EV Demand Forecasting Model Based on Travel Survey: a Brazilian Case Study

Global Electric Vehicle (EV) stock is expanding rapidly. As EV penetration grows, the power demand by EV for charging will increase, indicating a potential reshaping of load profile on the distribution network and likely impacts on the distribution transformers. This study forecasts EV load in a large residential network using specific probabilistic modelling approach. The EV load is estimated using light duty vehicles usage pattern based on a trip travel survey analysis. The trip survey is categorized into trip frequency, trip purpose, distance driven, trip time and duration of travel, which are all treated as distribution of probabilities. In addition to the trip survey analysis, the probabilistic model considers the EV type, model, battery capacity, state of charge, on-board charger level and fuel consumption to forecast energy demand by EV. The performance of the distribution transformers in response to the EV behavior are further investigated from the comparison of the proposed EV demand forecasting model with a generic Poisson model. The results show that, depending on EV penetration level, the distribution transformer can be overload, especially for EV participation higher than 50%.