A two-level optimization framework for battery energy storage systems to enhance economics and minimize long-term capacity fading

Abstract

This paper proposes a two-level optimization framework for a battery energy storage system to maximize revenue with consideration of the phenomena that cause battery's capacity fading. Instead of solving the scheduling problem as a singular problem, a two-level optimization framework is introduced. The upper-level optimization focuses on maximizing revenue by arbitrage in a real-time electricity market. With the determined operating schedule, the lower-level optimization, solved with particle swarm optimization, determines the optimal charging current that mitigates the side reactions while maintaining economic performance. With the optimal charging current and operating schedule, a pseudo-2D electrochemical model is used to simulate the battery behavior. To reflect the actual behavior of the lithium-ion battery, two side reactions, solid electrolyte interface formation and Lithium plating, are simulated. A case study using California energy prices is presented. With the proposed framework, the results show that the battery's lifespan is extended by up to 5.1 % and overall revenue increases by up to 9.8 % compared to a singular economic optimization utilizing the standard constant current, constant voltage charging protocol.