Route and grade sensitive modeling of fuel efficiency and emissions for diesel buses

The study first analyzed the characteristics of twenty-three chassis duty cycles to determine the most important parameters that affected FC and emissions. The analysis showed that cycle properties such as standard deviation of average speed, percentage of idle, characteristic acceleration, kinetic intensity, and stops per unit distance were closely related with cycle average speed. A cycle with low average speed contained stop-and-go behavior, with multiple accelerations and decelerations, whereas a high vehicle speed cycle was steadier. In this way, average speed was found to be inter-related to other cycle properties.;The experimental data were collected on a subset of twenty-three test cycles. Average speeds of these cycles were found to have substantial effect on distance-specific emissions of some species and FC from the thirteen buses. Distance-specific oxides of nitrogen (NOx) and FC demonstrated a good correlation with average speed. The highest NOx emissions and FC were observed on the slowest speed cycle, the New York Bus Cycle. The highest speed cycle, the Commuter Cycle, exhibited the lowest NOx and had the lowest FC. Hydrocarbon (HC) emissions from diesel and hybrid-electric buses and carbon monoxide (CO) and particulate matter (PM) from all these buses were too low to permit reliable determination of a trend.;A prediction model was developed to project the ratio of FC of an empty passenger bus and to FC of a full passenger bus. This predictive model was then compared with emissions and FC data collected on three drive cycles from 40-foot diesel and natural gas buses at empty passenger weight, half passenger weight, and full passenger weight. The experimental data showed that a 28% increase in weight yielded about 12% increase in FC but no significant increase in NOx emissions from natural gas buses, while for the diesel bus a 32% increase in weight yielded about a 19% increase in FC and about a 14% increase in NOx emissions, on average. However, changes in NOx emissions for corresponding small change in passenger weight did not follow any clear trend. CO, HC, and PM from these buses also did not follow any trend with passenger weight.;Weight effect was followed by analyzing the effect of road grade and terrain on FC of a transit bus. Two inputs, the type of terrain and the percentage of grade, needed to be determined in order to compute the increase in FC for grade. While it was comparatively simple to determine the terrain type, it was complex to determine the road grade for a bus route. This problem was compounded by the fact that all bus routes started and ended at the same point so that no grade could be ascertained. This model addressed this issue by assigning a sinusoidal road grade with a specified maximum positive or negative grade.;These analyses were combined to construct a predictive model for FC for a 40-foot diesel bus for a particular route, where the average speed was determined by an empirical relationship involving cycle duration, maximum speed, and number of stops per unit distance. The average speed was then used to calculate FC from the relationship developed in speed analysis, followed by determination of the weight correction factor for full as well as no passenger weight buses. Finally, a grade correction factor was applied to complete the FC model. The predictive FC model was expected to provide an insight into the planning and selection of bus technologies for route choice. It was noted that this model was developed for 40-foot diesel buses. Therefore, development of FC models for 40-foot CNG and hybrid-electric buses and 60-foot articulated buses of all technologies were recommended for cost-effective procurement of bus technologies and efficient bus scheduling. (Abstract shortened by UMI.)