistical and managerial factors within the
carrier's trucking company and on state laws regarding intra-state backhauls
This section estimates the effect of empty mileage on fuel productivity and
discusses possible policies that may minimize empty mileage.
The effect of empty mileage on fuel productivity can be estimated using the
variables and regression models that were developed by EEA and explained in
Section 3 3 If a vehicle experiences empty mileage in a given trip, then the
vehicle's operating weight is represented by AVWGHT, or empty weight plus
typical cargo weight. However, a vehicle on any given trip will incur some
empty mileage and some loaded mileage. In such situations, the vehicle's
operating weight will equal EQUIVWT, which is defined as the weighted sum of
AVWGHT and EMWGHT UTDWGT3 must be used to estimate fuel consumption (GPM)
when a truck operates empty, since it reflects a vehicle's operating weight
when empty WTDWGT2 should be used to estimate fuel consumption when a truck
3-18
-------
TABLE 3-5
LISTING OF FITTED MODELS
Model 1. GVW Class 6 (Gasoline Trucks)
IjGPM ~ -1 7801 - 0 0049*MDLIDX + 0.2346*LWTWGT1 -I- 0 4403*LCIDCTR + 0.0503*DLOCAL
(0.0165) (0.0019) (0.0247) (0 0391) (0 0133)
Model 2" GVW Class 6 (Diesel Trucks')
LGPM = -1 7854 - 0.0278*MDLIDX + 0 2209*LWTWGT1 + 0.0966*LHPCTR + 0.0337*DLOCAL - 0.0691*DAERO
(0 0181) (0 0017) (0.0343) (0.0512) (0 0123) (0.0227)
Model 3" GVW Class 7 (Gasoline Trucks)
LGPM = -1 6990 - 0 0092*MDLIDX + 0.1963*LWTWGT1 -t 0.5560*LCIDCTR + 0.0180*DLOCAL
(0 0235) (0 0025) (0 0316) (0 0602) (0.0193)
Model 4' GVW Class 7 (Diesel Trucks)
LGPM = -1 7435 - 0 0300*MDLIDX + 0.2336*LWTWGT1 + 0.0930*LHPCTR 4- 0.0635*DLOCAL - 0.0415*DAERO
(0.0172) (0 0017) (0 0314) (0.0522) (0.0125) (0.0228)
Model 5: GVW Class 8A (Gasoline Trucks)
LGPM = -1.5181 - 0.0102*MDLIDX + 0.1492*LWTWGT1 + 0.5729*LCIDCTR -t- 0 0416*DLOCAL
(0.3344) (0 0040) (0 0387) (0 0985) (0.0223)
Model 6: GVW Class 8A CDiesel Trucks)
LGPM ~ -1.6020 - 0 0112*MDLIDX + 0.1743*LWTWGT1 + 0.0630*LHPCTR + 0.0509*DLOCAL - 0.0573*DAERO
(0.0085) (0.0009) (0 0123) (0 0099) (0.0061) (0.0096)
Model 7- GVW Class 8B (Diesel Trucks)
LGPM = -1.5546 - 0 0112*MDLIDX + 0.0426*LWTWGT1 + 0.1113*LHPCTR + 0.0255*DLOCAL - 0.0181*DAERO - 0.0275*DECO
(0.0032) (0 004) (0 0058) (0 0072) (0.0029) (0.0043) (.0047)
Standard errors of the coefficients are shown in parenthesis.
For definitions of the variables see Table 3-4
-------
carries a typical load, since it reflects a vehicle's operating weight when
carrying cargo.
The potential gain in fuel productivity from eliminating empty mileage is
calculated as follows:
MPG. X fAVWGHT - EMHGHT1 - 1 ) X 100
[ (FRCLOAD * MPG,.) + (FRCNLOAD x MPG,)] X [EQUIWT - EMWGHT) /
- Percent gain in fuel productivity for the average fuel type
j truck in class i,
- Estimated MPG when carrying cargo,
- Estimated MPG when operating empty.
Payload when no empty mileage is incurred on an average trip is given by
[AVWGHT - EMWGHT], and payload when some empty mileage is incurred is given by
[EQUIWT - EMWGHT] . Therefore, the numerator of the expression in parenthesis
represents fuel productivity when empty mileage is not incurred in a given
trip, while the denominator represents fuel productivity when some empty
mileage .is incurred.
Before MPG can be estimated under either definition, it is first necessary to
determine the appropriate values for all the variables in a corresponding
regression model Table 3-6 presents the statistical means by GW class and
engine type combination for some of the variables that were discussed in
Section 3.3. These means can be interpreted as the average characteristics of
a typical truck in a GW class and engine type combination The mean for each
regression variable can, therefore, be plugged into the corresponding model to
estimate GPME and GPMc. For example, average GPME for GW Class 8B is
calculated from Model 7 in Table 3-5 as:
where ,
Gij
MPGC
MPGe
3-20
-------
TABLE 3-6
STATISTICAL MEANS OF SELECTED
VARIABLES BY GVW CLASS AND ENGINE TYPE COMBINATION
- GVW Class 6 -
Gasoline Diesel
- GVW Class 7 -
Gasoline Diesel
- GVW Class 8A -
Gasoline Diesel
GVW
Class 85
Diesel
U>
i
ro
# of Obs
5,
159
2,<
425
2,401
2,
238
2,
166
8
,973
19
ANNMIL
7,
160
19,
320
7,720
20,
235
8,
230
27
,725
65
MPG
5
.96
6
.94
5
68
6.
58
4
.83
5
.32
5
EMWGHT1
5
53
6
.70
6.
.58
7.
85
8
44
10
.94
13
AVWGHT1
11
.15
11
.32
14.
.34
14.
62
21
43
23
.11
35
FRCNLOAD
0
.36
0
.29
0
.36
0
28
0
.41
0
.33
0
EQUIVWT1
9
.00
9
.95
11.
.34
12.
68
16
.00
18
.91
28
MDLIDX2
2
21
5
.45
2
39
5.
33
1
84
3
.80
5
X w/DLOCAL=l
84
40
66
.10
86.
.20
66.
40
86
.30
63
.40
25
% w/DAERO=l
1
00
6
.30
1.
,40
6.
40
0
90
9
.80
22
% w/DECO=l
0
.10
2
.80
0.
80
3.
00
0
.20
5
.60
15
LHPCTR
-
-0
.0099
-0.
010
-
-0
.0485
-0
LCIDCTR
-0
.0077
-0.
0081
-
-0
0027
-
LWTWGT1
-0
.3863
-0
.2817
-0.
.3967
-0
2793
-0
6499
-0
.4865
-0
LWTWGT2
-0
1633
-0
1478
-0.
1483
-0.
1287
-0
.3510
-0.
.2801
-0
LWTWGT3
-0
.9035
-0
. 7085
-0.
9781
-0.
7936
-1
312
-1
0604
-1,
1 Expressed in tons.
2 This second block of variables give regression values used in estimating GPM.
-------
LGPMjs 1-5546 - 0.0112*(5.16) + 0.0426*(-1.1034) + 0.1113*(-0.0127)
+ 0.0255*(0.2540) - 0.0181*(0.2250) - 0.0275*(0 1590)
¦= -1.6628
where, MDLIDX - 5.16, LWTWGT3 - -1.1034, LHPCTR - -0.0127, DLOCAL = 25.4%,
DAERO - 22.5%, and DECO - 15.9% (see Table 3-6).
Taking c^e anti-log of -1.6628 gives 0.1896, or the estimate of average GPME.
The inv®rse 0.1896 defines MPGE, which in this case equals 5.2743.
The same procedure can be used to estimate MPGc for GVW Class 8B under the
WTDWGT2 definition of vehicle weight. Using LWTWGT2 instead of LWTWGT3, and
assuming that all other variables remain constant, MPGC is estimated to be
5.0582.
Having estimated MPGE and MPGC, G^ can be calculated from Class 8B's data in
Table 3-6 • Simply, payload when empty mileage is not incurred equals 22.20
tons. Payload when empty mileage is incurred equals 15.18 tons. It follows
that fuel productivity is 112.29 ton-miles per gallon under no empty mileage
and 77 . 81 ton-miles per gallon with empty mileage. A.s a result, G8B D is
estimated at 44.31%.
This procedure was performed for each GVW class and engine type combination.
The reSults are presented in Table 3-7. It is clear that empty mileage has a
severe detrimental effect on fuel productivity. Avoiding empty mileage can
potentially increase fuel productivity by 33 to over 60 percent. However,
there many reasons why these increases will not be achieved. The ability
of a eerier to avoid empty mileage greatly depends on logistical and
manage^3-! factors. First, a carrier may not have perfect information
regardinS the availability of a load, and miss the opportunity to carry it
Even if perfect information exists, a carrier may choose to return empty
rather than wait for the next available load if time restrictions are a
concern- Second, irregular route, specific commodity carriers may never
encounter traffic for which a corresponding backhaul exists within the
constr^nt:s commodity and point-to-point restrictions. Third, backhaul
3-22
-------
TABLE 3-7
THE EFFECT OF EMPTY MILEAGE
ON FUEL PRODUCTIVITY
GVW Class
Class 6 Gasoline
Class 6 Diesel
Class 7 Gasoline
Class 7 Diesel
Class 8A Gasoline
Class 8A Diesel
Class 8B Diesel
Fuel Productivity1 With
Empty Mileage (WTDWGT1")
22.95
23. 77
28.82
33.65
38.06
44. 37
77 53
Fuel Productivity1
Without Empty Mileage
(WTDWGT2)
33 67
32.56
44.17
45.04
61.51
64.63
112.29
Potential Gain From
Eliminating Empty Mileage
51 33%
36.98%
53.26%
33.80%
61.61%
45.66%
44.31%
1 Ton-miles per gallon
-------
restrictions at the state level may not allow certain carriers to avoid empty
mileage. Currently, approximately 30 states have economic regulations that
prohibit intra-state backhauls For example, a carrier originating in Chicago
with destination to San Antonio is not allowed (under Texas law) to transport
a load from San Antonio to Houston on his/her return to Chicago. Such
restrictions are of particular significance in larger states Finally,
backhaul restrictions stemming from the "Garbage Bill" limit some carrier's
ability to minimize empty backhauls. The "Garbage Bill" does not, for
example, allow carriers who haul food produce in one direction to haul
hazardous waste in the other For these reasons only a fraction of the
potential benefit from eliminating empty mileage will be recoverable.
Trucking companies can alleviate the detrimental effects of empty mileage by
introducing managerial and logistical techniques that identify available loads
and better manage a carrier's route. For example, innovation has led some
firms to find market niches where empty mileage can be minimized, while
computerization has allowed managers to better coordinate routes and identify
available loads Individual truckers, on the other hand, may not be able to
implement innovative management processes or computerize their operations
because of cost constraints and practical reasons. These truckers are
specially hurt by empty mileage since fuel costs often account for a large
portion of their total cost. Individual trucking practices are usually one-
man operations that operate at small profit margins and cannot incur the
additional cost of computerization or innovation.
From a policy perspective, a federal mandate that eliminates or relaxes
backhaul restrictions at the state level may prove worthwhile Such a policy
is likely to have a significant impact in large states, like Texas and
California, where distances between markets are greater and trucks are often
forced to incur empty backhauls for many miles Also, government sponsored
clearinghouses that locate and inform individual truckers on an on-time basis
about the availability of backhaul loads may prove to be helpful in
eliminating the information problem. These clearinghouses would specifically
help individual truckers and small trucking companies that have not
3-24
-------
computerized their operations. Such a program would be cost effective if a
fee structure is designed that is economical to users and a revenue source to
government.
3.5 FUEL PRODUCTIVITY UNDER AN ONCAPPED BRIDGE FORMULA B
This section quantifies fuel productivity gains from eliminating the 80,000
lbs GVW cap. Under this scenario, GVW would be controlled by existing axle-
weight limits and by Bridge Formula B. As was shown in Section 3.2, allowable
GVWs under the Bridge Formula are determined by the number of axles on a
vehicle and the distance between the extremes of any group of 2 or more axles.
Eliminating the 80,000 lbs GVW cap would not affect conventional 5-axle
tractor-semitrailers, since these configurations cannot exceed 80,000 lbs
without violating current axle-weight limits However, 6-axle tractor-
semitrailers and Twin 28s would be directly affected by this policy. Under
the uncapped Bridge Formula B scenario, 6-axle tractor-semitrailers would
typically be allowed to operate at 86,000 lbs GVW, while 5-axle Twin 2Bs would
be allowed to 'gross-out' at 91,500 lbs. Seven-axle, 8-axle, and 9-axle Twin
28s would be allowed to operate, without special permit, at 99,500 lbs,
104,500 lbs, and 110,000 lbs, respectively. The analysis assumes that no
changes in state length limits will take place, and fuel productivity
calculations are based on 48 ft semitrailers for tractor-semitrailer
configurations and 28 ft trailers for double trailer configurations to assure
that the benefits reflect nationwide operation. The nationwide operation of
LCVs will not be affected. By definition these configurations are double
trailers that have at least one trailer longer than 28 ft violating most
states' length limits (see Table 3-1). The fuel productivity benefits of LCVs
are recognized later in the section through fuel productivity comparisons with
5-axle, 6-axle tractor-semitrailers and Twin 28s operating under the uncapped
Bridge Formula B scenario.
Fuel productivity changes resulting from the elimination of the 80,000 lbs GVW
cap can be estimated by employing the regression model for GVW Class 8B that
is specified in Section 3 3 Because GVW Classes 6, 7, and 8A do not include
5-axle ot greater vehicle configurations, trucks in these classes are not
3-25
-------
affected by a policy that eliminates the 80,000 lbs GW cap. Moreover, not
all vehicles in GW Class 8B will be affected by a policy that eliminates the
80,000 lbs GW cap. Carriers that often carry light-density freight will
rarely be weight-limited. These carriers will be limited by volumetric
capacity and restrictions on the vehicle's size.** On the other hand, those
vehicles in GW Class 8B that carry high-density freight are often weight-
limited by the 80,000 lbs cap.
To distinguish between the operational and physical characteristics of weight-
limited vehicles versus all vehicles in GW Class 8B, EEA employed TIUS'
PCARWT variable Those vehicles in GW Class 8B that exhibited a PCARWT value
of greater than 75% were identified as being weight-limited by the GW cap.
Out of 19,974 vehicles in this class 3,276 met this criteria. However, TIUS
data for GW Class 8B does not distinguish between 5-axle, 6-axle, 7-axle, 8-
axle, and 9-axle vehicle configurations To get around this problem EEA
assumed Che following:
• MDLIDX to be 11 and DAERO to be 1 for all configurations so
that fuel productivity comparisons are made only between new
vehicles equipped with drag reduction devices.
• For all configurations, the values for DLOCAL and DEC0 can be
approximated from the statistical frequencies of weight-limited
vehicles in GW Class 8B (i e., those 3,276 trucks that met the
PCARWT greater than 75% criterion).
• Horsepower ratings under the uncapped Bridge Formula B scenario
increase by the same percentage as GW for increasing axle
configurations.
Table 3-8 presents the average physical and operational characteristics for
these vehicle configurations under both the capped and uncapped Bridge Formula
B scenarios (LCV data is shown for later reference). Statistical analysis was
performed, on weight-limited vehicles in Class 8B to determine the values for
Volumetric capacity restrictions are not relevant in the immediate
analysis since size limits remain constant But such restrictions are relevant
when comparing fuel productivity between LCVs and conventional configurations
Carriers that are currently transporting light density freight with Twin 28s will
be able to increase payload by shifting to LCVs that have longer trailers
3-26
-------
TABLE 3-8
PHYSICAL AND OPERATIONAL CHARACTERISTICS
OF WEIGHT-LIMITED 5-AXLE OR GREATER TRUCKS
5-Axle 6-AxIe 5-Axle 7-Axle 8-Axle 9-Axle Rocky Mt Turnpike
Semi. Semi. Twin 28 Twin 28 Twin 28 Twin 28 Doubles Doubles
Empty Weight (lbs)1
26,800
28,300
30,000
34,200
36,700
39,200
36,000
45,000
GVW Under 80,000 lb
GVW Cap (lbs)
80,000
80,000
80,000
NA
NA
NA
NA
NA
GVW Under Uncapped
Bridge Formula (lbs)
80,000
86,000
91,500
99,500
104,500
110,000
108,500
131,701
Payload Under 80,000 lb
GVW cap (lbs)
53 ,200
51,700
50,000
NA
NA
NA
NA
NA
Payload Under Uncapped
Bridge Formula (lbs)
53,200
57,700
61,500
65,300
67,800
70,800
72,500
86,700
LHPCTRc2
0.1068
0.1068
0.1068
NA
NA
NA
NA
NA
LHPCTR^2
0.1068
0.1791
0 2411
0 3249
0.3739
0.4252
0.4115
0.6053
MDLIDX
11
11
11
11
11
11
11
11
% w/DL0CAL=l
17.6
17.6
17 .6
17.6
17 .6
17.6
17.6
17.6
% w/DAEROl
100
100
100
100
100
100
100
100
% w/DEC0=l
22.0
22.0
22 .0
22.0
22.0
22.0
22.0
22.0
1 Empty Weights reflect tractors pulling dry vans
Empty weight varies by trailer types and length of trailers
2 LHPCTR estimates reflect AVWGHT of 35.95 tons for Class 8B trucks (Table 3-6).
NA: Not applicable since these configurations seldom, if ever,
operate below 80,001 lbs.
-------
DLOCAL and DECO. The horsepower control variable for the uncapped Bridge
Formula B scenario (LHPCTRy) was scaled according to the percentage change in
GVW between increasing axle configurations. Typical empty weights for
different configurations were determined from conversations with staff at Jack
Faucett Associates and the American Trucking Association, and reflect tractors
pulling dry vans. The numbers for GVW under the uncapped Bridge Formula B are
those that: were derived in Section 3.2.*** Average operating GVW under the
80.000 lbs cap for 5-, 6-axle tractor-semitrailers and 5-axle Twin 28s is
assumed to be 80,000 lbs since the analysis reflects weight-limited vehicles.
For 7-axle, 8-axle, and 9-axle Twin 28s, GVW under the 80,000 lbs cap has no
practical meaning, since these configurations seldom, if ever, operate below
80.001 lbs.
Having established the typical physical and operational characteristics of
these vehicle configurations, fuel productivity estimates can be derived for
both scenarios Regression Model 7 was used to calculate GPM under both GVW
scenarios and, together with payloads in Table 3-8, fuel productivity
estimates were derived Table 3-9 demonstrates the results of the
calculations. In practice, most weight-limited carriers facing an 80,000 lbs
GVW limit operate 5-axle tractor-semitrailers This is supported by the
estimates in Table 3-9, which indicate that weight-limited operations of other
than 5-axle tractor-semitrailers result in fuel productivity losses to the
carrier. Higher empty weights of 6-axle tractor-semitrailers and 5-axle Twin
28s allow less payload to be carried when GVW is limited to 80,000 lbs.
In contrast, when the 80,000 lbs GVW cap is eliminated, Table 3-9 shows that
the least productive configuration is a conventional 5-axle tractor-
semitrailer. Lifting the GVW cap will result in substantial increases in the
use of 6-axle tractor-semitrailers and Twin 28s, as weight-limited carriers
shift away from 5-axle tractor-semitrailers to these configurations to take
Note that unlike the empty mileage analysis, estimating fuel
productivity changes from eliminating the 80,000 lb GVW cap is not concerned with
EQUIVWT since the concern is on fuel productivity differences when carrying a
load.
3-28
-------
TABLE 3-9
ESTIMATES OF FUEL PRODUCTIVITY GAINS
FROM ELIMINATING THE 80,000 lb GVW CAP
Fuel Productivity1
Fuel Productivity1 Under Uncapped
Truck Configuration Under 80.000 lb GW Cap Bridge Formula B
5-Axle
Tractor-Semis
142.86
142
86
6-Axle
Tractor-Semis
138.83
153.
.29
5-Axle
Twin 28s
134.27
161.
,76
7-Axle
Twin 28s
NA
169
61
8-Axle
Twin 28s
NA
174,
.74
9-Axle
Twin 28s
NA
181
07
Ton-miles Per Gallon
NA: Not applicable since these configurations seldom, if ever,
operate below 80,001 lbs.
3-29
-------
advantage of higher fuel, productivity. For example, the potential benefit to
a weight-limited carrier of shifting from a 5-axle tractor-semitrailer to a 6-
axle tractor-semitrailer is an increase in fuel productivity of about 11.
Shifts to 5-axle Twin 28s can result in a fuel productivity gains of
approximately 13%, while shifts to 9-axle Twin 28s can result in a gain of
roughly 27%. In the short-run, however, Twin 28s may not be practical for
certain carriers because the ease of shifting to Twin configurations is
constrained by the following factors:6
• Previous investment in expensive equipment of long life,
particularly tanks, but also refrigerator units and concrete
mixers,
• The extra cost of handling Twins, which increases as a
proportion of total cost for short hauls,
• The extra cost of handling Twins in tightly constrained
terminal areas,
• Possible capacity constraints of shippers and receivers because
of limitations of space in existing facilities, particularly of
van operations that serve dense urban areas with high space
cost,
• Some shippers may be slow in changing production processes or
modifying their facilities to take advantage of potential
productivity improvements.
In the long run, the fuel productivity benefits of Twin 28s are expected to
outweigh the costs of operating these configurations for those carriers that
can benefit most from the added payload capacity, Terminal areas and existing
facilities will be redesigned to accommodate the handling of Twin 28s, current
equipment will depreciate and be replaced with Twin 28s, and shippers that
were previously slow in recognizing the productivity benefits of 7-axle, 8-
axle, and 9-axle Twin 28s will have reacted to potential gains
It should be noted that the analysis presented above describes maximum fuel
productivity attainable under each scenario. The analysis assumes that new
trucks operate at maximum payload capacity under both the capped and uncapped
Bridge Formula B scenarios and that trucks incur no empty mileage on a given
trip. ALso, the estimated vehicle weight coefficient is lower in Model 7 than
3-30
-------
in those Models for lighter truck classes, suggesting that operating weight
has less of an impact on the fuel consumption of Class 8B trucks than on the
fuel consumption of trucks in lighter GVW classes. From an engineering
standpoint, however, the effect of operating weight should be much higher
(approximately 0.30) than the estimated value (0.0426), since this effect is
expected to increases as the weight of the vehicle increases. EEA has
conducted extensive analysis to identify inconsistencies inherent in the data,
but must conclude that other unidentifiable compensating effects are driving
Class 8B' s LWTWGT1 coefficient to such a low level. With this caveat in mind,
EEA is confident that the analysis does provide reliable estimates of
potential fuel productivity gains from eliminating the 80,000 lbs GVW cap.
The Fuel Productivity of LCVs
Under assumptions made in this study, the nationwide operation of LCVs would
not be affected by the elimination of the 80,000 lbs GVW cap because LCVs have
longer trailer units that violate most state length limits. However, in those
states where length restrictions allow their operation, carriers operating
both size and weight-limited loads will benefit by shifting to LCVs when the
80,000 lbs GVW cap is lifted and permits are not required for their operation.
This section characterizes the fuel productivity of Rocky Mountain Doubles and
Turnpike Doubles through comparisons with 5-axle, 6-axle tractor-semitrailers
and Twin 28s operating under the uncapped Bridge Formula B scenario.
In those states where length restrictions do not prohibit the operation of
Rocky Mountain Doubles and Turnpike Doubles, lifting the 80,000 lbs GVW cap
will affect carriers operating weight-limited Twin 28s and carriers operating
size-limited 5-, 6-axle tractor-semitrailers or Twin 28s. As shown earlier in
Section 3.5, weight-limited operators of 5- and 6-axle tractor-semitrailers
benefit by shifting their operations to Twin 28s. Weight-limited carriers
operating Twin 28s, on the other hand, can gain by shifting their operations
to Rocky Mountain and Turnpike Doubles since these LCVs have longer distances
between extreme axles and are thus allowed higher GVWs under Bridge Formula B.
Size-limited carriers operating either 5-, 6-axle tractor-semitrailers or Twin
3-31
-------
28s also benefit if allowed to switch to LCVs. Trailers pulled by LCVs are
longer and have higher volumetric capacities.
Fuel productivity estimates, under a weight-limited scenario, for Rocky
Mountain and Turnpike Doubles can be calculated using regression Model 7 and
the physical and operational characteristics that are shown in Table 3-8 EEA
estimated the fuel productivity of Rocky Mountain and Turnpike Doubles to be
185.80 and 216.42 ton-miles per gallon, respectively. Comparing these
estimates to those for Twin 28s under the uncapped Bridge Formula B (Table 3-
9) provides approximations of the fuel productivity gains possible from shifts
to Rocky Mountain and Turnpike Doubles. Such shifts are estimated to result
in increases of 2 6% to 15.0% for Rocky Mountain Doubles and of 19.0% to 33.8%
for Turnpike Doubles. Of course, these benefits are not attainable on a
nationwide scale, but are attainable in those states where length restrictions
will not prohibit the operation of these LCVs when the 80,000 lbs GVW cap is
lifted and special permits are no longer required.
Shifting from 5-, 6-axle tractor-semitrailers and Twin 28s to Rocky Mountain
and Turnpike Doubles will also benefit those carriers that frequently carry
size-limited cargo Size-limited carriers are constrained by the volumetric
carrying capacity of the vehicle Table 3-10 shows the typical dimensions and
the volumetric capacities of the hauling units of 5-, 6-axle tractor-
semitrailers, Twin 28s and these LCVs The volumetric capacities of Rocky
Mountain and Turnpike Doubles are greater than those of other truck
configurations. For example, shifting from any Twin 28 to a Rocky Mountain
Double results with a gain in volumetric capacity of 2,941 cubic ft, while a
shift to Turnpike Doubles results with a gain of 5,406 cubic ft.
However, gains in volumetric capacity must be translated into fuel
productivity benefits This translation requires data describing the typical
freight densities at which size-limited vehicles operate These densities can
then be used as a baseline to calculate the potential payload gains that
result from shifts to these LCVs. Typical payloads for cube-limited tractors
pulling dry vans were determined from publications by Jack Faucett Associates.
3-32
-------
TABLE 3-10
TYPICAL DIMENSIONS AND VOLUMETRIC CAPACITIES
FOR 5-, 6-AXLE TRACTOR-SEMIS, TWIN 28s, AND LCVs
5-Axle 6-Axle 5-AxIe 7-Axle 8-Axle 9-Axle 7-Axle 9-Axle
Semi. Semi, Twin 2B Twin 28 Twin 28 Twin 28 LCV LCV
Trailer(s) Width (in)
Trailer(s) Height1 (ft)
Trailer(s) Total Length
(ft)
Estimated Volumetric
Capacity (cubic ft)
102
13 5
48.0
102
13.5
48 0
102
13.5
56 0
102
13 5
56.0
102
13 5
56.0
102
13 . 5
56.0
102
14 5
76.0
102
14. 5
96.0
5,508 5,508 6,426 6,426 6,426 6,426 9,367 11,832
1 Most states limit the maximum height of the trailer(s) on tractor-semis
and Twin 28s to 13 1/2 feet However, LCVs usually are allowed to pull
trailers that are 14 1/2 feet high
-------
The typical payload for cube-limited 5-, 6-axle tractor-semitrailers and Twin
28s are estimated at 24,500 lbs and 28,600 lbs, respectively.7 At an average
payload of 28,600 lbs, cube-limited Twin 28s operate with a typical freight
density of 4.45 pounds per cubic ft. Recognizing that Rocky Mountain and
Turnpike Doubles have volumetric capacities that are 2,941 and 5,406 cubic ft
greater than Twin 28s, average payload for these LCVs can be approximated if
the freight density for cube-limited Twin 28s is held constant. At a freight
density of 4.45 pounds per cubic ft, a Rocky Mountain Double can carry
payloads that are approximately 13,100 lbs greater than the typical Twin 28,
while a Turnpike Double can approximately carry an additional 24,100 lbs when
compared to cube-limited Twin 28s. Therefore, Rocky Mountain Doubles are
expected to operate with an average payload of 41,700 lbs, while Turnpike
Doubles are expected to operate with an average payload of 52,700 lbs.
Having determined the average payload weight of size-limited 5-, 6-axle
tractor-semitrailers, Twin 28s, Rocky Mountain, and Turnpike Doubles, fuel
productivity estimates can be derived for size-limited vehicles using Model 7
and the typical physical and operational characteristics presented in Table 3-
11. Data for size-limited trucks in GVW Class 8B was used to determine the
values for DLOCAL and DEC0. LHPCTR was scaled in accordance with increasing
GVWs across axle configurations. The values for MDLIDX and DAERO implicitly
assume that fuel productivity comparisons are made between new vehicles.
Table 3-12 shows fuel productivity estimates for size-limited 5- and 6-axle
tractor-semitrailers, Twin 28s, Rocky Mountain Doubles, and Turnpike Doubles
Fuel productivity for size-limited Twin 28s slightly decreases with increasing
axles. This is because at a constant freight density, Twin 28s carry the same
payload -weight without regard to the number of axles on the vehicle. But as
the number of axles increase the empty weight and total weight of the vehicles
increase without a payload gain As weight increases MPG decreases while
payload remains constant. So, fuel productivity decreases as the number of
axles increase. However, the fuel productivity benefits from shifting to
3-34
-------
TABLE 3-11
PHYSICAL AND OPERATIONAL CHARACTERISTICS
OF SIZE-LIMITED 5-AXLE OR GREATER TRUCKS
5-Axle
Semi.
6-Axle
Semi.
5-Axle
Twin 28
7-Axle
Twin 28
8-Axle
Twin 28
9-Axle
Twin 28
Rocky Mt
Double
Turnpik
Double
Empty Weight (lbs)
26,800
28,300
30,000
34,200
36,700
39,200
36,000
45,000
Estimated Payload
Weight (lbs)
24,500
24,500
28,600
28,600
28,600
28,600
41,700
52,700
GVW Under Uncapped
Bridge Formula (lbs)
51,300
52,800
58,600
62,800
65,300
67,800
77,700
97,700
LHPCTR1
-0 3376
-0.3088
-0.2045
-0.1353
-0.0963
-0.0587
-0.0776
-0.3066
MDLIDX
11
11
11
11
11
11
11
11
X w/DLOCAL=l
14. 7
14.7
14. 7
14.7
14.7
14.7
14.7
14.7
% w/DAERO—1
100
100
100
100
100
100
100
100
% w/DECO=l
25 .4
25 4
25.4
25.4
25.4
25.4
25.4
25.4
1 LHPCTR estimates reflect AVWGHT of 35 95 tons for Class 8B trucks (Table 3-6).
-------
TABLE 3-12
FUEL PRODUCTIVITY ESTIMATES FOR SIZE-LIMITED
VEHICLES AND POTENTIAL GAINS FROM SHIFTS TO LCVs
Fuel Productivity1 Potential Gain From Potential Gain From
With Uncapped Shifting to Rocky Mt Shifting to Turnpike
Truck Configuration Bridge Formula Doubles Doubles
5-Axle Tractor-Semis.
70.52
59 7%
94.8%
6-Axle Tractor-Semis.
70.24
60 4%
95.6%
5-Axle Twin 28s
80 70
39.6%
70.2%
7-Axle Twin 28s
79.84
41 1%
72.1%
8-Axle Twin 28s
79. 36
41.9%
73.1%
9-Axle Twin 28s
78.87
42.8%
74.2%
Rocky Mountain Doubles
112.64
NA
NA
Turnpike Doubles
137 .38
HA
NA
1 Ton-Miles Per Gallon
NA - Not Applicable
-------
Rocky Mountain arid Turnpike Doubles from Twin 28s and tractor-semitrailers are
substantial to cube-limited carriers. As shown in Table 3-12, shifts to Rocky
Mountain Doubles result in potential fuel productivity gains that range from
39.6% to 60.4%. Shifts to Turnpike Doubles result in gains that range from
70.2% to 95.6%. Of course, these benefits are not attainable on a nationwide
scale, but only in those states where length restrictions will not prohibit
the operation of these LCVs when the 80,000 lbs GVW cap is lifted and special
permits are no longer required.
As in the analysis of fuel productivity gains from eliminating the 80,000 lbs
GVW cap, the estimates calculated for LCVs and size-limited conventional truck
configurations are derived under the assumption that trucks operate at maximum
payload and cubic capacities, incurring no empty mileage. Therefore, the fuel
productivity estimates presented under the discussion of LCVs are also
absolute maximums. The LCV discussion also assumes that fuel consumption of
Rocky Mountain and Turnpike Doubles can be estimated from the regression model
for GVW Class 8B (i.e., Model 7) Given that these LCVs have distinct
physical and operational characteristics than Class 8B trucks, Model 7 may not
be altogether representative of LCV fuel consumption.
To account for this possibility, Model 8 using data for TIUS trucks rated at
above 80,000 lbs GVW (defined as Class 9 trucks) was formulated. It is
evident from the data, however, that Class 9 includes many different truck
configurations and is not wholly defined by LCVs. To single out LCVs, Class 9
trucks were further disaggregated by type of use - rough, agricultural, and
regular. Model 8 only uses data for regular use Class 9 trucks and,
therefore, is expected to roughly represent LCVs.
Statistical results of Model 8 are presented in Appendix A, while the fitted
model is shown below.
LGPM — -1.4320-0 0092(MDLIDX)+0.0616(LWTWGTl)+0 1489(LHPCTR)-0.0591(DECO) .
Using this model and data from Tables 3-8 and 3-11, an alternative estimate of
3-37
-------
fuel productivity for Rocky Mountain and Turnpike Doubles can be derived.
Under a weight-limited scenario, the fuel productivity of Rocky Mountain
Doubles is estimated at 165.30 ton-miles per gallon and that of Turnpike
Doubles is estimated at 189.8 ton-miles per gallon. Under a size-limited
scenario, fuel productivity is estimated at 102.21 and 123,07 ton-miles per
gallon, respectively **** In either case, LCV fuel productivity is much
lower when estimates are derived using Model 8. However, since it is not
inconclusive that Class 9 (regular use) represents only LCVs, the estimates
derived from Model 8 should be regarded as conservatively low, since other
configurations with inefficient engines may be included.
3.6 SUMMARY OF RESULTS
This section of the report has investigated two factors that greatly influence
fuel productivity, as defined by ton-miles per gallon of fuel consumed. These
factors are. 1) the prevalence of empty mileage, or mileage incurred when a
truck is riot carrying payload, and 2) the effect of lifting the 80,000 lbs GW
cap that is imposed by the Federal government on trucks travelling the
interstate highway system and other federally-aided roads Empty mileage
affects all commercial trucks, while the 80,000 lbs GVW weight limit
constrains the payload capacity of heavier trucks, usually 5-axle or greater
configurations.
Empty mileage has a severe detrimental impact on fuel productivity, since when
a truck operates empty fuel productivity is equal to zero EEA determined
that avoiding empty mileage can potentially increase fuel productivity by 35
to over 60 percent, depending on GVVJ class and engine type combination.
However, there are many reasons why these potential increases will not be
realized; such as, the lack of perfect information regarding available hauls,
time restrictions, and backhaul restrictions at the state level
**** Estimates for Rock Mountain Doubles using model 8 are based on a LHPCTR
of 0 1371 (weight-limited) and -0.1968 (size-limited) For Turnpike Doubles they
are based on a LHPCTR of 0.3309 (weight-limited) and 0 0322 (size - limited)
AVWGHT for class 9 regular use trucks is 47.3 tons.
3-38
-------
Nevertheless, there are measures that can be taken by trucking companies and
government to alleviate the impact of empty mileage. These measures are:
• The introduction of managerial and logistical techniques by
trucking companies that identify available loads and better
manage a carrier's route.
• A federal mandate that eliminates or relaxes backhaul
restrictions at the state level, which is expected to have a
disproportionate impact in larger states where carriers are
often forced to incur empty backhauls for many miles.
• Government sponsored clearinghouses that locate and inform
individual truckers about the availability of backhaul loads
The current 80,000 lbs GVW cap results in the inability of a carrier to use
some portion of the vehicle's potential capacity by restricting the weight of
the items being shipped Payload capacity is determined by the difference
between the truck's maximum practical GVW and it's empty (or tare) weight
Carriers operating Class 8B trucks that haul high density freight are often
weight-limited by the 80,000 lbs GVW cap. The weight limit does not allow the
carrier to load the truck to its optimum density where weight and size
capacities are fully utilized Such foregone capacity penalizes fuel
productivity and translates into an opportunity cost to the carrier.
A policy that eliminates the 80,000 lbs GVW cap, and limits GVW through axle-
load limits and Federal Bridge Formula B, will result in substantial increases
in the use of 6-axle tractor semitrailers and Twin 28s, as weight-limited
carriers shift away from 5-axle tractor-semitrailers to these configurations
to take advantage of higher fuel productivity. This higher productivity
results because axle-weight limits constrain the maximum practical GVW of 5-
axle tractor-semitrailers to 80,000 lbs even under an uncapped Bridge Formula
B scenario On the other hand, 6-axle tractor-semitrailers and Twin 28s would
be allowed to operate beyond 80,000 lbs under Bridge Formula B and current
axle-load, limits if the GVW cap is lifted. Payload gains of shifting to these
configurations are expected to be greater than the empty weight penalties that
these heavier configurations impose.
3-39
-------
REFERENCES
1. American Trucking Association, Summary of Size and Weight Limits.
Alexandria, Virginia, July 1990.
2. Transportation Research Board, Special Report 225, Truck Weight Limits:
Issues and Options. Washington, D.C., 1990, pgs 43-44.
3. Transportation Research Board, Special Report 211, Twin Trailer Trucks.
Washington, D C. , 1986, pg 49.
4. Transportation Research Board, Special Report 225, Truck Weight Limits-
Issues and Options. Washington, D.C , 1990, pg. 160.
5. Transportation Research Board, Special Report 225, Truck Weight Limits-
Issues and Potions. Washington, D.C., 1990, pg. 50.
6. Transportation Research Board, Special Report 225, Truck Weight Limits-
Issues and Potions. Washington, D.C., 1990, pgs. 66-67.
7. Trucking Research Institute, Productivity and Consumer Benefits of Longer
Combination Vehicles. Washington, D.C., May 14, 1990, pg. B-6.
3-40
-------
4. TECHNOLOGICAL IMPROVEMENTS TO TRUCKS
4.1 OVERVIEW
In the medium and heavy duty truck segments considered in the analysis, the
diesel engine has a commanding market share. The medium-duty truck market
(19,000 to 50,000 lb) is over 80 percent diesel, while the heavy-duty market
has been completely dieselized for over a decade. It is widely anticipated
that by -the end of this decade, the gasoline engine may be limited to some
small specialized niches, or be used after conversion to compressed natural
gas use. This analysis concentrates primarily on diesel powered medium and
heavy-duty truck technology and the potential for improvement in fuel economy.
The analysis utilizes 1987 as a base year, and TIUS fuel economy data for 1987
are used for baseline fuel economy estimates.
Current diesel powered trucks are already very fuel efficient relative to
their weight. For example, a fully loaded 80,000 lb truck can attain a fuel
economy of 7 to 8 MPG on the highway, which translates to 280 to 320 ton-miles
per gallon. In contrast, a gasoline powered car which weighs 5000 lbs or less
fully loaded can attain a fuel economy of 26 to 30 MPG or 65 to 75 ton-miles
per gallon. It is difficult to increase the diesel trucks fuel economy by
very large amounts in the future without affecting payload capacity
Using the laws of motion and conservation of energy it is simple to show that
fuel consumption (the inverse of fuel economy) is given by
FC ~ bsfc [Ejj + EA + Ek] + bsfc Eac + G(tx + tb) .
Id
Where, bsfc is the average brake specific fuel consumption of the engine,
t)d is the transmission efficiency,
Er is the energy required to overcome rolling resistance,
4-1
-------
Ea is the energy required to overcome aerodynamic drag,
Ejj is the energy lost to inertia forces during acceleration or
climbing gradients,
Eoc is the energy used by accessories,
G is idle fuel consumption,
tA is time at idle,
tb is time of braking.
For large diesel trucks, Eae is generally small relative to EA, EK, and ER
Depending on the cycle, the value of G(t4 + tb) , which represents fuel used
during periods of no useful engine output, can be quite small since G, idle
fuel consumption per unit time, is low for a diesel engine.
The above equation points to methods to improve fuel consumption Improve-
ments to bsfc can be accomplished by:
• increasing engine thermodynamic efficiency,
• reducing friction loss,
• reducing pumping loss, and
• increasing turbocharger efficiency
Fuel consumption can also be reduced by reducing the vehicle related parame-
ters of weight, drag and rolling resistance or reducing accessory loads. Idle
fuel consumption can be reduced by reducing idle time, i.e. switching off at
idle, and by reducing the engine displacement. In the case of weight reduc-
tion, we are referring to the empty weight of the truck. Reduction in empty
weight will reduce truck total weight, allowing either an absolute weight
reduction or an increased payload capacity (when trucks are limited by the
weight rather than the size of the payload) .
4.2 FUEL CONSUMPTION SENSITIVITY
The sensitivity of fuel consumption over any specific driving cycle to a
particular independent truck attribute, such as weight, can be represented as
the derivative of the equation relating fuel consumption to these variables,
4-2
-------
and physically represents the percent change in fuel consumption due to a unit
change in the variable (e.g., a 10 percent change in the weight results in a
'X percent' change in fuel consumption).
Volvo provided a general set of values for the energy demanded (and, hence,
fuel consumed) over a typical European long haul cycle. These values are:
Inertia (Weight) 48.9 %
Aerodynamic Drag 17.4 1
Rolling Resistance 24.6 X
Drivetrain Losses 5 0%
Accessories 3.8 1
The above figures do not explicitly refer to idling and braking energy loss.
Moreover, Volvo stated that these figures are for a highly aerodynamic truck,
and tfrsC average values for aerodynamic losses on a U.S. highway cycle were
considerably higher. Other manufacturers suggested that, depending on the
cycle, aerodynamic drag could account for 30 to 50 percent of energy loss,
while inertia related losses were in the 30 to 40 percent range, at highway
speeds.
Volvo and Navistar provided data from computer simulations of the same
truck/erigine combination loaded to different weights Volvo provided data on
two trucks, a medium duty (Class 7) truck and a F12 longhaul truck rated at
100,000 lb GVW (50 tons). Class 7 trucks had a nominal fuel consumption of 30
1/100 kni at 15 tans GVW and a sensitivity o£ 1.12 1/100 km per ton in city
driving. This leads to a sensitivity coefficient of 0.55, i.e. a 10 percent
weight decrease results in a 5.5 percent fuel consumption decrease, The F12
truck had a nominal consumption of 50 1/100 km at 55 tons GVW on a long haul
driving cycle, and a sensitivity of 0.65 1/100 km per ton. This represents a
sensitivity coefficient of 0.715, which appears to be very high relative to
other opinions. However, this is consistent with the energy breakdown
provided by Volvo, as weight reduction reduces both inertial and rolling
resistance energy loss, which Volvo determined as being 73.5 percent of total
energy-
4-3
-------
Navistar provided data on fuel economy for a Class 6, a Class 8A, and a Class
8B truck over two driving cycles, at different loaded weights. The Navistar
data is shown in Table 4-1. Using these data EEA calculated the sensitivity
coefficients for different truck classes as follows:
GVW Class
6
7/8A
8B
Citv
0.307
0.369
0.523
Highway
0.146
0.188
0.298
These coefficients are much lower than those shown by Volvo, and are more
consistent with the results from regression analysis of TIUS data. For
example, TIUS data shows a weight sensitivity of roughly 22 percent for Class
6/7 trucks, which is between the city and highway coefficients shown above.
However, the TIUS data shows a much lower sensitivity for Class 8B trucks that
are at odds with predicted trends shown above, as well as with the absolute
magnitude. Actual testing conducted on Class 8B trucks by Freightliner
Corporation showed that a 80,000 lb truck with a fuel consumption of 4 MPG
(0.25 GPM) had a fuel consumption decrease of 0.0029 GPM per ton weight
increase, for a sensitivity of 0.46, intermediate to the city/highway values
predicted by Navistar's simulation.
Aerodynamic drag reduction sensitivity factors have not yet relied on accurate
measurements of CD reduction versus fuel economy on a fixed test cycle The
only comprehensive data on CD reductions and fuel savings is based on study of
a roof fairing for a 80,000 lb tractor trailer, conducted by GMC 2/ The
measured drag coefficient for the base (no aerodynamic device) tractor trailer
was a CD of 0.770. Typical reductions of CD were measured at different yaw
angles with the aerodynamic device and a mean (wind averaged) value of ACD was
4-4
-------
TABLE 4-1
SENSITIVITY OF FUEL ECONOMY
TO TRUCK LOAD
Class 6 Truck, 155 HP 7.3 L Navistar dlesel.
Gross Weight Citv F/E Highway F/E
8,000 9.48 9.50
13,000 8.72 9.14
18,000 8 17 8.79
23,000 7.63 8.50
28,000 7.15 8.24
Class 8A Truck, 210 HP DT 466 Navistar diesel.
10,000
9.41
8.68
15,000
8.22
8 40
20,000
8 .13
8.14
25,000
7.60
7 88
30,000
7 .14
7.41
35,000
6.72
7 20
40,000
6.34
7 01
Class 8B Truck, Cummins NTC engine.
30,000 6.59 8 42
40,000 5.99 8.04
50,000 5.49 7 69
60,000 5.05 7.35
70,000 4.68 7 04
80,000 4.35 6.75
Source: Navistar Simulation City Speed: 18.6 MPH
Highway Speed: 53.0 MPH
4-5
-------
found to be 0.170. Fuel savings were measured on track tests conducted at 55
mph. Fuel economy increased from 6.3 mpg without the device to 6.93 mpg with
the device, providing a sensitivity of 0.453 for drag reduction. Truck
industry experts confirm that a 10 percent drag reduction provides about 5
percent fuel economy benefit at highway speeds closely matching the calculated
coefficient. The Volvo data indicates a coefficient of only 0.25 which
appears low but is consistent with the high inertial loss claimed by Volvo.
Since the sum of aerodynamic drag forces and inertial & rolling resistance
forces represents the total force to be overcome, the drag sensitivity
coefficient is related to the inertial & rolling resistance (weight) sensitiv-
ity coefficient. The reduced weight sensitivity coefficient translates to a
higher drag sensitivity for medium-duty trucks. At highway speeds, medium-
duty trucks can have a coefficient for drag as high as 0.75. This is
consistent with engineering expectation since the product of drag coefficient
and frontal area is nearly constant for all trucks from Class 6 to 8B while
weight varies by a factor of 4. Hence, at the lightest weight, drag is a much
more significant factor than weight at the same speed.
Power consumed due to aerodynamic drag scales as the cube of speed. At city
speeds of 19 mph, the highway sensitivity coefficient for a Class 8 truck
should decrease by the cube of the speed ratio, hence
Sensitivity at 19mph - 0 453 x i 19 i 3
- 0.019 (55)
The smaLl coefficient has not been verified by confirmatory testing. A
similar calculation for Class 6/7 truck at 19 mph indicates a sensitivity
factor of 0.031.
Tire roLling resistance changes have a large effect on fuel economy It
should be noted that total rolling resistance is a function of both truck
weight and the tire rolling resistance coefficient. The weight effect
calculated includes both the rolling resistance effect and the reduced inertia
4-6
-------
loss effect, while the change in the rolling resistance coefficient effects
only the tire losses. A study by Goodyear3' on the correlation between tire
rolling resistance and fuel economy provided data on the sensitivity to the
rolling resistance coefficient. The study was performed on a 78,700 lb GW
tractor trailer at 60 mph. The study showed the following relationship
between the rolling resistance coefficient and fuel economy:
Rolling Resistance Fuel Economy
Value % Change Value % Change
Bias ply 43.9 Base 4.10 Base
Radial A 39.9 -9.1 4 27 42
Radial B 37.7 -14.1 4.34 5.8
The factors indicate an average rolling resistance sensitivity of 0.4. Volvo
provided data on a 80,000 lb tractor-trailer on a long haul European cycle
with tires of different rolling resistance, and the data is shown in
Figure 4-1. The data indicates a sensitivity of 0.237. This lower figure is
also indicated in the Goodyear study cited above, as the sensitivity to tire
rolling resistance decreases with decreasing speed and weight. For example,
an empty truck was cited as having a sensitivity of 0.17 to 0.2 at 60 mph
Coefficients in the range of 0 25 to 0.35 are cited by many experts. The
sensitivity factors scaling with speed or weight are not well studied. It is
also likely that the rolling resistance of drive tires that transmit torque to
the road have a greater impact on fuel economy than those used in axles that
do not transmit torque Unfortunately, there is no data to support separate
or independent analysis of those effects.
There is general and widespread agreement that typical driveline efficiencies
are 93 to 95 percent (i.e., 5 to 7 percent of engine torque is lost in the
driveline) for manual transmission equipped vehicles. Automatic transmissions
are currently sold in some medium duty trucks and in buses, which may have
higher Losses.
4-7
-------
Figure 4-1
Long haul
Cerptra I E u. t o p e
Rolling resistance (N/ton)
14.83
Produoad »t VOLVO
-------
Lastly, accessory drive losses for engines equipped with a variable fan drive
are typi-c&^y *n t*ie ^ to ® percent range for medium duty trucks, and somewhat
lower to 5 percent) range for Class 8B trucks. Truck manufacturers suggest
that vafial)^e ^an dri-ves a^e used in over 90 percent of diesel engine powered
trucks.
A 3 j£gIGHT REDUCTION
Weight reduction measures are actively being undertaken by truck manufacturers
in botfr medium-duty and heavy-duty truck classes. In most cases, the weight
reducti011 largely due to two specific reasons:
• the substitution of plastic and aluminum components for steel
components
• the use of more modern, high BMEP engines that weigh much less
than older engines of the same power rating.
Weight reduction by material substitution has been widely introduced by all
major manufacturers in the last three or four years The cab (for non-sleeper
type c^ks) is generally a common design used across all trucks from Class 5 to
8B and plastic hoods and fenders have been introduced by Freightliner,
Paccari and Ford- For example, Ford uses a Reaction Injection Molded (RIM)
hood it1 their new Aeromax trucks. The icems are considered to be cost-effec-
tive b»sed only on production cost and have been introduced regardless of the
fuel s3v^n§s potential. Aluminum is, however, a relatively expensive option
The Alun,^nuni Association has conducted a study of weight savings in certain
load bearing parts as shown in Table 4-2, and this study indicates a weight
saving °^ UP t0 ^00 lbs on a Class 8 tractor Currently, none of these parts
are ma^e from aluminum, although Navistar offers special truck models wich
aluminum frame rails as an option. The estimates in Table 4-2 may be
optimistic. as i-s the usual case with supplier provided data.
NavisCfl-r". however, does offer an aluminum cab that offers a weight saving of
160 to 200 lb over a steel cab, depending on Che model The cost for this
option *-s about $2000, equivalent to $10 to $12 per pound saved Navistar
indicaced that aluminum frame rails were appropriately the same cost as a per
4-9
-------
TABLE 4-2
WEIGHT SAVINGS FOR SPECIFIC ALUMINUM FARTS
Part
Front Bumper
Front and Rear Hubs
Front Wheels
Drive Axle Wheels
Rear Axle Carrier Housing
Rear Axle Carrier Housing
Rear Axle Carrier Housing
Frame Rails
Fuel Tank (100 gallons)
Air starter with aluminum air
reservoir
Weight Savings Per Vehicle. Lbs
37
220
22
90
71
110
131
200
30
47
4-10
-------
pound saved basis. The high prices are, to a degree, a function of the
limited sales volume of these options. In general, neither Volvo nor Navistar
expected or had specific programs for (large) weight reduction through
material substitution, but expected 100 to 200 lbs in weight savings over the
next ten years. Volvo stated the weight reduction potential was 500 kg (1100
lbs) for the tractor, and up to 500 kg for the trailer for a Class 8B
combination truck. Volvo stated that Class 6/7 trucks had only a 200 kg
weight reduction potential. However, it was not clear that these potential
reduction levels could be obtained in a cost-effective manner.
Weight reduction by the use of lightweight engines of high specific output is
already occurring A detailed listing of engines by horsepower category and
their weights is provided in Table 4-3. The 350 to 450 HP range is unlikely
to see any major weight reduction except in certain specific cases, as, for
example, by replacement of the Mack EM-9 V-8 engine with six cylinder engines
Engines in the 350 to 450 HP range are widely used in the Class 8B truck
market by owner operators and fleet owners who need extra power to negotiate
the steep gradients in many Western states. The 290 to 350 HP range is more
commonly specified by fleets in Eastern states that are concerned about fuel
costs. In 1990/91 Caterpillar and Cummins introduced the so called '10 litre'
engines that weigh 1900 to 2050 lb with ratings of 300 to 350 HP The engines
can directly replace the popular Cummins NTC Series and Caterpillar 3406
engines, with a weight savings of 500 to 600 lbs. However, these engines have
only about 30 percent of the Class 8B market in 1991, but can be expected to
increase market share in the future. The 240 to 290 HP segment is usually
specified for 60,000 lb Class 8B trucks or by the 'supermedium' 50,000 lb
Class 8A. trucks In this class, weight savings may be significant only if
owners choose the more traditional 'medium-duty' engines such as the DTA466,
or high output versions of the Caterpillar 3116 These engines will have
lower durability than.the L10 and 3176, and also have lower efficiency so that
the weight reduction will not be a major factor affecting fuel economy.
Similar concerns are valid for the 190-240 HP engines used in Class 7 and some
Class 8A. trucks where moving from a Cummins C-series engine to a B-series may
4-11
-------
TABLE 4-3
COMPARISON OF ENGINE WEIGHTS
1987 vs 1991
(Selected Examples)
HP
Range
1987 Engine Models
Model Weight (lb)
1991 Engine Models
Model
Weight (lb)
350 to 450
Cum NTC (400) 2530
Cat 3406 (400) 2840
DDC 8V-92TTA 2415
Mack EM-9 (400) N/A
Cum NTC (400) 2530
Cat 3406B (425) 2840
DDC Series 60 (450) 2670
Mack E7 (400) 2165
290-350
Cum NTC (315) 2520
Cat 3406 (310) 2760
DDC 6V-92TTA (300) 2020
Mack EM-6 (335) 2165
Cum L10 (330) 1870
Cat 3176 (325) 1945
DDC Series 60 (350) 2630
Mack EM-7 (350) 2165
240-290
Cum L10 (270) 1950
Cat 3306 (270) 2040
DDC 6V-71TA (270) 2175
Mack EM-6 (275) 2160
Cum L10 (270)
Cat 3176 (275)
1950
1945
Navistar DTA 466 (270)* 1475
190-240
Cat 3208N (210) 1340
Cum 6CT 8.3 (210) N/A
GM 8.2T (205) 1120
Nav. DTI466 (210) 1475
Ford 7.8L T (210) 1395
Cat 3116 (250)
Cum BTA 5 9 (230)*
Nav. DT466 (210)
Ford 7 8L (210)
1198
880
1475
1395
<190 HP
Ford 6.6T (170) 1310
Nav DTA310 (175) 1235
GM 8.2N (170) 1095
Ford 6.6T (170)
Cum BTA 5.9 (180)'
Nav. 7 3L (175)
Cat 3116 (185)
1310
880
790
1190
Engine HP in parentheses
Engines do not have the same durability as others in this class
4-12
-------
be impractical due to the reduced durability. The Caterpillar 3116 and the
Cummins C series engines are replacing the older Navistar DT466 and
Caterpillar 3208 models in the medium-duty Class 6 and 7 trucks. In the lower
than 190 HP engine range used in Class 6, trucks can shift to some of the new
light-heavy-duty engines such as the Cummins BT 5.9 litre, also with some loss
in durability. On a class average basis, industry experts anticipate that
engine changes (from the 1987 baseline) will contribute to the following
weight reduction by class:
• Class 8B - 250 lbs
• Class 8A - 200 lbs
• Class 7/6 - 150 lbs
The reductions are in addition to the reductions forecasted for material
substitution.
4.4 AERODYNAMIC DRAG
Aerodynamic drag coefficients of heavy-duty trucks for a tractor trailer
combination are usually quite high, historically, the drag coefficient (CD)
has been in the 0.75 to 0.8 range While the simple wind deflector mounted on
the cab roof or the trailer nose cone has been available since the late -
1970's, more modern integrated cab/roof fairing designs have become available
from many manufacturers in the mid-to-late 1980's. The new aerodynamic cabs
were pioneered by Kenworth, and most manufacturers have since followed suit.
New models in Navistar's popular 9000 series (Class 8E) have 30 percent lower
drag than the earlier trucks of 1980-1985 vintage when equipped with the full
aerodynamic package and a 102 ft trailer. In the medium-duty segment,
Navistar has also introduced more aerodynamic cabs that have 7 to 12 percent
lower drag, mostly from the change in the shape of the hood.
Navistar's simulation handbook provides a guide to the relative changes in
drag coefficients achieved by the different aerodynamic devices in a Class 8B
tractor trailer combination. The 9600 series is a 'cabover' design while the
7100 is a conventional tractor, and with a high van trailer, the device
4-13
-------
specific drag reduction from the basic cabs (which are already lower drag
designs) are as follows:
9600 7100
Cab roof fairings 12% 102
Side fairings 3.52 3.72
Roof and Side fairings 13.52 192
Full roof and side fairings 182 N/A
TypicalLy, the full roof and side fairing 'aero package' costs $1400 to $1500
for trucks without sleeper cabs, and about $2300 for trucks with a sleeper
cab. Navistar offers a highly aerodynamic 8300 cabover model that, when
equipped with the same aero package, is about 10 percent more efficient than
the 96QO and 7100 models listed above. In general, these drag reductions are
typical for industry average and special 'aerodynamic' trucks, respectively
Potential for further improvements in Class 8B truck aerodynamics exist
Navistar has displayed a prototype truck called 'IDEA' that has If percent
Lower drag than the current best truck (the 830C series] . Navistar identified
the potential for drag imp coven en ts m the tractor alone of up to £5 percent,
while an integrated tractor-trailer can have the potential for up to a AO
percent reduction over the current new truck fleet average. However, many of
the features required, such as tractor-trailer gap seals, tractor skirts and
rounded van corners are not popular due to the payload reduction potential
incurred., as well as the lack of flexibility in switching tractors to differ-
ent trailers. Volvo agreed with Navistar's assessment, and believed that
tractor related reductions are more likely to be realized by 2001. Navistar
suggested that a 15 to 20 percent additional drag reduction would occur due to
market forces by 2001, These values refer only to Class 8B trucks with van
bodies, which account for about % of sales. In other truck types such as
flat-becLs, stake-beds, livestock haulers, petroleum tankers, etc., drag
reductions are likely to be in the 7 to 10 percent range.
Smaller reductions in drag are forecast for medium duty trucks. Currently,
most medium duty trucks do not even feature rounded corners for the van body
4-14
-------
Navistar's simulation analysis suggests significant potential for drag
reduction simply by introducing a radius of curvature for the top and vertical
corners in a van. A rounded corner van body was found to have a 30 percent
drag reduction relative to a square corner van body. Roof fairings for
medium-duty trucks can produce drag reductions of up to 4 percent, but are
difficult to 'tune' due to the lack of space to mount the deflector. Problems
with consumer acceptance of rounded van bodies, as well as lack of development
of van aerodynamics suggests these improvements are unlikely to be realized.
Navistar suggested that 5 to 8 percent additional aerodynamic drag reduction
would be realized from market forces alone.
Volvo and Navistar suggested that the van aerodynamics problem had not been
well addressed because van bodies were independently manufactured by small
manufacturers, who did not have the resources to develop low drag designs.
Tractor- trailers integration was another area where manufacturers'
representatives believed that further scope for cost-effective improvement was
available, but market forces were insufficient to cause these improvements to
occur
4.5 ROLLING RESISTANCE REDUCTION
The advent of radial tires brought about significant decreases in rolling
resistance, and a large majority of trucks now use radial tires Bias-ply
tires are limited to some rough terrain applications (construction) or in
garbage haulers where there is potential for sidewall damage. Low profile
radials were the next improvement available in tires, and currently have about
35 percent of the market. Low profile radials have 8 to 10 percent lower
rolling resistance than conventional radials, and reduce the operating height
of the truck. Low profile radials also have lower weight, and hence a tax
advantage as tires are taxed by weight. These factors suggest that low
profile radials will essentially displace conventional radial tires over the
next 5 years.
4-15
-------
Additional advances in rolling resistance reduction will come about from
improved rubber formulations, new tire cord materials and the development of
new tread designs. Based on confidential data received from the manufactur-
ers, Navistar estimated that rolling resistance reduction of 20 to 25 percent:
was possible for tractor tires in the Class 8B long haul trucks For trailer
tires, or non-traction tires, they estimate a reduction in the range of 10
percent. The 10 percent figure was also estimated as likely for medium duty
truck applications. Some tires have already shown significantly greater
reductions already, but these tires have compromised durability and/or
handling properties. For example, Continental sold a special tire in the
European market that had a 28 percent reduction in rolling resistance co-
efficient in comparison to a similarly sized standard tire, but had to
discontinue sales due to problems with heat buildup. Future improvements are,
however, not expected to impact durability.
The tread depth and number of wheels also impact tire rolling resistance.
Tire rolling resistance has been found to vary inversely with tread depth;
many special fuel economy prototypes often used 'shaved' tires to maximize
fuel economy. Hence, special snow tires or traction tires tend to have
significantly higher rolling resistance compared to standard tires. Volvo
also found that the increasing number of wheels increases total tire rolling
resistance. Data for a 78,000 lb GVW tractor-trailer, shown in Figure 4-2,
suggests that increasing the number of wheels from 18 to 26 decreased fuel
economy by one percent. The popular Class SB tractor-trailer typically uses
18 tires , and some tire manufacturers have sought to replace the 4 tires on
each of 4 axles with 2 tires, called 'super single' tires, with significant
improvements in fuel economy. However this has not been popular since a tire
blowout causes an unacceptable loss in payload capacity and driveability In
Europe, where three axle trailers are more common, the super single tire has
received greater acceptance. Manufacturers do not foresee substantial
increases in market penetration for super single tires to 2001 in the U.S..
4-16
-------
Figure k-2
fitted
5.80-
5.60-
5.40-
5.20-
5.00
Long haul
Central Europe
18
22
Number of wh««l«
26
lNl-tt-M &S.13 BY TlkOlt
•« VOLVO
-------
4.6 IMPROVEMENTS TO THE ENGINE
As noted in Section 4.1, improvements to the diesel engine can be accomplished
by improving thermodynamic efficiency and turbocharger efficiency or by
decreasing friction and pumping loss. The net effect of these improvements is
to reduce the brake specific fuel consumption, or bsfc. While bsfc can be
measured at specific load/speed points, EEA has utilized bsfc as measured
over the EPA transient test cycle, which is designed to replicate engine
speeds and loads over a city and highway driving schedule. As a matter of
interest, the engine manufacturers association (EMA) provided estimates of
bsfc by engine type to the year 2002. Table 4-4 reproduces the EMA
submission.
It is instructive to compare the actual bsfc attained in 1991 under the
prevalent: 5.0 g/BHP-hr N0X standard with EMA's projections for 1992 EEA
obtained detailed data from several select engine manufacturers and the data
is tabulated in Table 4-5. The bsfc of engines is a function of both Che
horsepower rating and the rated RPM of the engines Increased horsepower
ratings result in lower bsfc as frictional and pumping losses become a smaller
fraction of total output. Lowering the rated RPM results in decreases in
friction loss. It can be seen from Table 4-5 that most manufacturer's
products are similar, although some very modern designs such as Caterpillar's
311b and 3176 models appear to have lower bsfc relative to older engines at
similar HP/RPM ratings It also appear that the EMA projections for 1992 are
very close to the actual values attained in 1991, with the medium-heavy-duty
engines currently at about 0 420 lb/BHP-hr compared to a predicted 0.413
value, and the heavy-heavy at about 0.365 lb/BHP-hr compared to a predicted
0.353. The light-duty engines are split between the older ID1 design Navistar
with high bsfc relative to the more modern, turbocharged/aftercooled DI engine
from Cummins With the expected conversion of Navistar's 7.3 L to DI, the EMA
bsfc projection of 0.466 lb/BHP-hr could easily be attained.
Given th.e relative accuracy of the EMA projection, it is interesting to
examine the projection to 2002. Between 1992 and 2002, no improvement is
4-18
-------
TABLE 4-4
ENGINE MANUFACTURERS ASSOCIATION
BRAKE SPECIFIC FUEL CONSUMPTION
HEAVY-DUTY DIESEL ENGINES
(AS MEASURED ON TRANSIENT TEST CYCLE)
Truck NOx
Class Standard 1977 1982 1987 1992 1997 2002
IIB 10 7 - .527 .524 .451 .451 .451
thru 7.0 - .527 .524 .453 .453 .453
IV 5 0 - .527 524 466 465 .464
V 10.7
thru 7.0
VIII A 5.0
.467 457 .412
.482* .472* .418*
.516" .507* 454*
.382 373 .367
.386 377 370
.413 .401 .398
10.7
VIII B 7 0
5.0
.429 .400 .343
.438* .407* .348*
.491* .432* .374*
.321 310 .308
330 .319 316
.353 .340 .336
* Based, on less than total production.
Source- EMA, 1983
4-19
-------
TABLE 4-5
BSFC FOR SELECTED
1991 ENGINES (LB/BHP-HR)
Class
Model
HP
RPM
BSFC
Light Heavy
Navistar 7.3L
185
2800
0 550
Cummins 5 1L
100
2500
0.400
Medium Heavy
Ford 6.6L
170
2300
0 440
Navistar DTA 360
185
2700
0 445
Caterpillar 3116
185
2600
0 396
Ford 7.8L
210
2300
0.404
Navistar DTA 466
230
2400
0.418
Heavy-Heavy
DDC Series 60 (11.1L)
275
1800
0.372
(<350 HP)
DDC 68-92
300
1800
0.448
Cummins L10
310
1800
0.378
Caterpillar 3176
325
1800
0 344
Heavy-Heavy
Caterpillar 3400 PEEC
460
1900
0 358
(>350 HP)
DDC Series 60 (12.7L)
450
2100
0.367
4-20
-------
forecast for the light-heavy duty class of trucks, a 3.6 percent for the
medium-duty class and a 4.8 percent improvement for the heavy-heavy-duty
class. Our analysis, described below, shows that the EMA projections are
relatively reasonable
As noted, engine bsfc improves with increasing the brake mean effective
pressure (bmep) of the engine since friction and pumping loss do not increase
in proportion to bmep, while output does. Some of the advantage of the newer,
small displacement engines is due to Che increased BMEP. For example.
Caterpillar's 3176 rated at 325 HP has a peak torque BMEP of 300 psi which is
10 to 15 percent more than the BMEP of the typical 14 litre engine, and it
displays one of the lowest bsfc ratings of all engines for which data was
obtained..
There aire obvious structural and durability limits to the increases possible
in BMEP and Caterpillar believed that the 3176 had reached the limits of the
high BMEF strategy. Volvo and Mercedes were in general agreement that this
strategy produced some fuel economy benefit, but Cummins did not subscribe to
this view, as it believed its 14 litre and 10 litre engines had near identical
bsfc at the same rated HP.
Current thermodynamic efficiency of diesel engines is about 40% at the optimum
bsfc point. Volvo suggested the following improvements were passible to
current engines:
• improved charge air cooling,
• electronic control of engine timing,
• improved air utilization,
• adjustments to compression ratio, and
• variable geometry turbocharging
Air-to-air charge cooling is now (1991) available in virtually all heavy-duty
diesels except in the light-heavy category and further improvements to change
air cooling will result in very small benefits to fuel economy, probably less
4-21
-------
than 0.5 percent. Between 1987 and 1991, however, a very large fraction of
the medium-heavy-duty engine segment began utilizing intercooling with a fuel
economy benefit of 5 percent.
Electronic control of fuel injection allows optimization of injection timing
and "shaping" of the "torque rise" curve. Some of the 1991 engines have
electronic injection timing control offered as an option, such a Caterpillar's
PEEC model, while it is standard on some engines such as the Detroit Diesel
Series 60. Comparison of bsfc data from mechanically controlled and
electronically controlled engines do not consistently support a fuel economy
advantage but the effect of the controls can be to change driving practices
and allow the incorporation of speed control, which are discussed in Section
4.9 and 4.10.
Improvements to air utilization can be achieved by moving the piston top ring
closer to the upper edge of the piston and by optimizing valve lift and
duration. At this point, no manufacturer is considering variable valve
timing, but it is a possibility only for the post-2000 time frame. Improved
air utilization will help in reducing particulate emissions, but is expected
to provide very little benefit to fuel economy.
Compression ratio (CR) increases are possible for the future, according to
some manufacturers. Caterpillar and Volvo believed that CR could increase by
0.5 to 0.8 over the next decade, providing a 1/2 percent increase in bsfc.
Cummins and Mercedes believed that N0X emissions would limit further CR
increases and no benefit would result.
Improvements to turbochargers can arise from better matching of turbochargers
as well a.s variable geometry turbocharging. Most of today's engines select
"off-the-shelf" turbochargers that compromise peak efficiency and matching of
characteristics over the load/RPM range. Variable geometry turbochargers have
the capability to improve the matching characteristics over the operating RPM
range, and improvements to turbine and compressor efficiency could add 1.5 to
2 percent in fuel efficiency at operating points that are not close to the
4-22
-------
peak efficiency point. The net effect over the transient cycle is estimated
to be 1/2 to 1 percent benefit in bsfc. In urban driving conditions where the
engine spends much of its time at conditions where current turbochargers are
inefficient, the full 1.5 to 2 percent benefit may be realized.
Hence, the net benefit from improved thermodynamic efficiency is quite small
The most optimistic estimates place the benefit at about 3 percent while the
least optimistic at less than 1 percent.
Friction reduction in diesel engines is expected to occur by:
• improved component design,
• reduction of oil and cooling water flow, and
• reduction of governed RPM.
Current levels of friction in diesel engines relative to engines output is
already quite low. Mercedes provided data showing that a turbocharged diesel
engine rated at 285 HP had a total loss from friction, pumping and accessory
loads of 70 HP. Of this total, only about 30 HP was in friction loss. Hence,
the complete elimination of all friction, a practical impossibility, would
result in output increasing to 305 HP, and bsfc decreasing by 10 percent at
full load. However, the particular values cited are for a low BMEP engine
(135 psi). At a BMEP of 200 psi, complete elimination of friction will not
increase bsfc by even 7 percent at full load. However, at part load, the
friction loss (which stays nearly constant in absolute terms) becomes a much
larger percent of output, and friction reduction has a bigger effect on fuel
economy. An illustrative effect of friction of engine efficiency is shown in
Figure 4-3 for the Mercedes engine discussed above.
Actual friction reduction and internal loss reduction potential is quite
small. Mercedes has been experimenting with a special naturally aspirated
engine using a 2-ring piston, special fuel efficient oils, a low flow water
pump and a low flow oil pump, which reduced friction and accessory power loss
4-23
-------
Figure 4-3
Effect of Reducing Friction on
Engine Efficiency
BMEP bar
-------
by 5.2 HP (or about 10 percent). Some of the changes are inappropriate for
turbocharged engines, such as the 2-ring piston. Other friction reduction
technologies include roller-cam followers, that are widely used in most
domestic engines, but can be adopted for some medium duty engines, and
improvements to the fit and shape of the piston and cylinder liner by improved
manufacturing technologies. Most of the manufacturers interviewed for this
study suggested that 5 percent reduction in friction was likely by 2001, while
a 10 percent change was the highest conceivable limit of friction reduction.
A 5 percent reduction in friction translates into a 0.75 to 1 percent fuel
economy benefit averaged over a range of load/speed conditions
Friction increases rapidly with engine RPM, and 10 percent decrease in RPM
will bring about a proportionally larger reduction in friction, as the
dependence is non-linear. Most heavy-heavy duty engines already operate in
the 1800-1900 RPM range, although models rated to 2100 RPM are still avail-
able. Models rated at 1600 RPM as special "economy" models have been avail-
able since the early 1980's but have not been very popular since the drive-
ability of these engines saves fuel both by friction reduction and by limiting
truck speed. The new electronic controls in most engines make the 1600 RPM
engine redundant, as these new engines shape the torque curve to permit the
driver to cruise at 1500 RPM, while retaining the benefits of an 1800 RPM
engine during acceleration or in city traffic.
In the medium-duty applications, ratings at 2800 RPM or 2600 RPM were
traditionally popular, but new models rated at 2300 RPM or 2400 RPM have
become available in the last few years. These models have higher torque than
their 2600/2800 RPM counterparts, and require transmissions and axles with
higher ratings. The lack of availability and increased cost of transmissions
and axles has held back the market penetration of these lower RPM rated
engines Electronic controls have not yet been adopted, and are likely to
increase penetration in the future Reductions in the peak RPM to 2300/2400
RPM from 2600 RPM can result in fuel economy increases of 3 to 4 percent with
no change in vehicle speed.
4-25
-------
Pumping loss is not a large fraction of total output for diesel engines,
Mercedes estimated pumping loss to be equivalent to half of friction loss at
full load. At a given RPM, engine absolute pumping loss is not a strong
function of load, and therefore is a larger factor at light loads, much like
friction. Pumping loss can be reduced by controlling airflow through the
engine to prevent "excess" air due to turbocharger mismatch at different
speeds and loads. In addition, the use of pulse-tuned intake manifolds and
tuned exhausts can reduce pumping loss at specific load/speed points. These
technologies provide very small benefits in fuel economy, and have been widely
adopted in most heavy-heavy engines. Variable volume intake and exhaust
systems have not been considered for heavy-duty diesels, due to their poor
cost/benefit.
The use of 4-valve heads reduces the pressure drop across the valve orifice,
and is common in many of domestic heavy-heavy engines However, most of the
engines rated below 270 HP, and most imports of all horsepower ratings, use
2-valve engines. Conversion to 4-valve aids in reducing the pumping loss, and
also in increasing the bmep of a specific engine. BMEP increases are,
however, accounted for separately in the analysis.
Reductions in pumping loss in total over the next 10 years are likely to be
less than 5 percent in medium-heavy duty engines. This translates into a half
percent increase in fuel economy
The net benefit from all improvements to conventional diesel engines will vary
by engine type, as the heavy-duty diesels (used in Class 8B trucks) already
incorporate technology that the lighter engines do not. A summary of the
potential gains is provided in Table 4-6. In 1991, virtually all of the
heavy-heavy-duty engines feature air-to-air intercooling and some already are
of the high BMEP design As shown in Table 4-6, a 7.5 percent improvement to
bsfc is forecast over the period 1987-2001 relative to a 1987 baseline, at
constant emission standards. The imposition of the 5 0 g/BHP-hr standard in
4-26
-------
TABLE 4-6
IMPROVEMENTS TO ENGINE
BSFC, 1987-2001
Light-Heavy
Medium-Heavy
Heavy-Heavy
Conversion to DI
80
12.0
0
0
0
0
10% Governed speed
reduction
80
4.0
50
2.5
20
1.0
High BMEP design
0
0
50
1.5
80
2.4
Improved turbocharger
matching
80
8.0*
100
0.5
100
0.5
Air-to-Air
intercooling
100
5.0
80
4.0
30
1.5
Improved Thermodynamic
efficiency
100
1 0
100
1.0
100
1.0
Friction reduction
100
1.5
100
1.0
100
1.0
Pumping loss reduction
100
1.0
100
0.5
20
0.1
Total bsfc Benefit
32 5
11.0
7.5
Effect of 5.0 N0X Std.
0
-2.0
-2 0
Effect of 1998 NOx Std.
-3.0 to -5.0
•3.0 to -5.0
-3 0 to -5.0
Conversion from naturally aspirated.
4-27
-------
1991 reduced fuel economy by about 2.0 percent, approximately counterbalancing
the gain achieved through the introduction of air-to-air intercoolers. Hence,
net gain of 5.5 percent is forecast at the 5.0 g/BHP-hr NO,, standard between
1987 and 2001. However, the 1998 N0X standard of 4.0 g/BHP-hr could impose
significant bsfc reduction, and it is possible that the loss could be as large
as 5.0 percent, negating all of the technology benefits. This forecast
closely corresponds to the opinions of most manufacturers of heavy-duty
diesels. Starting from a 1987 bsfc of 0.365, the bsfc at a 5 0 NO, standard
will be 0.346 lb/BHP-hr in 2001.
Larger improvements will be available for the medium and light diesels. In
1987, most medium duty diesels were not intercooled, but the new emissions
standards for 1991/94 have forced a large majority of these engines to adopt
air-to-air intercooling. In addition, the adoption of lower governed speeds
will result in more engines moving to the 2300-2400 RPM ratings from the
current 2500-2600 RPM ratings, with some engines moving to 2100 RPM A net
bsfc reduction of 9.0 percent relative to 1987, at a 5.0 N0X standard is
forecast. This implies that bsfc will decline from 0.425 lb/BHP-hr in 1987 to
0.390 lb/BHP-hr in 2001.
The largest gains in fuel efficiency will occur for the light-heavy engines
that are used only in a few Class 5/6 trucks and school buses. This largely
stems from the conversion of the IDI diesels in this segment to Direct
Injection (DI) Currently, these engines are mostly indirectly injected V-8
designs rated at 2800 to 3000 RPM. By 2001, it is anticipated that they will
be replaced by turbocharged and aftercooled, direct injection diesels, with a
32.5 percent decrease in bsfc for the segment as a whole Starting from an
average bsfc of 0.540 lb/BHP-hr, the average bsfc will decline to 0.408
lb/BHP-hr. This forecast is the only one that differs significantly from
the 1983 EMA forecast; yet, the presence of the new Cummins BT5.9 engine
with a bsfc of 0 400 lb/BHP-hr in this class suggests that the forecast is
conservative.
4-28
-------
At this point, it is difficult to estimate the effect of the 4.0 g/BHP-hr N0X
emissions standard required by the revised Clean Air Act for 1998 and beyond.
Many diesel engine manufacturers believe that further NO,. reduction from
current levels would be very difficult to achieve without timing retard.
Timing retard would result in significant reduction in fuel economy of 3 to 5
percent. However, Navistar was an exception, and its staff stated that fuel
economy effects of the 4.0 N0X standard would be relatively small. Manufac-
turers need not use any timing retard if several new technologies under
investigation prove successful. Among these technologies, the new zeolite
catalysts appear promising in that tests with diesels have shown N0X reduc-
tion sufficient to meet even a 3.0 g/BHP-hr standard. Of course, these test
results are preliminary and the durability of these catalysts is unknown.
Nevertheless, there is potential ta meet Che 4 .0 g/BHP-hr t?Qx standard with no
fuel economy penalty. Other technologies, such as incorporation of exhaust
gas recirculation, reduced compression ratio and/or variable valve timing may
be used singly or in combination to attain the 4 0 g/BHP-hr standard with less
fuel economy penalty than if the standard was obtained by timing retard alone.
4,7 1URB0C0MFOTIND DIESEL ENGINES
Turbocompound dLesel engines were extensively researched by the Department of
Energy (DOE) in the late 1970's and early 1980's. Despite a successful
technology development program that suggested significant fuel efficiency
benefit, the technology was not commercialized. The DOE-Cummins joint
development program that was essentially complete in 1982 involved the
assembly of a 450 HP turbocompound diesel that met the 6.0 g/BHP-hr NOx
standard . The turbocompound engine provided a fuel consumption reduction of
15 to 16 percent over a production 1982 NTC-400 horsepower engine used as a
reference. However, the turbocompound engine used a number of component
refinements, and Cummins determined that the benefit of turbocompounding alone
was 4.2 to 5.3 percent. Additional advances in the exhaust manifold design,
arid insulation of the exhaust flow path resulted in an additional 6 percent
improvement over the initial turbocompound design. A minimum bsfc of 0,298
lb/BHP-hr was attained at 1500 RPW.
4-29
-------
At meetings with manufacturers held in conjunction with this analysis, most
engineers contested the large values of fuel economy benefit initially
reported by Cummins. Mercedes stated that their tests had shown a 2 percent
benefit In fuel efficiency with heat insulation of the exhaust manifolds and
ports. Caterpillar was more optimistic, and cited recent development showing
a 4.7 percent lower specific fuel consumption at rated speed and a 3.3 percent
benefit at peak torque RPM. Cummins engineers also stated that a 4 percent
benefit for turbocompounding may be representative over a typical driving
cycle.
Turbocompounding has other benefits. Since the turbine obtains power from
waste heat, the engine emissions per unit of useful work are decreased
Typically, the turbine output is in the range of 10 to 12 percent of recipro-
cator output. If absolute engine-out emissions stay constant with and without
turbocompounding, a proportional 10 to 12 percent reduction in brake specific
emissions is implied. Indeed, Caterpillar found that at 4 0 g/EHP-hr NOx, the
turbocompound engine had 8.0 percent lower bsfc at rated RPM and 3 5 percent
lower bsfc at peak torque RPM.
It has also been suggested that heat insulation of the cylinder would be
particularly useful with turbocompounding. Proponents of ceramic components
have discussed the heat insulation of the cylinder head, piston top, and
cylinder liners as a means to recover the heat wasted to the coolant
Performance assessments of ceramic components for low heat rejection engines
completed in the 1985-1988 time frame suggested additional fuel efficiency
benefits of 3 to 4 percent at full load, and 'up to 13 percent' at part load
for high swivel engines. Tests conducted by heavy-duty diesel engine manufac-
turer ha-ve failed to produce such benefits. In fact, most of the manufactur-
ers interviews had very negative perceptions of ceramics for use as heat
insulation for the cylinder. Mercedes stated that the very high temperatures
of the ceramics had the following negative effects-
• decreased volumetric efficiency,
• increased N0X emissions,
4-30
-------
• potential increase in brake specific particulate emissions due
to lowered intake airflow,
• and no significant reduction in heat transfer.
While Cummins was in general agreement with Mercedes and saw no benefit to
ceramic "hot internal parts" at a 4.0 g/BHP-hr standard, Caterpillar was
notably more optimistic. Caterpillar suggested that a ceramic cylinder head
was a distinct possibility even without turbocompounding, and that with
turbocotnpounding, ceramics could increase the total bsfc benefit from 5 to 7
or 8 percent at rated speed.
All manufacturers stated that some turbocompound engines would be introduced
in 1994/1995 but this technology would not have high market penetration by
2001.
4.8 DRIVETRAIN OPTIMIZATION
The drivetrain parameters are selected for a given truck gross weight and
engine combination to meet a variety of performance requirements, such as fuel
economy, 'on-grade' startability, capability to negotiate a grade at a
selected speed and vehicle top speed requirements. Once a customer has
selected an engine with a specific peak torque and RPM rating, the transmis-
sion and drive axles must be selected to optimize among the various require-
ments. The selection is based on the power/torque rating of their compo-
nents, tine ratio coverage of the transmission, the number of gears, and the
axle ratio. Historically, the choice was sometimes constrained by availabili-
ty of transmissions, and parameters were selected to optimize overall "perfor-
mance", even at some slight loss of fuel efficiency.
In Class 8B trucks, inefficient drivetrain parameters were chosen by fleet
operators to limit driver top speed rather than match for best efficiency. At
highway cruising speed on level roads (e.g., 55 to 60 mph), it is most fuel
efficient to operate the engine closer to peak torque RPM, which is typically
60 percent of rated RPM. However, selection of gear and axle ratios to
achieve this RPM results in a truck with a capability to exceed 70 mph at
4-31
-------
rated RPM. Many fleet operators specify a numerically high axle ratio to
prevent overspeeding. A more fuel efficient axle ratio selection with
additional road speed control can provide 3 to 5 percent improvement at the
same reference speed. This improvement is possible for the portion of the
feel where, currently, axle ratios are misspecifled. Anecdotal information
suggests that about 30 to 50 percent of Class 8B trucks may have unoptimized
drive - trains,
In medium-duty trucks, greater scope exists for drivetrain optimization, since
the choices of gear ratios, gearbox torque capability and axles were supply
limited. For example, virtually no transmissions were available with a rating
between 650 ft-lbs of torque and 1000 ft-lbs. This situation has been
changing with the introduction of transmissions from Spicer and Fuller with
ratings of 750 ft-lb and 975 ft-lbs, for example. As a result, the medium
duty truck customer can choose the low RPM, high torque engine models without
paying a very large cost penalty for transmissions and driveshafts with
ratings substantially higher than required. However, Che fuel economy benefit
associated with using lower governed speed engines is accounted for in
Section 4.7
Other improvements to the transmission include the incorporation of single-
plate rather than double plate clutches. New cerametallic materials allow
single plate clutches to effectively replace double plate clutches, resulting
in a small (0 5 percent) increase in city cycle fuel economy More recently,
Eaton has introduced a new automated 9-speed transmission called 'Econoshift'
that automates the shifting in the top two gears so that the correct gear will
be chosen automatically at speeds above U5 mph. Eaton stated that the
Econoshift would cost $1500 more than a traditional transmission but pay for
itself in 18 months, suggesting a fuel economy improvement as large as 5
percent in Class 8B trucks. It is not clear how Eaton arrived at this
benefit, as the benefit would depend on the reference baseline.
Increased levels of market penetration are forecasted for automatic transmis-
sions in. the lightest medium-duty trucks and bus markets. Although the
4-32
-------
automatic transmission can be less efficient Chan a manual transmission that
is shifted correctly, actual driver behavior may make the automatic transmis-
sion more efficient In certain applications.
Drivetrain component suppliers are also integrating their axle and transmis-
sion/clutch offering, partly in response to a narrowing of the supplier base
for each manufacturer. In the past, consumers could specify clutch, transmis-
sion and axles from different suppliers with some loss of configuration
optimization. The bundling of those components could result in an improvement
in drivetrain optimization.
It is difficult to estimate the benefits of drivetrain optimization, as the
current extent of misspecification is not well understood. In addition the
benefit of some improvements is dependent on driver behavior that is common
now, not on what it will be in the future. We have relied on manufacturer
opinions to estimate the fleet average benefits by truck class and city-
highway cycle, as follows, in terms of percent improvements to MPG
City Highway
Class 6/7 2 0 2 0
Class 8A 1.5 1.5
Class SB 1.5 2.0
These benefits do not include any benefit associated with low RPM engines for
Class 6/7/8A trucks
4.9 ELECTRONIC CONTROL
Electronic controls are being widely incorporated into trucks, partly in
response to new emission standards that are easier to meet with these
controls . In 1987, very few trucks had any electronic controls, but in 1990
industry experts stated that about 20 percent of Class 8B trucks and 5 percent
of medium duty (Class 6/7/8A) trucks had electronic control As noted,
electronic controls of fuel injection does not improve the bsfc of the engine
4-33
-------
on the EPA test cycle significantly. However, the principal benefits of the
electronic systems are associated with:
• shaping of the torque curve to allow drivers to operate at
lower RPM in top gear,
• gear shift indications,
• road speed governor functions,
• engine shutdown and protection systems,
• extended idle shutoff,
• driver monitoring,
• and engine/vehicle diagnostics.
An example of such an electronic system is Mack Trucks' VMAC computer control
system. The system is optional at a cost of $3000, and is available for
Mack's Class 8B trucks.
The benefits of the system are dependent on the baseline. If a driver
maintains speeds at or below legal limits and selects the appropriate gear for
cruise, the system provides benefits only from torque curve shaping. Larger
benefits will be obtained relative to the average driver who may be
overspeeding on the highway. Proper shifting during city driving can also
save fuel, and shutting the engine off rather than subjecting it to extended
idle will also save fuel. These benefits are difficult to estimate as there
are no detailed analysis of current inefficiencies in driver behavior that are
publicly available. Estimates of fleet average impacts are largely based on
anecdotal evidence, and not though any actual analysis.
Benefits were estimated by industry experts for electronic injection timing
shaping the torque rise of the engine, speed control to 65 mph, shift control
and extended idle shutdown. Total benefits are as follows:
Class 6/7/8A
Class 8B
City
3 0
2.5
Highway
5.0
6 0
4-34
-------
These benefits are percentage increases in MPG for the fleet as a whole due to
adoption of electronic control.
4.10 OTHER IMPROVEMENTS
Small improvements in engine and drivetrain related components will occur due
to evolutionary changes in design, leading to small improvements in fuel
economy. The use of synthetic lubricants in the engine, gearbox and axle lead
to small decreases in drivetrain friction, that also can benefit fuel economy.
The largest power drain from accessories comes form the cooling fan. Thermo-
statically activated cooling fans operate only when required by the engine and
can improve fuel economy significantly; however, most new trucks already
incorporated this device by 1978. Modest improvements can be made to the air
compressor, water pump and power steering hydraulic pump For example, Eaton
has recently unveiled a variable assist power steering system that cuts the
pump's power absorption by up to 50 percent. Currently all of the accessory
loads take up 7 to 10 percent of total engine power output, and a slightly
larger share of fuel consumption While gains from improvements to existing
accessories will improve fuel economy in the range of 0.5 to 1.0 percent,
there is a tendency to increase the level of accessories to improve driver
comfort and safety These considerations can lead to an increase in accessory
loads from demands for air suspension, antilock brakes and other interior
power equipment. As a result, net accessory power decrease is unlikely to
decrease and could actually increase over the next decade
Synthetic lubricants have been available for over 10 years but have not
achieved significant market penetration. Lubricant manufacturer sponsored
testings showed significant benefits in fuel economy from their use - up to 4
percent - but truck industry experts suggest that 1 to 2 percent benefits are
more appropriate Improvements to conventional lubricants have also occurred
through the use of friction modifications and viscosity under improvements.
Such improved non-synthetic oils are more likely to find widespread acceptance
due to their lower price. A fuel economy gain if 1 percent in city driving
4-35
-------
and 0.5 percent on the highway is expected to result from lubricant
improvements to 2001.
4.11 TOTAL IMPROVEMENT IN FUEL EFFICIENCY
In order to predict the total impact of all technological improvements to fuel
economy, it is necessary to obtain a penetration weighted estimate of each
technology's contribution to fleet fuel economy. If, between 1987 and 2001,
the market penetration of technology i increases by mi, and rA is the percent
fuel economy gain associated with the application of that technology to an
individual truck, then 11^.^ is the percent gain in fuel economy for the fleet
as whole. The total fuel economy improvement is'
T - I xL
i
This assumes that there are no favorable or unfavorable synergies across all
technologies. The equations governing fuel consumption demonstrate the fact
that some fuel economy improvements are additive, and others are
multiplicative so that the above equation is conservative.
Table 4-7 summarizes the derivation of total fuel economy benefit for Class 8B
trucks. Since the sensitivities for city and highway driving are so
different:, the benefit over each cycle is computed separately. For Class 8B
trucks, the largest benefits on the highway come from the combination of
aerodynamic improvement, and the use of electronic road speed governors/high
torque rise engines. Net benefits m city driving are much smaller since
aerodynamic benefits are negligible and road speed governors have no effect at
these speeds Nevertheless, in the absence of a 1988 N0X standard, the net
fuel economy improvement for these trucks is 22 percent (assuming a 70/30
highway/city weighting) If the N0X standards penalty is as high as 3 to 5
percent, the net increase to 2001 will be only 17.5 percent, corresponding to
1 2 percent increase per year that is nearly identical to the historical
experience between 1978 and 1987.
Table 4-8 shows that same calculations for Class 6/7 medium-duty trucks
4-36
-------
TABLE 4-7
IMPROVEMENTS TO CLASS 8B
TRUCKS, 1987 - 2001
(Percent Improvements in Fuel Economy)
Technolo gv Market City F/E Highway F/E
Penetration Increase
Weight Reduction (0.75X)
100
0.40
0.
.25
Drag Reduction Van Bodies
50
0.28
6
75
Other
35
0.07
1
60
Rolling Resistance
65
1.85
2
90
Engine Improvements
100
5.50
5.
.50
Turbocompound
10
0.30
0
50
Drivetrain Optimization
N/A
1.50
2.
,00
Electronic Control
100
2.50
6
00
Lubricants
100
1.00
0
50
13 00
26
00
Potential Effect of N0X Standard
(-3.00)
(-5.
.00)
4-37
-------
TABLE 4-8
IMPROVEMENTS TO CLASS 6/7
TRUCKS, 1987 - 2001
Technology
Weight Reduction
Drag Reduction: Van Body-
Other
Rolling Resistance
Engine Improvements*
Drivetrain Optimization
Electronic Control
Lubricants
Market
Penetration Increase
100
45
55
45
100
100
100
100
City F/E Highway F/E
0.35
0.15
0.07
0.90
12.00
2 00
3.00
1.00
0.15
3.35
1.69
1 35
12.00
2.00
5 00
0 5
Potential Effect of N0X Standards.
19 47
(-2 5)
26 04
(-4 00)
Weighted for light-heavy diesels in Class 6
4-38
-------
The highway MPG improvement is as large as the one calculated for Class 8B
trucks, but the source of the improvement is quite different; in this case,
engine improvements are the main contributor. These engine improvements are
also available in the city cycle, and the net benefit at city speeds is also
quite large. The city/highway composite improvement is 21.h- percent, using a
weighting of 70 percent city/30 percent highway, without any emission penalty
and 18.5 percent if the penalty is as large as anticipated without any
breakthroughs in technology. This translates to an annual rate of growth of
1.3 percent, which is again in good agreement with the historic rate of growth
experienced in the 1978-1987 time frame.
In conclusion, it appears that technology is available to continue the
historic growth rates or even exceed them slightly with no government
intervention in the markets.
4-39
-------
Appendix A
Statistical Results of the
Regression Models to Estimate GPM
-------
MODEL 1
Dependent Variable*. "LGFM
EEA CVW Groo|>" 19 4 iQO~ 26 ,000 lha Typ« ot \ts»ad by vthUlt-Ga&oUne
Analysis of Variance
Sourc*
Modal
Error
C Total
OF
4
4133
4i3;
Root MSE
Dap Wean
C V.
5LCD of
Sijuiireo
24 94020
400 493/8
423 42397
Haan
Square
6 23)03
0 09690
0 31129
-1 £4409
16 83020
R-aquorfl
AdJ R-oq
F Valua
(4 346
0 0586
0 0577
rrob>F
0 0001
Pararaatar Eatlmataa
Parameter
Standard
T tor HO
Vatlalla
DF
Estimata
Error
Par«met«r-0
Prob > JT|
INTEHCEF
1
-1 360115a
0
016*4615
-107 M»®
0 0000
1«L I OX
X
-0 00492&
0
00193149
-2 552
0 0106
LWTVCT1
1
0 234 590-
0
024 7410B
9 462
0 0001
LCIDCTR
1
0 440347
0
03900437
11 266
0 0001
DLOCAL
1
0 050265
0
01325301
3. 793
0 0002
Correlation of EaLJnstaa
CORR5
IHIERCEP
MDLJDX
LHTWCTl
LC1DC1R
DLOCAL
IMERCEP
1 0000
-0 4103
0 5077
-0 0409
-0 6663
MDLIDK
-D 41B3
1 oooo
-D 1040
-0 2122
0 1 158
IMTHG Tl
0 5877
-0 1040
1 0000
-0 1311
0 0250
LCIDCTR
-0 0409
-0 2J22
-0 1311
1 0000
0 0649
DLOCAL
-0 6063
0 1150
0 0250
0 0649
1 0000
-------
MODEL 2
Dependent Variable: LGPH
EEA GVW Gxcup-19,50Q-26.000 Lba Typ« of fual uaad fay v«hlcla-01••aL
Analysis of Variant:*
Soxirco
ModaL
Error
C Total
Root H3E
Dap Haan
C V
Sun of
DF
Squares
5
2* 00105
2006
13* 36*55
2011
15a 36)59
a
25881 1
-l
99331 1
-12
98381
Ha an
Square
4 &0021
0 Q64S6
R-icpiar*
hdj R»q
P Value
71 665
153*
1494
Prober
0 0001
Faramattr Estlnuiaa
>
I
ro
Variable DF
IHTERCEP
MDLIDX
LVfTWGTI
LHPCTR
DLOCAi-
DAERO
Psrunattr
Eailuta
76^384
027829
220667
096602
033743
069142
Standard
Error
0160 509*
00174202
03*2?740
0512199)
01224845
02273169
T lor R0
Par«B«tar*0
-98 908
-15 975
6 444
1 666
2 750
-3 042
Prob > (r|
0.0000
0 0001
0 0001
0 0594
0 0060
0 0024
CORRfl
IHTERCEP
M)UDX
lwtvcti
LHPCTR
DLOCAL
DAERO
IHTERCEP
1 0000
-0 6522
0 5693
-fl 0283
-0 5189
-0 0076
HDLIOK
-0 6522
1 0000
-0 0620
0 1331
0 1532
-0 1641
lotion of Ea
LWTVCT1
0 1693
-o aezo
1 oood
0 0176
0 0207
0 0090
LHPCTR
-0 02 BJ
Q 1331
0 0176
1 0000
0 0106
-0 9453
D LOCAL
-0 5169
0 1531
0 0207
0 0106
1 0000
0 0286
DAERO
-0 0076
-0 16*1
0 0090
-0 0t53
0 0286
1 0000
-------
MODEL 3
Dependent Variable: LGPM
EEA GVW Croup-26 000-33,000 Lbo Typa of faal uoad bjr veh ic La-Caao Una
Analysis of Vftrlenc*
Sourca
Mod«l
Error
C Total
DF
4
2012
2019
Sum ot
Squares
12 781/2
1&1 6 8*901
194 won
Mean
Square
3 19543
0 09017
F Value
35 439
Pcob>F
0 0001
Root USE
Pep Hean
C V
0 3OO20
-I 7889S
16 7852?
R-»quar«
AdJ Biq
0 0657
0 D639
P-aromet-ar Estimates
>
I
Parameter
Standard
T for HD
Variable DP
Eatlaate
Erroc
PaxamatacO
Pcob > Ir I
IHTERCEP 1
-I 698965
0
02334247
-72 166
0 0000
MELIDX 1
-0 009167
0
00233633
-J 614
0 0003
LKTHGT1 1
0 196238
0
03159836
6 212
0 0001
LC1DC1R 1
0 555*994
0
06024826
9 228
0 0001
~LOCAL 1
0 018047
0
01933636
0 933
0 3508
CORRfl
IHTERCEP
MDL1DX
LWTWGT1
LC1DCTK
OLOCAL
IHTERCEP
1 0000
*0 4401
0 5451
-0 0227
-0 7205
Correlation
MDLIDX
-0 4401
1 0000
-0 1326
-0 1493
0 1339
of Eatlnataa
LWTWGT1
0 5451
•0 1326
1 0000
-0 1237
0 0369
LCIDCTR
-0 0227
-0 1493
-0 1237
1 0000
0 0202
DLOCAL
-0 7205
0 1339
0 0369
0 0202
1 0000
-------
MODEL 4
Dependent Variable: LCPM
EEA GVW Grcup-26,OOO-33,DDO lbs Iypa of fu«l ujed by vahlcla-Dlaael
Axialyala of
Variance
Sua of
Him
Source
DF
Squares
Square
Modal
5
28 90*99
3 78100
Error
1930
129 29563
0 06631
C Total
1935
138.2008*
Root USE
0
23750 R-
aquara
Dap Haul
-1
93881 AdJ K-«q
C V
13
28129
F Valua
67 187
Frob>F
0 0001
0 1B27
0 1806
>
Variable DF
INTERCEP
IOLIDX
LHTMGT1
LHFCTH
DLOCAL
DAEAO
Parameter Eatlmetea
Parameter
Eatloate
743511
03002}
233573
093002
0(346*
0*1537
Standard
Error
01716069
00169163
0313882*
03222*23
01246*07
02270766
T for B0
Faraneter-0
-101 599
-17 749
7 441
1 781
3 092
-1 623
Prob > |T|
0 0000
0 0001
0 0001
0 0731
0 0001
0.0683
Correlation of Ettinalia
CORRB
INTEHCEP
tDLIDX
LVTTHGT1
LHFCTR
DLOCAL
DAEKO
IHTERCEF
1 0000
-0 6271
0 5381
-0 0177
-0 53*2
-0 1007
HDLIDX
-0 6271
1 0000
-0 0712
0 1390
0 1153
-0.1037
LWTVCT1
0 3381
-0 0712
1 oooo
0 0174
0 0337
-0 0320
LHFCTR
-0 0377
0 1390
0 017*
1 0000
0 0229
-0 0158
DLOCAL
-0 33*2
0 1155
0 0337
0 0229
1.0000
0 0940
DAEBO
-0 1007
-0 1037
-0 0320
-0 0156
0.09*0
1 0000
-------
MODEL 5
Dependent Variable: LGPH
EEA GVH Grouj>**331 ODO~60, 000 Iba Typo ot CuaL uatd by vahici.**OaDoltKifi
AiialyiU of Variance
Sourc*
Mod* I
Error
c rotoi
Sims et
Ma an
Dr
Squares
Square
F Valua
Proh>F
4
6 33916
1 56479
17 240
0 0001
1639
150 66161
a <19L9i
1643
157 00096
Root KSr
Dip Haon
C V
0 303J9
-1 60008
-18 94626
R*«qu«r#
AdJ R-iq
0 0*04
0 0360
FaiaiMttr EsLlaut**
Vatlebl* OF
IHTERCEP
KH.IDX
LHTMGT1
LCIDCTR
DL0CA1
Puaaitir
Standard
T for E0
Frob » |t|
fUlMt*
Crxor
P«r«Ml*rH3
-1 518136
0
033(3690
-«5 403
0 0001
-0 010200
0
00398049
-2 562
0 010J
0 149245
0
03673395
3 853
a oooi
0 572693
0
09650945
5 616
o oooi
0.041550
0
02234609
1 659
0 0&31
>
U1
COKftB
1HTEHCEP
tfiLIM
LHTKJTI
LCIDCTR
~LOCAL
INTERCEP
1
-0
a
-o
-0
COOC
3066
7354
1225
5635
CoirtULion ot EiLlnBtiB
H>UCX LWHGT1
-0 3066
1 0000
-D 049a
0 0B39
o otas
0 735%
-0 049B
1 0000
-0 1352
0 0433
LCIDCTR
-0 1223
Q.0B39
'0 1532
i aooa
-D a093
DLOCAL
-0.3H3
0 Q6B3
0 0453
-0 <1092
1 0000
-------
MODEL 6
Dependent Variable: LXJFH
EEA GVW Croup-33,llOO-fcO(OOI> Lba Typ* o( EihI usH by v«hlcl«"Dkt»tl
Analyala of Vari«nc«
Sun &t Mian
SoutC*
DF
Ewjuara* Squat*
F V.lua
pmb>r
Hadml
5
33 SB* *2 6 J9&0-9
113.101
0 oooi
Zz c.oc
C Total
7775
300 58702
Root USE
0
2*505 R-aquar«
0 0679
0«p H««n
1
7080 7 Adj Ruq
0 0673
c.v
It
34692
Paranatal Eitlnitai
It-
er.
Varlabla DF
INTERCEP
KH.1DK
LWTWST1
LHPCTH
DI.GCA1.
DAERO
Paramatar
Eatlnata
-1 602002
-0 011171
0 174341
0 063006
0 030B64
-0 057297
Standard
Error
0 00831730
0 00068880
0 01223369
0 00966089
0 00611715
0 0096*660
T for HO-
Paramatar*0
-188 0BB
-12 568
14 223
6 369
a 316
-3 94Q
Prcb > |T|
0 0000
0 0001
0 0001
0 0001
D 0001
a oooi
Correlation of EaClaaLaa
CORRB
(HIERCEP
mi-ID*
LKIHCIl
LilPCIR
DLOCAL
daeho
1HTERCEP
1 0000
-0 470(1
0 6604
-0 0830
-0 4481
-0 1543
K3L1DX
-0 4700
1 .0000
-0 0623
0 1124
0 1173
-0 2312
LWIHGT1
0 6684
-0 0625
1 0000
-0 0361
0 0979
-0 0626
LHPCTH
-0 0630
0 1124
-0 0301
1 0000
0 1673
-0 0317
DLOCAL
-0.4461
0 1173
0 0979
0.1673
1 0000
0.1960
DADtO
0 1343
0 2312
0 0626
0 0317
0 1960
1 0000
-------
MODEL 7
Dependent Variable: LGpm
EEA CVW Group-60,000-80.000 Lb* Typ« of fu
cla-Plaaal
Sourca
Modal
Ittot
C ToLal
or
&
W01J
17081
Root &6E
Dap Hun
C V
AnaLjrsia of Variance
Sub of
Squaraa
SO 93)86
ittl «81>2
432 49388
Maan
Squara
8*8931
0 0223)
0 14949
'1 63480
-9 14399
R-aquara
Ad J H-aq
Faraaatar Eattutaa
F Valua
379 903
0 1171
0 1173
Frob>F
0 0000
>
I
Par«m«tor
Standard
T tor HO
Piob > |t|
Variable DF
Catlcata
Error
Paraaatar-0
inrEKCEF ¦
-1
0
00)133)2
-412 «09
0 0009
MjLIDE 1
-0 011201
0
0003(188
-JO 9J3
o ooo i
UHTWGT1 1
0 ««U13
0
00373790
J 402
0.0001
LKPCTR 1
0 >11380
0
CO 723494
13 381
o.ooai
HjOCU. 1
0 »I54«9
0>
•02)3001
8 »?]
d.oobi
DA£RO 1
-0 01BD8B
o
0til26»Z4
-4 :jj
a com
DECO 1
-0 027305
0
00473327
-i 811
0 0001
Correlation of Eitimtaa
CQ8HB
IHTERCEF
K>L1DX
LWTVtGTl
U1PCTB
D LOCAL
DAEHO
DEDO
intehcep
1 0000
-0 6330
0 6619
-0 0342
-0 2244
-0 0881
0 0439
W)UDH
-0 6330
1 0000
-0 1001
-0 0331
0 1613
-0 1792
-0 0907
LHTXCT1
0 6619
-0 1001
1 0000
-0 1030
0 1803
0 0624
0 0106
LHKTtH
-Q 0542
-O 033\
-O
1 MM
ft 1011
0 0JS2
0.0137
D LOCAL
-0 2244
0 1613
0 1803
0 1011
1 0000
0.0812
0 007]
DA£RO
-0 0881
-0 1792
-0 0624
0 0392
0 0012
1 0000
-0.7230
DECO
O 0459
-0 0967
0 0106
0 0157
0 0073
-0.7230
1.0000
-------
M3DEL 8
Dependent Variable LGPM
EEA GVW Group-Over 80,000 Lbs Typo of fuel used by vahlcLe~Dleael USETYPE"!
AnaLysis of Variance
Sum ol
Source
DF
Squares
Square
F Value
Prob>F
Model
4
1 20 <.81
0 30120
10 398
0 0001
Error
551
15 96073
0 02897
C TotaL
555
17 1655*
Root MSE
0
17020 R-square
0 0702
Dep Mean
-1
53* 73 Uj
R-sq
0 063*
C V
-11
08966
Parameter Estimates
Parameter
Standard
T for HO
Variabla OF
Estimate
Ecror
Paramatar-0
Prob > |T|
IHTERCEP 1
-1 *31978
0
02528792
-56 627
0 0001
MDLIDX 1
-0 009221
0
00216618
-* 257
0 0001
LWTVCT1 1
0 0616*8
0
02673900
2 306
0 0215
LHPCTR 1
0 1*8885
0
0*555618
3 268
0 0011
DECO 1
-0 05906*
0
02693986
-2 192
0 0288
Correlation of Estimates
CORRfi
IHTERCEP
MDLIDX
LWTWGT1
LhPCTR
OECO
INTERCEP
1 0000
-0 5083
0 6567
0 0091
0 0312
MDLIDX
-0 5083
1 0000
-0 1071
-0 0817
-0 2626
LWTVJGT1
0 8567
-0 1071
1 0000
-0 0781
0.0076
LHPCTR
0 0091
>0 0617
-0 0761
1 0000
-0 0312
DECO
0 0312
-0 2626
0 0076
-0 0312
1 0000
------- |