Research On Graded Optimal Operation Of Three-gorge And Gezhouba Cascade Hydropower Stations In Flood Season

In accordance with the requirements of water resource utilization for the Three-gorge and Gezhouba cascade hydropower stations in flood season, research on short-term optimal operation for the Three-gorge and Gezhouba cascade hydropower stations is carried out, which considers the characteristics of inflow of Three-gorge reservoir and tight water and hydraulic relation between Three-gorge and Gezhouba. In this research, multi-type characteristics are analyzed to propose various optimal operation rules for different discharge flow levels. Through building optimization dispatching model with Genetic Algorithm as solving method, optimization dispatching solutions for the Three-gorge and Gezhouba cascade hydropower stations are obtained. The research results can support actual scheduling. The main research works and results include the following aspects:First of all, multi-type characteristics of Three-gorge and Gezhouba cascade hydropower stations are analyzed and concluded.(1) In flood season, hydraulic head for electricity generation may be lower than the rated value. The maximum allowable output of a water turbine will be reduced by more than 50MW;(2) When hydraulic head of Gezhouba hydropower station is lower than 19.0m, operation efficiencies of water turbines will reduce significantly with lower hydraulic head, and the highest reduction will be more than 0.1/m;(3) Tail water level of Three-gorge reservoir is influenced by water level of Gezhouba reservoir, and its change will decrease with larger discharge flow of Three-gorge reservoir;(4) In flood season, the maximum water discharge capacity for electricity generation of Three-gorge hydropower station and Gezhouba hydropower station is 30000m3/s and 20000m3/s respectively.Secondly, three operation rules of Three-gorge and Gezhouba cascade hydropower stations are proposed aiming at maximizing the sum of electricity generation in an operation period, which are based on the average water discharge of Three-gorge reservoir and multi-type characteristics.(1) When the average water discharge is more than 30000m3/s in an operation period, Gezhouba reservoir should schedule its reverse regulation process according to the key value, initial reservoir level and daily water discharge progress of Three-gorge reservoir;(2) When the average water discharge is more than 20000m3/s and less than 30000m3/s in an operation period, Gezhouba reservoir schedules its reverse regulation process based on hydraulic head which is 17.0m;(3) When the average water discharge is less than 20000m3/s, according to water balance and typical reservoir level process, Gezhouba reservoir should schedule its reverse regulation process depended on the maximum water discharge capacity, initial reservoir level and other decision factors.Thirdly, optimization dispatching model of the Three-gorge and Gezhouba cascade hydropower stations built aims at maximizing the sum of electricity generation in an operation period, which considers operation characteristics of water turbines, relation of water balance, relation of water level and water volume, relation of water discharge and tail water level and other constraints. In order to improve searching efficiency, Genetic Algorithm is used to solve the model with optimal operation rules embedded in population initialization stage.Finally, with the Three-gorge and Gezhouba hydropower stations as an example, optimal operation rules in flood season are verified by a case study. The results show:(1) When the average water discharge is more than 30000m3/s, it can directly obtain the optimization dispatching solutions by using optimal operation rules without optimization;(2) When the average water discharge is less than 30000m3/s, initial population with optimal operation rules applied has a better fitness value. better operation solutions can be obtained with faster optimization speed;(3) Compared with the stochastic optimization, the optimization solutions which embed optimal operation rules are more practical with smaller changes of water discharge and output in two intervals as well as less times of unit commitment.