Investigating the characteristics of public electric vehicle charging station: A case study of California

This study aims to explore the characteristics of public electric vehicle (EV) charging stations as a California case study using multiple data sources. This is because EV charging stations differ in terms of charging capacity and the surrounding built environment. We first classify public EV charging stations based on their neighborhood built environments, access to amenities, and charging capacity using a clustering algorithm, Latent Profile Analysis (LPA). We then employ a fractional logit model (FLM) to analyze factors influencing EV adoption rates across communities, focusing on varying impact of different public EV charging stations. Some key findings are as follows. First, we identify five clusters of public EV charging stations as follows: 1) CBD level 2 (Business level 2), Residential level 2 and DC fast charging, 3) Coastal level 2 (Tourist area level 2), 4) Urban level 2, and 5) Suburban and Rural DC fast charging. Second, LPA results demonstrate that current EV charging stations are primarily concentrated in residential and urban areas, mainly consisting of level 2 chargers. Third, our empirical model results show a significantly positive correlation between residential level 2 and DC fast charging and the proportion of EVs in nearby communities. Lastly, Costal level 2 (Tourist area level 2) is also positively associated with the ratio of EVs in surrounding communities.