Robust Sizing of Solar-Powered Charging Station with Co-located Energy Storage

This paper studies optimal sizing of an electric vehicle charging station with multiple types of chargers. The station is coupled with a photovoltaic system and stationary battery to reduce the overall cost of delivering charging services. Our approach entails (a) generating sample paths from a stochastic process that represents the state of the charging station given the minimum number of chargers of each type needed to meet some quality of service requirements, (b) solving an optimization problem for each sample path to find the smallest battery capacity and number of solar panels that meet specific requirements, and (c) determining the robust size of co-located distributed energy systems by finding an upper envelope on the sizing curves obtained for different sample paths. Using real data, we find the least expensive size of the entire system that is robust to seasonality effects, travel patterns, and future increase in the penetration of electric vehicles.