異重流模式於石門水庫閘門操作之應用

Abstract

為研究石門水庫日益嚴重的泥砂淤積問題，本研究採用層平均二維模式探討水庫異重流運移型態。研究成果顯示，完成參數較佳化之模式能有效評估異重流運移速度、到達壩前時間及泥砂濃度遞減率，並發展高度正相關的回歸公式以預測異重流於庫區之運動情形。基於上述研究成果，庫區不同閘門之排砂效率也能合理評估。不僅於此，本研究進一步提出底孔修正公式及水庫操作策略以優化整體排砂效率。總結上述，本研究利用現有數值模式建立一套水庫防淤決策方法，模式的一致性能夠合理並穩定預測異重流運行狀態。未來可進一步結合即時監測數據，以達前期預警之效。

Sedimentation has been a serious problem of Taiwan reservoirs, and the rapidly trapped sediment caused by extreme flood events has affected the storage capacity. Therefore, reservoir deposition becomes a highly-discussed topic. In Taiwan, the timing of sediment release usually happens during typhoon periods. On the other hand, the density current would be produced due to the incoming hyper-concentrated flow. Hence, grasping the movement of the density current is a focused issue. In this study, a 2-D layer-averaged density current numerical model would be applied to simulate the density current flow during typhoon periods. Firstly, the field measured data were adopted to calibrate simulation parameters and to verify the simulation results. Next, three highly correlated regression equations are developed to calculate the travel velocity, arrival time, and concentration decay rate. Furthermore, a reasonable estimation of the sediment release amount can be established based on these outcomes. Lastly, the modification of the release efficiency at the bottom outlet, and the method of reservoir operation can be presented to improve the desilting strategy in the Shimen Reservoir. Above all, the flow pattern of the density current is estimable. Accordingly, this research can be adopted to predict the flow pattern of the density current due to high consistency of this model. In conclusion, this study can combine the real-time observed data to conduct the rolling prediction and provide the approximate tendency of the density current flow, and further be the reference to the early warning beneficial to reservoir management.