1 Old Dominion University
Norfolk, VA 23529
University of Virginia
351 McCormick Dr.
P.O. Box 400742
Charlottesville, VA 22904-4742
Paleti, Rajesh (ODU)
Total Project Costs
Plug-in electric vehicle (EV) sales are on the rise and growing EV adoption can be a key factor in helping regions achieve national- and state-level air quality standards for ozone and particulate matter, and ultimately carbon-emissions standards. In the Mid-Atlantic region, purchasing an EV over a gasoline-powered medium sedan can reduce transportation greenhouse gas emissions by 60 percent . The ability to predict which households in which neighborhoods are most likely to own such vehicles can provide important insights and opportunities for power-grid planning, transportation investments, and air quality policy-making. However, unlike the choice to purchase a traditional gasoline-powered vehicle, the decision to adopt an EV is complicated by technology familiarity, vehicle availability, and charging infrastructure provision. Previous quantitative research forecasting regional or household EV ownership trends tend to neglect one or more of these influence factors. The research proposed here aims to enrich the existing understanding of EV adoption by incorporating a combination of revealed-preference (RP) and statedpreference (SP) data that captures household demographics, vehicle traits, and zone-level infrastructure and land use considerations, all of which influence vehicle choice. The latent choice model proposed here minimizes SP data bias by recognizing differences in individual households’ vehicle choice set while incorporating a spatial analysis component which accounts for “neighbor effects” that are commonly present in early technology diffusion.
Research Goals and Objectives: The goal of this research is to establish a new choice modeling framework to evaluate new technology diffusion, which has been historically challenging due to limited data from early adopters and inherent biases in SP survey responses. The research proposed focuses on utilizing a combination of existing RP data and to-be-collected SP survey data to examine the effects of household demographic, vehicle, and transportation infrastructure characteristics on EV ownership. In this application, the EV adoption predictions will be joined with a county-level emissions model which takes into account regional grid mix, ambient temperature, household vehicle miles traveled (VMT), and driving conditions to anticipate the effect of EV adoption trends on carbon footprint in Virginia. This study will provide valuable planning information for transportation planners and utility providers on anticipating when and where new EV households will likely emerge.
Web Links to Reports and to the Project website
Report: Would you consider a Green Vehicle? ( Submitted: January 2019)