Integrated Multiscale Design, Market Participation, and Replacement Strategies for Battery Energy Storage Systems

Increased dependence on non-dispatchable energy sources has created a need for flexible energy storage systems capable of providing peak shaving and ancillary services. This work develops a computational framework to optimize sizing, replacement, and market participation strategies for battery energy storage systems. A multiscale linear programming formulation is proposed that spans real-time market decisions (minutes) to sizing and replacement decisions (years) while explicitly considering capacity loss due to degradation. These optimization problems include millions of variables and constraints, but can be efficiently solved using modern algorithms. Several case studies explore tradeoffs between market participation, degradation, power-to-energy ratio, and replacement strategies using historical energy and ancillary service price data from CAISO. Findings emphasize that simultaneously transacting both energy and ancillary services in day-ahead and real-time markets can provide a four to five fold increase in net present value compared to only buying/selling energy in the day-ahead market. Optimization also reveals market participation strategies that effectively manage degradation leading to an optimal energy-to-power ratio of less than 1 MWh/MW for NaS batteries. Results also show that developing ideal degradation-free batteries would improve net present value by 20%. This emphasizes that judicious management of degradation, sizing, and markets is critical.