Chance-Constrained Optimal Configuration of BESS Considering Uncertain Power Fluctuation and Frequency Deviation Under Contingency

• Published in: IEEE transactions on sustainable energy Vol. 13; no. 4; pp. 2291 - 2303
• Main Authors: Cao, Yongji, Wu, Qiuwei, Zhang, Hengxu, Li, Changgang, Zhang, Xuan
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 01.10.2022; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Battery energy storage system; Benders decomposition; chance-constrained optimization; Configurations; Constraint modelling; Contingency; Energy storage; Fluctuations; Frequency control; Frequency deviation; frequency regulation; frequency stability; Iterative methods; Life cycle costs; Life cycles; Load modeling; Optimization; Optimization models; Power smoothing; Power system dynamics; Power system stability; Renewable energy; system frequency response model; Uncertainty
• ISSN: 1949-3029; 1949-3037
• DOI: 10.1109/TSTE.2022.3192087

With the accelerating integration of variable renewable energies (VREs), power systems become more vulnerable to active power disturbances, and more drastic frequency dynamics emerge. The battery energy storage system (BESS) is able to handle the uncertainties of VREs, and the decreasing system inertia and frequency regulation capability. This paper proposes a chance-constrained optimal configuration scheme for the BESS to maintain both the uncertain power fluctuations and frequency deviation within predefined limits. First, the required frequency regulation capability of the BESS constrained by the maximum transient frequency deviation (MTFD) and quasi-steady-state frequency deviation (QSSFD) is estimated. Then, the kernel density estimation method is utilized to model the net power fluctuations of VREs and load. A multi-objective chance-constrained programming model accounting for the life cycle cost, energy arbitrage, uncertain power fluctuation, MTFD, and QSSFD is established to optimize the capacity of the BESS. Furthermore, the Bernstein approximation is utilized to process the chance constraint, and transform the optimization model into a deterministic form. Based on the linear weighted method and Benders decomposition, the optimization model is solved through alternating iteration. Case studies were conducted to validate the proposed scheme, showing superior performance in smoothing uncertain power fluctuations, and reducing frequency deviation under contingencies.