Probabilistic assessment of tunnel construction performance based on data

•Probabilistic model of tunnel construction time is introduced.•Data from tunnels built with conventional tunneling are statistically analyzed.•Novel probability distribution describing the real performance is suggested.•Effect of disturbances and failures on the uncertainty of time estimate is shown.

A probabilistic model for estimating tunnel construction time is learnt with data from past tunnel projects. The model is based on the Dynamic Bayesian Network technique. The model inputs are determined through an analysis of data from three tunnels built by means of the conventional tunneling method. The data motivate the development of a novel probability distribution to describe the excavation performance. In addition, the probability of construction failure events and the delay caused by such failures are estimated using databases available in the literature. The model is applied to a case study, in which it is demonstrated how observations from the tunnel construction process can be included to continuously update the prediction of construction time.