Risk mode lling techniques for complex tunnelling projects

Abstract: Although the outdated idea of calculating risk as the deterministic product of its likelihood and impact is not entirely a thing of the past, a multitude of organizations have adopted the usage of probabilistic methods for risk analysis, as they better reflect the inherent uncertainty of cost and risk estimates with ranges. But are these advancements far-reaching enough in a complex environment such as tunnelling projects, where inadequate modelling of cost and risk can result in estimate discrepancies that run into millions of euros? This paper explores two advanced probabilistic modelling techniques for tunnelling projects, a first one demonstrating how the cost and schedule uncertainty related to changes in excavation and support classes can be modelled with correlated variables. A second example models the risk related to explosive ordnance clearance on construction sites using a probabilistic event tree.