Energy arbitrage optimization of lithium-ion battery considering short-term revenue and long-term battery life loss

Abstract

To achieve the ambitious goal of net-zero by 2050, the power sector in many countries has been increasingly using renewable energy sources (RES). The power system balance will be significantly challenged due to the intermittency of RES. Lithium-ion battery energy storage systems (BESS) can effectively address the challenge by providing flexibility to fill the mismatch between the intermittent supply and the varying demand, generating revenues for BESS to payback the initial investment. However, every grid participation also results in an inevitable lifetime loss. From the asset owner’s perspective, therefore, it is important to balance the short-term revenue from the fluctuating electrical market with the long-term battery life loss. To achieve this in this study, first, a novel online-applied battery life model is established, which is computationally efficient to comprehensively consider the battery’s historical aging, state of charge (SOC), charge–discharge rate (Crate). Then, an online-applied sliding-window dynamic programming (SWDP) strategy is proposed to optimize short-term grid service revenues and long-term battery life losses. The SWDP is also compared with an offline-applied dynamic programming (DP) strategy. The simulation results show that the cost due to battery life decay is significant, which indicates the importance of tracking the battery life loss in the battery asset management. Additionally, the prediction accuracy of electricity price significantly influences the performance of SWDP strategy. The BESS’s total profit can reduce by half in the considered case study if the electricity prediction error is 12.5%.