Multi-objective optimization of a flexible compressed air energy storage system via discrete unit combination for accommodating volatile wind power

The direct coupling of compressed air energy storage (CAES) with highly fluctuating wind power often forces turbomachinery to operate under part-load, throttling, or variable-speed conditions, which may cause efficiency penalties and economic uncertainty. To improve wind power accommodation, this study proposes a flexible compressed air energy storage (FCAES) system that tracks volatile wind power through discrete combinations of rated-capacity compressor and expander units. In the proposed strategy, activated turbomachinery operates close to its nominal condition, while low- and intermediate-pressure buffer reservoirs are used to coordinate interstage mass-flow imbalance during frequent mode switching. A thermodynamic-economic model is established and validated against literature data, with a maximum relative error of 0.57%. Based on year-round 15-min wind-load data from Inner Mongolia, the rated power and the volumes of low-, intermediate-, and high-pressure reservoirs are optimized using NSGA-II and TOPSIS, with round-trip efficiency and payback period as dual objectives. The case results show that the maximum surplus and deficit powers are 151.03 MW and 122.97 MW, respectively. The optimized FCAES configuration achieves a system-level round-trip efficiency of 45.2% and a payback period of 4.84 years, indicating that discrete rated-power unit combination with buffer-assisted mass-flow coordination is a promising route for improving the techno-economic feasibility of wind-coupled CAES systems.