一個營建工程專案的生命週期中,施工階段最是難免因為各種原因造成履約爭議,當履約爭議發生的時候,契約雙方可以採取的處理機制,不外乎調解、仲裁或民事訴訟等等。但是,無論採取何種爭議處理機制,爭訟程序往往曠日費時,且所費不貲,作為權益損失的一方,不僅實際的利益遭受損害,工期的損失所造成的進度延遲亦難以填補。當履約爭議的主要原因為非可歸責於業主與廠商時(如工程範圍外之第三方),為降低實際損失的影響及潛在的風險,廠商應如何選擇一最佳策略模式及解決方案呢? 為探討相鄰兩工程標案之間,當發生如上述工程紛爭時,應該如何研擬因應對策及評估解決方案,以期規避工程逾期和利益損失之風險,本研究遂以臺北市捷運工程連續壁坍孔侵界個案,作為探討分析的案例,利用有系統的分析方法,求得實務上最佳之解決對策。 賽局理論適合用於採討多個主體之間,在內在及外在不同條件下,對彼此之間互動時之策略行為。本研究使用完全訊息動態賽局作為主,非合作完全靜態賽局為輔,針對廣慈統包商與其他利害關係人之間的關連性,決定決策樹的架構。在完成條件假設及收益函數估算後,逐一求出各個子賽局之優勢策略Nash均衡,最後收煉得到整體最佳之預測模式。 本研究案例最後以實際執行之工期及成本兩項風險因素,作為預測模式之驗證,工期沒有發生逾期,成本損失亦在可接受範圍內,所得結果確實符合預期。惟完全訊息靜態賽局諸多假設過於理想,後續研究仍須考量更多條件,才能得到更加精確的結果。
During the life cycle of a construction project, performance disputes are unavoidable due to various reasons. When a performance dispute occurs, the course of action that can be taken by the two parties is limited to mediation, arbitration or civil litigation. However, no matter what kind of dispute resolution method is adopted, the litigation is usually time-consuming and costly. As the party who loses the rights and interests, not only loss of physical benefits is experienced, but the schedule delay caused by the loss of construction time is also difficult to compensate. When the cause of the performance dispute is neither liable to the proprietor nor the contractor (e.g. a third party beyond the scope of the construction), how should the contractor select one optimal strategic model and resolution to reduce the effects of physical loss and potential risks? To discuss how to develop countermeasures and evaluate solutions between two adjacent construction projects when the above-mentioned project disputes occur, to avoid the risk of construction delays and loss of profits, this study is based on the Taipei City MRT project. The case of the diaphragm wall trench collapse crossing the boundary is taken as the case for discussion and analysis, using a systematic analysis method to find the best practical solution. Game theory is applicable for adopting the strategic behavior of multiple subjects interacting with each other under different internal and external conditions. This study mainly adopts dynamic games with perfect information, supplemented by non-cooperative static games to determine the structure of the decision tree based on the correlation between the Guangzi Turnkey contractor and other stakeholders. Upon completion of the conditions and assumptions to estimate earnings functions, the Nash equilibrium of the dominant strategy of each subsidiary game is obtained one by one, and finally the overall optimal projection model is obtained through refining. Finally, the actual construction period and costs are the two risk factors used for the verification of the projection model. The construction period was not delayed, the cost of loss is within an acceptable range, and so the results are actually consistent with the expectation. However, many assumptions of perfect information static games are too ideal, and more conditions should be considered in subsequent studies to get more precise results.