A systematic methodology for mid-and-long term electric vehicle charging load forecasting: The case study of Shenzhen, China

•A novel classification method for the EV fleet is proposed to predict the total load.•A modified Bass model is presented to predict the ownership of private EVs.•The regression model is used for the ownership prediction of other types of EVs.•The probability theory is adopted to model the charging profiles of EVs. More and more adoptions of electric vehicles (EVs) would bring a potential threat on the existing electric grid. In this context, a systematic methodology is presented in this paper to predict the additional loads resulting from EV charging in the mid-and-long term. It includes probabilistic models for describing the EV charging profiles and forecast models for predicting the future EV ownership. It is impractical to develop a method to simulate the charging profiles of the entire EV fleet due to the diversity of EV charging behaviors. As a consequence, the entire EV fleet is divided into four categories viz. private EV, electric taxi, electric bus and official EV so as to predict their charging loads respectively. The proposed method is conducted in the city of Shenzhen, which currently has the largest electric bus and electric taxi fleet in the world. Results indicate that the maximum value of the predicted EV charging profile in 2025 would occur at 21:30, reaching 1,760 MW under high oil price, which could elevate the existing load peak by 11.08 %.