State-based monthly operational rules for hydropower reservoirs to balance benefits, reliability, and vulnerability in hourly power supply

This study introduces a state-dependent, seasonally adaptive operational rule framework for cascaded hydropower reservoirs that, for the first time, incorporates detailed hourly power-demand balancing into multi-decadal, monthly-scale planning. This approach addresses a major limitation of previous studies, which often overlooked short-term operational changes, thereby overestimating long-term performance. The method aims to balance four key objectives—power generation, reliability, curtailment, and vulnerability—using a multi-objective evolutionary algorithm (NSGA-II) to generate a Pareto front of Non-Dominated Operational Rules (NDORs). Hypervolume (HV) is used to evaluate convergence and diversity, and a new priority-based filtering technique is developed to select a practically feasible rule. Applying this approach to the Xiaowan–Nuozhadu cascade in the Lancang River Basin over 50 years, the results show that maximizing hydropower production is the most compatible goal, followed by reliability, whereas reducing curtailment and vulnerability are largely conflicting objectives. The proposed NDOR decreases power curtailment and supply vulnerability while enhancing generation and reliability compared to the traditional Storage Target Rule (STR). Additionally, a Heuristic Proportional Method (HPM) enables hourly scheduling, achieving a 16.8-fold reduction in computational time compared to Gurobi-based optimization (GSP).