A Probabilistic Framework for the Estimation of EV Hosting Capacity of a Low Voltage Distribution Network Considering Spatio-Temporal Uncertainty

The forecasted uptake of Electric Vehicles (EV), if materialized, may create significant challenges in power system planning and operation, particularly in the low and medium voltage distribution network. Uncoordinated EV charging demands can significantly stress the distribution network to maintain the system voltage limits and distribution system capacity.Therefore, the EV hosting capacityof a network is required. This paper presents a framework to estimate a probabilistic EV hosting capacity of a distribution network considering the inherent (spatio-temporal) uncertainty in the location and timing of the EV charging. Measurement-based EV charging profiles have been developed from millions of Australian (the state of Victoria) EV charging events. These EV charging profiles have been distributed among 449 nodes of the distribution network (by considering the spatio-temporal uncertainty) to estimate the probabilistic EV hosting capacity. An index has been demonstrated to indicate the risk of voltage violation and the thermal limits of the distribution transformers. While the EV hosting capacity varies from feeder to feeder and hour to hour, the network problems would start to emerge at around 30% hosting capacity.