Asymmetric Logistic Model for Estimation of Mileage-Related Vehicle Depreciation Function of Roadway Characteristics

Abstract

This paper describes an approach for the development of prediction models for the estimation of mileage-related vehicle depreciation that can be used in the estimation of the benefits derived from transportation network improvements. The approach takes advantage of published online data for vehicle valuations. A new asymmetric logistic prediction model for total vehicle depreciation, including initial and mileage-related depreciations, is proposed and fitted to collected valuations data. The added benefit of this prediction model is that it takes into consideration both vehicle age (i.e., years since manufacture) and vehicle usage (i.e., miles of travel). Six small light-duty vehicles (SLDVs), five large light-duty vehicles (LLDVs), three two-axle trucks, one single-unit truck, and two combination trucks were considered in this study. Vehicle fuel sources included gasoline, diesel, gasoline-ethanol blend of up to 85% ethanol (E85), and hybrid-electric, resulting in 26 combinations of vehicle type and fuel source. Additionally, the developed models were adjusted to account for the effects of average speed of vehicle and roadway characteristics (e.g., grade, curvature) on vehicle depreciation. The practicality of the developed models for large sport utility vehicles (SUVs) and midsize cars was illustrated using select examples highlighting the models’ sensitivity to vehicle average speed and roadway characteristics.