Multi-stage distributionally robust scheduling of an integrated wind power and high-temperature thermal and compressed air energy storage system

• The potential of Hi-CAES in improving wind power dispatchability is evaluated. • A novel multi-stage distributionally robust optimisation method is developed. • Operation targets of Hi-CAES and power system are considered in the bi-level model. • An adaptive tolerance adjustment method is proposed to relax equality constraints. Compressed Air Energy Storage has been recognised as a feasible solution for addressing the power system challenge caused by high penetration of intermittent wind power. This paper proposes a combined system of wind power generation and a novel High-temperature thermal and Compressed Air Energy Storage (Hi-CAES), namely, WP-HiCAES. WP-HiCAES co-locates wind power with the Hi-CAES system, aiming to make wind power more dispatchable and achieve optimal cooperation with the power grid. A novel multi-stage distributionally robust optimisation (DRO) method is developed for optimal scheduling of a power system integrated with WP-HiCAES. The system optimisation is performed using a bi-level multi-stage distributionally robust scheduling (B-MDRS) model. The model is formed to have two levels of optimisation, aiming to maximise WP-HiCAES revenue and minimise power system cost. Numerical simulation results indicate that Hi-CAES elevates the discharge power capacity of A-CAES from 191.63 MW to 308.08 MW, offering higher flexibility in improving wind power dispatchability. The integration of the WP-HiCAES in the power system increases its revenue by 45.1 % and reduces the power system operational cost by 13.0 % when the wind penetration level rises from 30 % to 50 % in the case study. Comparison studies show that the proposed multi-stage DRO method is less conservative and saves more than 95 % computational time compared to other stochastic optimisation models. The encouraging results provide key support information for real-world Hi-CAES project implementations.