Pumped storage is crucial for maintaining energy balance and smoothing out the fluctuations from renewable sources. Yet, it is limited by its fixed capacity and lack of expandability post-construction, posing challenges to its long-term adaptability in the context of increasing installed renewable sources capacity. Underwater hydrogen storage, however, characterized by its green, low-carbon profile and ability for rapid energy release and long-term storage, complements pumped storage by enhancing the system's overall energy storage capacity and flexibility. Therefore, this paper proposes an innovative way for the pumped storage power station capacity expansion based on the underwater hydrogen storage introduction. On the basis of technical support of underwater hydrogen storage and time-series attribute consideration of uncertainties, a multi-objective distributionally robust optimization model with temporal correlation is constructed for underwater hydrogen storage planning and further scheduling pumped storage power station and underwater hydrogen storage to operate. Firstly, the system structure and operation mode after introducing underwater hydrogen storage into pumped storage power station are designed. Secondly, the temporal covariance conditions are introduced in a moment-based ambiguity set, with the aim of removing those distributions that do not match the temporal correlation of the historical forecasting errors samples. Finally, considering the “worst-case” distribution within the narrowed ambiguity set, an improved multi-objective distributionally robust optimization is constructed, which optimizes the capacity of each equipment in underwater hydrogen storage and the operation strategy of pumped storage power station and underwater hydrogen storage. Simulation mainly verifies: 1) it increases the economic revenue, electric load supply and photovoltaic output accommodation by 3.35 × 108 $, 2033.091 MW and 67584.054 MW, respectively, due to the introduction of underwater hydrogen storage for pumped storage power station expansion. 2) it improves cost savings, load supply reliability and photovoltaic output accommodation by 0.224 %,3.231 % and 2.722 % respectively, due to the introduction of temporal covariance to modify distributionally robust optimization model.
Read full abstract