Management of electric vehicles to provide energy flexibility under uncertainty

In recent years, the adoption of Electric Vehicles (EVs) has increased driven by governments initiatives to reduce greenhouse gas emissions. This increased adoption, while environmentally beneficial, has also led to a rise in electricity consumption. This paper analyses the advantages of bidirectional vehicle-to-grid (V2G) chargers for Charging Station Operators (CSOs): Electric Vehicles energy flexibility could be a valuable resource during grid congestion episodes. We analyze this flexibility proposing a two-stage stochastic programming model under uncertainty, regarding the presence of EVs at charging station. We utilize a mixed Long Short-Term Memory (LSTM) neural network to generate scenarios for each charger, then, blend and reduce them into the most representatives. The results demonstrate that electric vehicles can provide demand flexibility, resulting in potential cost reductions of up to 70%, contributing to prevent power grid congestion during periods of low generation or high demand.