Optimal sizing and operation of community hybrid energy storage systems

Configuring a community energy storage system (CESS) helps balance energy supply-demand and increase the self-consumption rate of distributed renewable energy based generation on the user side. However, a CESS primarily relies on battery storage, whose high economic cost makes large-scale development and practical application difficult. Given this background, the optimal sizing and operational strategy for a community hybrid energy storage system (CHESS) is proposed in this paper, which comprises the slow-response energy storage device (SRESD) and the fast-response energy storage device (FRESD). Firstly, considering the community's willingness to deploy storage and the impact of storage on smoothing load fluctuations, a multi-objective optimization model aiming to maximize community profits and minimize load fluctuations is proposed, which is transformed into a single objective optimization by incorporating a penalty factor. Then, the Density-Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm is employed to derive typical community load and electricity price curves which are used to schedule the monthly charge and discharge periods for the SRESD, subsequently developing the optimal sizing and annual 8760-hour operational strategy for a CHESS based on this schedule. On this basis, three storage configurations (CHESS, SRESD, and FRESD) are compared, and it is found by simulation results that the CHESS combines the advantages of both SRESD and FRESD, offering larger arbitrage opportunity with the peak-valley price spread and being more effective in managing load fluctuations. Finally, the impacts of declining battery costs on optimization results are examined and experimental verification based on real data from Australia is carried out, with the feasibility and sustainability of the proposed sizing and operational strategy for a CHESS demonstrated. • A hybrid community energy storage system balances load and improves community profits. • Clustering methods determine the system's annual 8760-hour operation strategy. • Simulations show cost advantages and better use of peak-valley electricity prices. • Real-world data confirms the system's effectiveness and long-term sustainability.