Predictive control and sizing of energy storage to mitigate wind power intermittency

This paper studies how the control algorithm impacts the required capacity of battery energy storage system (BESS) to mitigate wind intermittency. We study a futuristic scenario in which wind generation is traded in the energy market on an hourly basis, and the wind power producer has to procure reserves to handle wind fluctuations. The wind power producer can avoid paying for conventional reserves by charging/discharging BESS to compensate for wind surplus and deficit. We develop control algorithms using the model predictive control (MPC) methodology, which incorporates wind forecasts for the next few hours when determining wind scheduling. The MPC algorithm is developed by solving a non-linear optimization problem to minimize operation costs to the wind power producer. In addition to operation costs, the MPC algorithm considers two practical aspects: the efficiency loss of BESS and the smoothness in wind power scheduling. BESS sizing is studied by parametric analyses. Simulations show that BESS is more effective in reducing operation costs and reducing wind curtailment than conventional reserves. In addition, MPC is a horizon-based control algorithm and can preview future information in its control action. Simulations also show that MPC consistently outperforms an instantaneous heuristic algorithm that does not use future information. Therefore, we confirm that MPC can reduce the BESS capacity required to cover wind uncertainties. Copyright (c) 2015 John Wiley & Sons, Ltd.