Capacity Planning of a Park-Level Integrated Energy System Considering Seasonal Salt-Cavern Hydrogen Storage and Adaptive Representative Days

Park-level integrated energy systems with high shares of wind and photovoltaic power face pronounced seasonal source–load mismatches, renewable energy curtailment, and low-carbon operation challenges. This paper proposes a capacity planning method considering seasonal salt-cavern hydrogen storage and adaptive representative days. An electricity–heat–cooling–hydrogen coupled system is established by integrating renewable generation, conventional conversion units, short-term storage, electrolyzers, fuel cells, and salt-cavern hydrogen storage, together with waste-heat recovery and tiered carbon trading. To represent interseasonal hydrogen transfer under representative-day modeling, a seasonal hydrogen inventory formulation based on weighted net hydrogen changes is developed, considering cushion gas, storage bounds, injection and withdrawal efficiencies, and flow-rate limits. A season-specific adaptive K-medoids method based on CRITIC evaluation is further proposed to determine the number of representative days, while zero-weight extreme days are introduced to verify capacity feasibility under boundary conditions. The optimization objective is to minimize annualized total cost. Case studies show that removing seasonal hydrogen storage increases total system cost by 23.63%, raises wind and photovoltaic curtailment from 1.81% to 13.09%, and increases carbon emissions by 13.74%. The proposed method improves economic, renewable-energy-utilization, and low-carbon performance.