Reliability analysis and risk assessment of deep excavations using random-set finite element method and event tree technique

Abstract In the recent past years, the importance of reliability analysis and risk assessment in solving Geotechnical Engineering problems is underlined in literature. This study is aimed to show the feasibility of Random Set Finite Element Method (RS-FEM) and Event Tree Analysis (ETA) in assessing the reliability and risk to fatality of urban excavations. In order to fulfill this aim, a case study (22-meter excavation) which is stabilized with soil anchors is considered in North of Tehran, Iran. Prior to risk assessment, the failure probability of the deep excavation is estimated using RS-FEM, which is a relatively novel method in analyzing the reliability of deep excavations. Findings of the reliability analysis suggest that the upper and lower bounds of failure probabilities (Pf) for the considered case study are 0.0002 and 0.00001, respectively. Subsequently, using ETA, different possible scenarios, which may lead to fatality, are defined and proper probabilities are assigned to each scenario. Finally, for risk quantification purposes, the outcome probability of each scenario is multiplied by a considered value of statistical life as well as an aversion factor. Overall, findings recommend that the presented methodology (i.e. RS-FEM coupled with Event Tree Analysis) in this paper can be employed for quantitative risk assessment (QRA) in deep urban excavations. Nevertheless, further research is recommended for other case studies as practicing the aforementioned method can constitute common sense on QRA of deep excavations.