Reliability aspects of rock tunnel design with the observational method

According to Eurocode 7, two accepted approaches for managing uncertainty in tunnel design are reliability based methods and the observational method. Reliability-based methods account for uncertainty by acknowledging the random variation of the input parameters; the observational method does this by verifying the expected behavior from an initial design during the course of construction. However, in the framework of the observational method, as defined in Eurocode 7, no guidance is given on the selection of suitable parameters for observation and how they can be linked to the limits of acceptable behavior and, at a sufficiently early stage, the decision for implementing contingency actions. Furthermore, no guidance is given on how to verify that the structure fulfills society's required safety level. In this paper, we present a design procedure for shotcrete-supported rock tunnels that combines reliability-based methods with the observational method. The design procedure applies a deformation-based limit state function for the shotcrete support that is based on the convergence confinement method. We suggest how the requirements in the observational method, as defined in Eurocode 7, may be satisfied for this application. In particular, we focus on the structural reliability aspects. The structural reliability of the preliminary design is assessed with Monte Carlo simulations by calculating the expected deformations of the tunnel. The appropriateness of the preliminary design is then verified by observing the actual deformations during the course of construction. The observed deformations are used to predict the future behavior of the tunnel and to update the assessed probability of unsatisfactory behavior. If the defined deformation-based alarm limit regarding the structural reliability is exceeded, predefined contingency actions are put into operation. The procedure is illustrated with a shotcrete-lined circular rock tunnel and practical aspects in satisfying the reliability requirements with the observational method are discussed.