Methodology for Deriving Vehicle Activity Parameters from Travel Survey Databases

Abstract

Vehicle activity parameters are key inputs to mobile source emissions modeling. Emission factors are sensitive to these inputs, which vary by geography, time of day, and trip purpose classifications. The state of the practice in emission factor modeling relies heavily on default parameters imbedded in the models. Therefore, there is a need for deriving these inputs for the various geographic regions to provide realistic values. Methods are described for deriving vehicle activity parameters from travel surveys, such as starts distribution, fraction of vehicle miles traveled (VMT), trip length distribution, soak time distributions, and the VMT mix distribution. This methodology is applied for deriving vehicle activity parameters from Nationwide Personal Transportation Survey databases and conducting a sensitivity analysis of the MOBILE6 model with respect to these derived inputs. Inputs derived from travel surveys are compared with MOBILE6 default inputs; wide variations in the inputs are observed. A sensitivity analysis of MOBILE6 related to travel survey inputs reveals that emission factors can vary by as much as 26% when compared with those obtained with default inputs. A statistical analysis has determined that the difference in emission factors is significant enough to warrant the development of locale-specific inputs.