Charging, steady-state SoC and energy storage distributions for EV fleets

A recent worldwide uptake of electric vehicles (EVs) has led to an increasing interest for the EV charging situation. A proper understanding of the former is required to understand charging needs and to dimension the corresponding infrastructure. In the paper, we develop models that allow us to approximate the steady-state distribution of State-of-Charge (SoC) levels for EVs at the beginning of the day and infer its dependence regarding the daily relative range, defined as the ratio of mean daily-driven distance to the maximum range. The framework combines: (i) a generic parametric model for decision to charge in terms of the SoC and ; and (ii) a simulation model in which we utilize longitudinal log-data for a fleet of cars to track charging events and SoC over time. The model brings about several interesting use cases, of which two stand out. Firstly, it provides a simple parametric way of circumventing transient behaviour in SoC distributions when applied to simulation frameworks. Secondly, it offers a clear method to infer crucial information regarding EV fleets and the total energy storage potential. Such information is useful for vehicle-to-grid (V2G) applications in that it provides expected lower and upper bounds for the energy that can be stored and charged. The model is applied in a Danish context, and it is suggested that while a full electrification of the vehicle fleet could lead to significant stress on the power grid, it will at the same time hold a large potential for deploying V2G as an important mechanism to dampen the effect of temporal high-demand. The potential for V2G stems from a low battery utilization between charging events of approximately 40%, which in turn provides a large storage buffer that could be harnessed with little to no impact on EV utilization. Results also indicate that tapping into just half of the available buffer could serve the EV demand and also meet close to 50% of the household consumption.