A layered architecture for EV charging stations based on time scale decomposition

We present a layered decomposition approach that permits a holistic solution to the storage, scheduling and pricing problems of Electric Vehicle (EV) Charging Stations. By exploiting time scales, these problems can be decomposed and solved layer by layer. In the top layer, at a long time scale, with grid power price and renewable energy represented by their long-term averages, and total demand following the price-demand curve, the optimal pricing scheme is obtained. The real-time charging and discharging operation of the battery, is considered in the middle layer. With average number of customers arriving determined by the price set at the top layer, the middle layer determines the optimal amounts of energy to buy from the grid and to use for charging. At the bottom layer, the scheduling policy of EV charging is determined while satisfying the total battery consumption obtained at the middle layer. We illustrate the algorithms with a simple example using ERCOT data, demonstrating the implementability of the architectural solution in real-time market operation of an EV charging station. Simulations show that the architectural decomposition does not incur any significant cost penalty.