Probabilistic and Fuzzy Fault-Tree Analyses for Modelling Cave-in Accidents

Construction injury accidents result from different causes. Risk evaluation for cave-ins using tradi-tional Fault-Tree Analysis (FTA) can be difficult, especially since the variables resulting in cave-ins are unique; in addition, historical data, when available, are often incomplete. In construction, the assess-ment of risk is based on linguistic terms using subjective judgment of linguistic values such as severe, very likely, etc. Such linguistic terms are best modelled using fuzzy set theory. The traditional FTA method has been widely used to calculate the probability of the top undesired event, which is based on the historical data of the occurrence and the severity of the basic events. FTA implementation into construction projects needs to be modified since assessment of contributing events to cave-in accidents is based on managerial experience using experiential subjective expressions. This paper introduces a fuzzy triangular model to assess risks associated with excavation work in advance and helps manage-ment prepare solutions in advance.