Stochastic modelling of fast DC charging stations with shared power modules

Bayram, I. Safak and Sevdari, Kristian (2024) Stochastic modelling of fast DC charging stations with shared power modules. In: 2nd IEEE International Conference on Renewable Energies and Smart Technologies 2024, 2024-06-27 - 2024-06-28, University of Prishtina.

[thumbnail of Bayram-Sevdari-2024-Stochastic-modelling-of-fast-DC-charging-stations-with-shared-power-modules]
Preview
Text. Filename: Bayram-Sevdari-2024-Stochastic-modelling-of-fast-DC-charging-stations-with-shared-power-modules.pdf
Accepted Author Manuscript
License: Strathprints license 1.0

Download (1MB)| Preview

Abstract

Fast DC charging stations are becoming increasingly necessary for wider electric vehicle uptake. In standard DC chargers, each charging unit has its own charging power and cannot be shared with another electric vehicle. Depending on the electric vehicle type, the maximum DC charging power varies (e.g.50 kW for small sedans and $\geq 100$ kW for SUVs) in parallel to battery chemistry and capacity. If the charger power is higher than the maximum charging capacity of an EV, then, charging resources are wasted for other vehicles which can accept high charging currents. On the other hand, recent advances in power electronics enable centralized inverters to supply power to multiple DC chargers and shift the load between them dynamically. To that end, we propose a stochastic model for a fast charging station in which the charging power modules are centrally located and electric vehicles are connected via external charging sockets. The charging station serves multi-class customers based on their charging power, random arrival and service durations. The system is modelled with a multi-rate Erlang loss system and a methodology to calculate the probability of meeting customer demand is presented. Case studies are presented to provide insights on how the station performs under varying station settings.