Class-8 Heavy Truck Duty Cycle Project Final Report - Center for ...

three rows of Table 21 show the average (same as the ones presented in Table 19), standard deviation
and sample size of each of the distributions. Notice that the minimum sample size is 850 observations
(Duals-NGSWBTs) with a maximum of 1,956 observations (100-mile segments) for the NGSWBTs-
Duals case. The fourth row in Table 21 presents the mean of the difference between the sample
average fuel efficiency of the truck- tire combination of the corresponding column and that of the
Duals-Duals case. The fifth row shows the standard deviation of the difference which, together with
the mean and sample size, permits the computation of the test statistic value z (a one-tailed test based
on the alternative hypothesis shown in Eq. 12 and provided in row six of Table 21). For the three
truck- tire configurations that have some or all NGSWBTs, the null hypothesis, which states that
there is no difference in terms of fuel efficiency that can be explained by different tire configurations,
can be rejected with 100% confidence in favor of the alternative hypothesis, which indicates that on
average the fuel efficiencies obtained when using the Duals-Duals combination is lower than in the
other cases. This strongly suggests that fuel efficiencies in cases using some or all NGSWBTs are
higher than in the case of a truck equipped with all dual tires. In other words, the percent
improvements in fuel efficiencies that are achieved when using all NGSWBTs or combinations of
duals and NGSWBTs are statistically significant. And as demonstrated in the last sub-section, even if
the databus does not measure fuel consumption with total precision, as long as the information
collected from the six databuses present the same systematic bias, then the percent differences are still
accurate,.
Table 21. Fuel Efficiency Comparison by Truck-Tire Configuration (Null Hypothesis: All Fuel Efficiency
Distributions Are the Same)
Sample Average Fuel Efficiency 1 (mpg)
Duals-
Duals
Truck-Tire Configuration
Duals-
NGSWBTs
NGSWBTs-
Duals
NGSWBTs-
NGSWBTs
6.60 7.00 7.03 7.22
Sample Std. Dev. 1.21 1.32 1.19 1.19
Sample Size 1,798 850 1,956 871
Mean of Difference with Duals-Duals 0.40 0.42 0.61
Std. Dev. of Difference with Duals-Duals 0.05 0.04 0.05
Test Statistic Value (Z) 7.41 10.77 12.43
Reject Ho at Confidence Level =
1
Fuel efficiencies computed using collected databus information.
6.3.2.2 Effect of Vehicle Weight on Fuel Efficiency
100.00% 100.00% 100.00%
The extensive database and powerful search engine created for this study allows researchers to study
other aspects of fuel efficiency., For example, studies involving the effects that different tractor/trailer
tire combinations have on the fuel efficiency of Class-8 trucks, when additional variables are taken
into account, would be of interest. One very important variable affecting fuel efficiency is the weight
of the vehicle.
The previous sub-section showed that there were improvements of the fuel efficiency of the types of
vehicles studied here when at least some of the tires (either those mounted on the tractor, the trailer,
or both) were NGSWBTs. Those results were obtained without controlling for any particular variable
(other than the type of tires) since it could be assumed that any of the participating vehicles would
experience the same situations (terrain, weather, load) over the course of the year-long experiment. It
is, however, important to determine whether the observed increases in fuel efficiency due to the type
of tires used is maintained once the data is controlled for vehicle weight. That is, it is important to
determine if the gains on fuel efficiency due to the type of tires used are more significant when the
91