Resumen
Operators often have a dilemma in deciding what water levels the over-year hydropower reservoirs should drawdown at the end of dry seasons, either too high to achieve a large firm hydropower output during the dry seasons in the current year and minor spillage in coming flood seasons, or too low to refill to the full storage capacity at the end of the flood seasons and a greater firm hydropower output in the coming year. This work formulates a third-monthly (in an interval of about ten days) hydropower scheduling model, which is linearized by linearly concaving the nonlinear functions and presents a rolling strategy to simulate many years of reservoir operations to investigate how the water level at the end of dry seasons will impact the performances, including the energy production, firm hydropower output, full-refilling rate, etc. Applied to 11 cascaded hydropower reservoirs in a river in southwest China, the simulation reveals that targeting a drawdown water level between 1185?1214 m for one of its major over-year reservoirs and 774?791 m for another is the most favorable option for generating more hydropower and yielding larger firm hydropower output.