圖書典藏及影音數位平台

水庫淤泥之清淤方式除了機械挖掘外，採取水利排砂為最經濟之方法，本署為改善颱洪期間石門水庫原水濁度過高的問題，已經完成電廠防淤第1期改善工程的基本設計及水工模型試驗，另電廠防淤第2期正辦理基本設計審查作業，本計畫將以三維數值模擬探討電廠防淤改善工程之細部流場分析，包含：負壓區、管壁剪應力、流體對彎管作用力、防淤效率等幾何條件校核水工模型試驗（模型比尺為1/25的等比模型），探討石門水庫電廠發電機不同組合操作下壓力鋼管之壓力分佈情形；以及水庫泥砂在不同高程的濃度分佈條件下，防淤出水口泥砂濃度隨時間的變化情形，以作為石門水庫電廠防淤改善工程之水理與防淤效率推估，並藉由三維流場數值模擬與物理模型相互印證比較細部水理流況，作為設計與營運管理參考依據。另以三維流場數值模擬，利用數值模擬結果，結合不確定性分析方法，評估不同流場條件下壓力鋼管內之壓力分佈情形。

In the rehabilitation project for pressure steel pipes of the power plant in Shihmem Reservoir, the pipe #2 is rebuilt to be a sand escape. It is expected that silt in the reservoir can be moved to the downstream so as to enhance the capacity of Shihmen reservoir and associated power plant according to the strategy and operation of the pipe #2. The purpose of project is to analyze the detailed characteristics of the flow field, including pressure field, velocity field, force and shear stress on wall of pipe, by means of a 3D numerical simulation model with boundary conditions resulting from the hydraulic model test carried out by Water Resources Planning Institute in 2008. This project primarily performs uncertainty analysis for the zones in which the corresponding pressure is less than one atmospheric pressure based on different water levels, silt concentrations and conditions of outlet. The results of numerical experiences show that the pressure values at positions No.18 to No. 19 are less than one atmospheric pressure. This is poorly found from the accomplished hydraulic models tests and relevant research reports. For the model calibration and validation, this project develops a way to compare the physical model with the numerical model by taking into account the infiltration ratio of trash rack K value and roughness coefficient R value. The results indicate that the pressure values at different positions calculated by the numerical model are consistent with the results from the physical model. This implies the numerical model can capture the behavior of pressure estimated by the physical model and provides the detail analysis for the flow field. According to results from the numerical model, the fluid in the pipe #2 is significantly influenced by the resulting hydrostatic pressure from water level of the reservoirs, so that the variation of fluid field merely affect flow in the pipe #2. Therefore, the fluid field in the pipe #2 in independent on operation condition of the pipe #1, such as opening one out of two gages or both gages. The results from the case study show that the there are low pressure values at the positions A4 and A5 in the inside branch pipe, especially for the right side in which the corresponding pressure is less than one atmospheric pressure (approximately 9m), for the 245m of water level. Similarly, as the water level is 251m, the pressure approaches to 14m, namely, it is less than one atmospheric pressure. In addition, the probability of pressure less than one atmospheric at the position in the right side of branch pipe of the pipe #2, the distance of which is 410 m from the inlet, approximates 83% under the consideration of pure water. Therefore, it is worthy paying more attention on the uncertainty analysis for the pressure in the steel pipes under the consideration of various conditions. Since the 3D numerical model adopts results from the hydraulic model test as conditions in this project, it is necessary to compare the associated results with measurement taken in the pipe. Therefore, lots of further works are suggested to be made that the detail and practical flow and pressure variation in the pipes would be simulated and discussed using the proposed procedure of the model calibration and verification with the real pressure values and discharge.