Risk assessment methodology for a deep foundation pit construction project in Shanghai, China

Risk assessment and risk management for deep foundation pit engineering are essential for quality and safety in civil engineering owing to the needs of urban construction projects. However, uncertainty and fuzziness continue to challenge studies of the probability and consequences of risks in this area. Therefore, a fuzzy comprehensive evaluation method based on Bayesian networks (BNs) is proposed to assess the risks of deep foundation pit construction. This methodology has five main parts: modeling of BNs, determination of occurrence probabilities of risk events, assessment of consequences, calculations of risk value and membership degree of risk rating, and definitions of risk acceptance criteria. In a case study, primary data analysis of a simple accident database from a deep foundation pit construction project in Shanghai is used to provide the probabilities of basic events and approximate distributions of risk consequences. The probability of every risk event is calculated by using deductive BN techniques. Then the consequence of each event is calculated by using fuzzy analysis (i.e., statistical consequence distributions and weight coefficients of risk events are determined through the database). A fuzzy comprehensive evaluation model with a membership function is also presented, and each risk event in the deep foundation pit construction his rated. In addition, risk precautions and control measures are suggested on the basis of the risk assessment results and are applied to risk management in deep foundation pit construction.