Spatio-Temporal and Correlation Modeling of EV Charging Data for Hosting Capacity Analysis

The growing number of residential Electric vehicles (EVs) are posing a significant challenge for grid operators around the world. This paper presents spatiotemporal (ST) modeling approach to address system uncertainties and copula theory to model correlations of EV charging demand with EV users’ travel parameters (such as EV departure time, trip distance, and EV arrival time) and solar PV generation. Copula theory is employed to represent the correlations between these parameters, validated using the Kolmogorov-Smirnov (KS) test and quantile-quantile (q-q) plots. The case study presents EV hosting capacity in an Australian distribution network, adapting the modeled EV charging demand with real-world measured load demand profiles. The network EV hosting capacity is determined subject to voltage violation risk (VVR), line overloading, and transformer capacity violation. According to the results, the EV hosting capacity was found to be limited to 25% EV penetration due to the transformer capacity limit primarily. However, the EV hosting capacity could reach 50% without violating the VVR limit. The simulation results also indicate that for the chosen test network and load profile, the EV hosting capacity is slightly higher on weekends than on weekdays of the simulation period.