Optimal location of EV charging stations in a neighborhood considering a multi-objective approach

•Presents a methodology to locate semi-fast charging stations in distribution system.•Includes constraints of CS service zones delimited by a hierarchical clustering.•Considers the distribution system and the EV's demand in a multi-objective approach.•Evaluate CS location for mid and long-term planning based on the CS occurrence rate.•Presents a real case distribution network including both MV and LV networks. Despite the environmental and economic benefits of Electric Vehicles (EVs), distribution network operators will need to understand the location where the charging infrastructure will be placed to ensure EV users’ needs are met. In this sense, this work proposes a methodology to define the optimal location of EV semi-fast charging stations (CS) at a neighborhood level, through a multi-objective approach. It applies a hierarchical clustering method to define CS service zones, considering both technical and mobility aspects. Besides, it considers uncertainties related to the EV load profile to determine the CS capacity, based on the user's charging behavior. A Pareto Frontier method is deployed to support the decision-making process on CS optimal location, considering utility and EV users’ preferences. The results indicate that the best CS locations for mid-term EV penetration can also fit into long-term planning, with higher EV charging demand. Thus, these locations would be good candidates for the power utility to make initial investments, regarding both planning horizons. A real distribution system case is used to demonstrate the applicability of the results.