Aggregating load shifting potentials of electric vehicles for energy system models

This paper presents and classifies different methodologies for aggregating the load shifting potentials of several plug-in electric vehicles (EVs) for consideration in energy system models. These approaches range from modeling single charging events to approximation of EV fleets. We conclude that most of the applied approaches show shortages and that their suitability depends on the number of time steps of the energy system model and the EV fleet size. For distributed systems and limited time span, modeling single EVs is suitable and can provide exact results when used in optimization models. For large EV fleets and centralized optimization approaches, approximation methods seem to be appropriate. Aggregating the flexibility of multiple EVs to polytopes exactly describes the solution space for the optimization but is an NP-hard task. Consequently, approximation methods to this so-called Minkowski sum are presented, which may improve the consideration of EVs’ flexibility in energy system models.