Risk Analysis And Probabilistic Modeling In Geotechnical Engineering

Geotechnical design using Factors of Safety recommended by Codes of Practice is based on largely unknown ?margins of safety?, usually without consideration for the importance of the project (i.e., potential consequences of failure) directly or explicitly. On the contrary, probabilistic modelling can be a useful tool in risk management as it provides the means to measure the risk to be managed. This is achieved by calculating the probability of failure associated with a specific design, i.e., by relating the classical factors of safety with the corresponding probability of failure and by using a cost optimization procedure to achieve ?optimal? designs, accounting for the expected costs in case of catastrophic failure. This paper describes the methods used to calculate the probability of failure and its correlation with the classical Factor of Safety using sampling techniques and a simplified method based on the moments of the factor of safety considered as a random variable. These procedures are illustrated in an example which investigates the stability of a slope.