Short-term peak-shaving strategies for cascade hydropower against evolving net load in systems with high VRE penetration

The integration of high-penetration variable renewable energy is reshaping net-load profiles into multi-peak, multi-valley structures, creating a growing mismatch between dynamic peak-shaving demand and executable hydropower dispatch. To address this, this paper proposes an integrated framework that establishes a quantitative link between net-load morphology and hydropower response behavior through three interconnected innovations. First, quantitative indices are defined to characterize net-load structural evolution under increasing VRE penetration. Second, a short-term cascade hydropower scheduling model is developed to explicitly balance peak-shaving effectiveness and operational smoothness, coupled with an enhanced algorithm to ensure executability. Third, based on the proposed indices and scheduling framework, representative dispatch modes are identified and quantitative switching criteria are established, revealing systematic mode evolution under increasing VRE penetration. A case study in China demonstrates that: 1) the net load exhibits a staged structural reconstruction, in which the dominant valley shifts from nighttime to midday, thereby progressively compressing the continuous operating space of conventional units; 2) the proposed method achieves comparable peak-shaving performance while substantially improving executability, reducing hydropower output fluctuation indices by 23.5% (dry season) and 29.6% (wet season); 3) dispatch patterns evolve systematically with renewable penetration —the feasible mode set contracts beyond identified thresholds, revealing how net-load restructuring propagates into hydropower operating behavior. These findings provide both scientific understanding of flexibility-demand evolution and actionable decision boundaries for adaptive dispatch in transitioning renewable-dominant systems.