One-way car-sharing (OWCS) systems although offer flexibility to their users over two-way systems, they introduce significant operational complexities. In electric OWCS, a vehicle without sufficient battery level cannot be assigned to a trip. A regular electric vehicle battery charger requires 6 hours to fully charge an empty battery. To ease the operations, electric OWCS companies are adopting fast chargers. Using fast chargers, a vehicle can be fully charged within 30 minutes of time. Given the higher cost of fast chargers and the associated benefits associated with their use (higher vehicle utilization, reduction of the number of costly vehicle relocations et.), it is important to investigate what is the optimum number and location of fast charging stations.
We are proposing an Integer Programming model for determining the optimal number and location of fast chargers needed. The model considers the demand for trips, relocation activities, battery requirements of trips, and battery levels of vehicles. Since the model is computationally intractable, we consider a clustering approach to generate approximate solutions. The proposed model and algorithms are tested on a car-sharing system operating in Nice.