Bilevel stochastic optimization of pumped-storage hydropower with dynamic hydraulic short-circuit operation

Dynamic control of hydraulic short circuits (HSC) can significantly enhance the economic performance and operational flexibility of pumped storage hydropower plants. This study demonstrates that incorporating optimized HSC strategies into plant scheduling increases net revenue by 4–6%, reduces unit start-ups by approximately 30%, and improves reservoir head stability compared with conventional pumped storage operation. These benefits are particularly pronounced in electricity markets characterized by high price volatility. To achieve this, a bilevel stochastic optimization framework is developed for the operational scheduling of pumped storage plants equipped with HSC capability. The upper level determines strategic decisions, including maintenance planning and HSC operational policies, while the lower level performs short-term dispatch under electricity price uncertainty. The hydraulic short circuit is explicitly modeled, enabling partial decoupling between hydraulic flow and electrical power output, thereby providing enhanced flexibility for head management and load balancing. To efficiently solve the resulting nonlinear stochastic optimization problem, a hybrid computational framework is proposed. The solution approach integrates sequential convex programming (SCP), sensitivity-based parameter updates, ADMM-based multi-scenario coordination, and trust-region stabilization to ensure numerical robustness and convergence. Compared with a global mixed-integer nonlinear programming (MINLP) solver, the proposed approach achieves near-optimal solutions while requiring less than 2% of the computational time. The results confirm that integrating dynamic HSC control within a bilevel stochastic scheduling framework significantly improves both economic performance and operational stability, especially under highly volatile market conditions.