Cost optimization of energy purchase for EV fleets based on a Markovian EV charging model

This work analyses potential energy purchase strategies for an ICT-enabled and active charge management of a large fleet of electric vehicles in order to minimize applicable costs for the purchase of energy at Day-Ahead or Intraday spot markets. The optimization potential for energy purchase is leveraged through a Markovian electric vehicle charging model and on the basis of empirical data for mobility patterns of vehicles as well as actual spot market data. Two scenarios with different charging characteristics of the EV fleet are investigated. In a commuter scenario, where the fleet of EVs charges during daytime (7:00AM-3:30PM), we found that volatility in spot market prices from Q4/2011-Q3/2012 may have allowed for cost optimization of up to 13% compared to entirely unmanaged charging. In a parcel delivery service scenario, the fleet of EVs charges during nighttime (6:00PM-6:00AM), which allows for cost optimization of up to 34% based on the same period for spot market data.