Multi-objective sizing and real-time scheduling of battery energy storage in energy-sharing community based on reinforcement learning

Abstract

Reducing peak demand and enhancing self-sufficiency have made the battery energy storage system (BESS) essential in microgrids when combined with photovoltaic (PV) systems. Previous studies have shown that BESS usage remains expensive, and it is more practical to share it within a community rather than installing individual systems. Therefore, this study aimed to propose a framework for BESS sizing and scheduling in an energy-sharing community based on reinforcement learning. To validate the proposed framework, a case study was conducted based on the BESS ownership scenarios (i.e., individual-owned BESS (IOB) and community-shared BESS (CSB)) considering the following purposes of potential stakeholders: (i) enhancing SSR; (ii) reducing peak demand; and (iii) increasing economic profit. The maximum SSR of CSB was 12.4% higher than IOB, while the peak load of CSB was 0.07% lower than IOB. As the BESS size increased, the total cost of BESS increased, and the total profits on electricity decreased. The economic profit increased by about 38% when CSB was used compared to IOBs. The proposed framework offers more promise for a broader range of building loads and PV generation within the energy-sharing community. Ultimately, this study would allow prosumers, community practitioners, and policymakers to better understand the sharing mechanisms in the community and provide decision guidelines for ownership types, operational strategies, and sizing of BESS and PV systems.