Credible capacity gain identification method of peak-shaving scheduling of cascade hydro-wind-solar complementary system

Accurate assessment of peak-shaving credible capacity in hydro-wind-solar systems under climate extremes remains a critical challenge. This study develops a dynamic evaluation framework integrating multidimensional uncertainty modeling and sensitivity analysis, validated in China's Lancang River Basin (five cascade hydropower stations + 6 GW virtualized wind/solar farms). Using NASA meteorological data (2013–2015), historical hydropower records, and renewable output models, we propose a three-dimensional scenario generation method combining C-vine copula and Monte Carlo sampling to quantify spatiotemporal correlations among runoff, wind, and solar power. A Generalized Additive Model (GAM)-based sensitivity analysis reveals key findings: The average Peak-Shaving Credible Capacity Gain (PSCCG) in spring is 4811 MW, which is 58 % higher than the 3038 MW in autumn, primarily due to increased variability in renewable energy production. Additionally, PSCCG increases with the expansion of wind and solar installed capacity, validating hydropower's stabilizing role. Variance decomposition analysis highlights solar power's dominant contribution (78 %) to PSCCG enhancement, significantly exceeding wind power's 14 %, reflecting distinct impacts of different energy source characteristics on grid regulation demands. We recommend prioritizing solar-hydropower hybrids in spring-peak regions and embedding PSCCG into grid reliability standards to support China's 2030/2060 carbon goals.