A LIFE CYCLE RISK MANAGEMENT AND PREDICTION SYSTEM FOR CONSTRUCTION JOINT VENTURES

Abstract

This research develops the life cycle risk management and prediction (LCRMP) system for construction joint ventures (CJVs). The system focuses on the dynamic of risk characteristics throughout the CJV life cycle. It consists of two subsystems: the multi-objective risk management (M-ORM) subsystem and the multi-determinant risk prediction (M-DRP) subsystem. The first subsystem was modified from the ISO risk management process. The 30 CJV risks were identified and analyzed, and the treatment options for each individual risk were investigated. The second subsystem was established by first deriving 48 CJV determinants, which were used as the framework of multi-determinant matrices. These matrices were then analyzed by the analytic hierarchy process (AHP). The inputs of both subsystems were derived from the results of relevant past research as well as the questionnaire surveys and the in-depth interviews with a large panel of CJV experts in Thailand. The Delphi method was also integrated into the data collection processes to increase the accuracy of the results. The results indicated that the risk characteristics in each phase of CJV life cycle were quite dynamic in terms of the number of risks as well as their consequence and likelihood of occurrence. Based on nonparametric statistics, there were 21 risks, the consequence and likelihood of which were sensitive to the organization structures of CJVs, namely, the cooperative governance joint venture (CG-JV) and the separate governance joint venture (SG-JV). This research contributes to CJV risk management in many ways. The CJV partners can use the characteristics of risks throughout the CJV life cycle and appropriate risk treatment options from the multi-objective risk management subsystem to develop a comprehensive risk management for their CJV administration. The relations between the CJV organization structures and their risk characteristics can be used to design an appropriate CJV organization for a certain contractor. The contractors can apply the multi-determinant risk prediction subsystem to predict the consequence and likelihood of risks based on the 48 underlying project determinants.