Uncertainty of available range in explaining the charging choice behavior of BEV users

Available range (AR) uncertainty is prevalent in battery electric vehicles (BEVs). AR depends on traffic conditions, weather conditions, road conditions, driving style, etc. This study aims to examine the role of AR uncertainty in BEV users' en-route charging and charging route choice behavior. A stated preference survey is designed and conducted to explore the two consecutive choice behaviors: (1) en-route charging and (2) charging route choice. Four model specifications for en-route charging and three for charging route choice are defined. AR uncertainty is, for the first time, considered in explaining en-route charging choice behavior, in addition to the conventional choice attributes such as travel time, charging duration, initial AR, average AR, etc. Socio-demographic and vehicle-related factors are also considered in the behavioral modeling. Furthermore, random parameters logit with error components model is utilized to capture the full panel effects. Results indicate that AR uncertainty significantly affects users' en-route charging choice behavior, thus being indispensable in behavioral modeling. In addition, average AR can modify the asso-ciation between AR uncertainty and choice decision. Demand for en-route charging will be overestimated when AR uncertainty is not considered. Furthermore, there is considerable inter-respondentand intra-respondent heterogeneity in the en-route charging choice. The effect of AR uncertainty is stronger for female and high-income BEV users. The findings will contribute substantially to better en-route charging demand estimation and optimal deployment of public charging stations, particularly for medium-to long-distance trips.

Automated summary

This study examines the impact of available range (AR) uncertainty on battery electric vehicle (BEV) users' en-route charging and charging route choice behavior. The findings indicate that AR uncertainty significantly affects users' en-route charging choice behavior and should not be overlooked in behavioral modeling. The average AR can modify the association between AR uncertainty and choice decision. The results have practical implications for estimating en-route charging demand and optimizing the deployment of public charging stations, particularly for medium-to long-distance trips.