河防構造物功能風險分析架構之研究

Abstract

本研究目的主要為建置一個評估河防構造物防洪功能失效之風險分析架構。因河防構造物包括堤防、護岸、丁壩等，其中堤防廣泛用於城市或流域之防災減洪工作，故本研究主要著重於評估評估流域在水文、水理及地文條件皆改變之情況下，依據保護標準之降雨量或洪水位所設計之堤防可能承受防洪功能失效之風險，也就是最高洪水位超越堤防高程之發生機率(失敗機率)。此一風險分析架構主要可區分為四部份：(1)風險因子之辨識與衍生：採用故障樹分析方法，界定出河川治理規劃之水文水理分析過程中，會造成河防構造物無法達到預期防洪功能之風險因子，可區分為水文、水理及地文因子，並採用拉丁高次取樣法衍生各風險因子；(2) 推估最高洪水位：將風險因子衍生值輸入降雨-逕流與水理模式中推估其最高洪水位； (3) 最高洪水位與風險因子關係式之建立：使用多變量迴歸分析法建立最高洪水位與風險因子之關係式；(4)風險計算：採用不確定性分析方法計算溢堤之失敗機率。 本研究以基隆河為應用案例，分別探討風險分析方法之適用性、各風險因子對堤防防洪能力之影響程度；堤防不同出水高及員山子分洪道設置對堤防溢堤風險之比較。依據分析結果可知，高等一階二矩法(AFOSM)較其他兩種方法適用於評估河防構造物之風險分析，其所需最高洪水位與風險因子關係式以非線性型態較能夠適當反應最高洪水位與風險因子之變化趨勢。各風險因子以最大無因次降雨比率與200年之降雨量為影響風險分析結果之重要因子，且堤防出水高及員山子分洪道皆可有效達成降低防洪功能失效風險之功能。綜合上述結果可知，本研究所建置之河防構造物風險分析架構可適用於河川堤防防洪功能之評估，期使未來其分析結果可作為制定河川治理規劃之參考。

The purpose of this study is to develop the flood risk assessment framework for flood control structures in fluvial system, including the embankment, dike, groyne and so on. Since the embankments are wildly used to prevent the urban and watershed from flooding, this study focuses on the risk analysis for the flood-control ability of the embankments, which mainly calculates the failure probability of the water level greater than the embankment. The failure probability probably results from the uncertainties of the rainfall depth or flood of specific protection criterions for the design of hydraulic structures, caused by the variation of hydrological, hydraulic and geometrical conditions in the catchments. The proposed risk analysis framework is grouped into four parts, (1) identification and generation of risk factors: using the fault-tree-analysis, the risk factors in the hydrologic and hydraulic routing can be identified, that is hydrological, hydraulic, and geometric factors. The risk factors are generated by Latin hypercubic sampling (LHS) method; (2) estimation of the maximum water levels: the maximum water levels are estimated by the hydrologic and hydraulic analysis with the generated risk factors; (3) establishment of the relationship between the maximum water levels and risk factors: the maximum water levels relationship with risk factors is established by the multi-variates regression analysis; and (4) calculation of the failure probability: the failure probability of the maximum water level greater than embankment is calculated by the risk and uncertainty methods. In this study, the proposed risk analysis framework is applied in the study area, Keelung river watershed, to evaluate the adequacy of uncertainty methods, the sensitivity of risk factors to the flood control capacity of embankment, and the effect of different freeboards and Yuan-Shan-Zi flood-diversion channel. In view of the results of numerical experiments, the advanced first-order-second-moment (AFOSM) method is more adequate to the risk analysis for the flood control capacity of hydraulic structures in fluvial system, of which the nonlinear relationship between the maximum water levels and risk factors can describe the behavior of the maximum water levels varied with the risk factors. The 200-yr rainfall depth and the maximum dimensionless rainfall ratio are more sensible than remaining risk factors on the flood control ability. Additionally, the freeboards of embankments and Yuan-Shan-Zi flood-diversion channel are able to effectively reduce the failure probability. In summary, the proposed risk analysis framework is demonstrated to be able to be used in the risk analysis for the prevention flood ability of flood-control structures in the river systems. Hence, it is expected the results from the proposed risk analysis framework would be referred in the river treatment and planning.