Project completion time is one of the important indicators for managing/scheduling construction projects. This paper proposes a method for evaluating the reliability of project completion time by using the Universal Generating Function technique. The proposed method is used to overcome the limitations of existing project reliability evaluation methods by providing accurate results while requiring low computational burden. Also, unlike many of the methods revealed in the literature, in the proposed method, the activity duration does not need to follow any specific distribution. Our methodology consists of five main parts: 1) determining the u- function of the activity durations 2) modeling the stochastic project network as a parallel-series multi-state system, 3) determining the u-function of the project network paths to determine the probability of failure in each path, 4) determining the probability of simultaneous failure of each pair of paths, and 5) computing reliability bounds for project completion time by integrating the NRB and UGF methods. Results of the proposed method were compared with those of the Monte Carlo simulation method indicating that the reliability level obtained from both methods are indifferent at 0.01 significance level, while the computation time of the proposed method is significantly lower.
