Modeling EV charging behavior with hybrid choice models

Smart charging programs adjust the time of day of electric vehicle charging to reduce congestion in the grid, reduce the cost of electricity, and maximize renewable energy use. As General Motors partners with electric utilities to design and implement these programs, it is important to understand customers’ preferences and motivations for enrollment. A discrete choice experiment was conducted to quantify how customers trade off among program characteristics and incentives when deciding whether to enroll in demand response, managed charging, and fixed schedule programs. The data were used to estimate two hybrid choice models, which outperform a benchmark conditional logit model and systematically account for unobserved preference heterogeneity using discrete–continuous mixtures. The results indicate that monetary incentives and environmental benefits increase the likelihood of choosing a smart charging program. Meanwhile, there is notable preference heterogeneity regarding non-monetary program options and enrollment perks. Latent environmental concern was constructed as a useful dimension for differentiating customers: those with higher environmental concern tend to exhibit higher valuations of reducing emissions and maximizing renewable energy use. We further linked preferences to five personas based on sociodemographic clustering. The findings enable targeted marketing efforts that highlight the environmental benefits of smart charging to the customer groups most likely to be environmentally concerned.