Probabilistic model of electric vehicles’ flexibility in residential areas

• Flexibility is modelled in terms of a potential for rescheduling charging sessions. • Inherent randomness in EV charging patterns is addressed in the developed model. • Main factors which affect EV flexibility are identified and modelled. • Limits of EV flexible capacity are obtained from the model-based calculations. • Developed model is of particular interest for distribution network planning. This paper presents a probabilistic model for the flexibility assessment of electric vehicles in residential areas. The flexibility is modelled and evaluated in terms of the aggregate charging demand which can be rescheduled to prevent exceeding network capacity. Developed model and corresponding calculations are based on the algebra of random variables. Main factors which affect the flexibility of electric vehicles are identified and modelled first. Modelled quantities and their mutual links are then incorporated into the probabilistic model. Cumulative distribution of charging duration, discrete probability of charging on particular day and discrete probabilities of possible start charging time are presented in the paper as the probabilistic model outputs. Finally, limits of available flexible capacity [kW] are calculated and presented as the main flexibility indicator for specified test cases. Presented results illustrate a potential for rescheduling charging sessions in different cases/scenarios. Developed model is expected to be of particular interest for distribution network planning during the 2020s and early 2030s.