Life-Aware Operation of Battery Energy Storage in Frequency Regulation

The rapid growth of renewable generation in power systems imposes unprecedented challenges on maintaining power balance in real time. With the continuous decrease of thermal generation capacity, battery energy storage is expected to take part in frequency regulation service. However, accurately following the automatic generation control (AGC) signal leads to more frequent switching between charging and discharging states, which may shorten battery life. Because battery life is a consequence of long-term operation depending on the depth of discharge, it is difficult to model battery health in frequency regulation problems. This paper establishes an online operation policy in response to the real-time AGC signal considering battery health. Based on the empirical relation between cycling number and depth of discharge, a cost function is suggested to approximate the impact of charging-discharging action on battery life in the long run. Then, Lyapunov drift-plus-penalty technique is employed to derive an explicit online policy compromising the revenue of frequency regulation and battery life without requiring future AGC data. The offline operation problem after knowing the entire sequence of AGC signal is cast as a difference-of-convex program and solved by a sequential convexification algorithm. The gap between the offline and online optimum is shown to be a constant, which quantifies the value of knowing future information. Case studies based on AGC data from the PJM market validate the proposed method.