Back analysis of settlement of Teven Road trial embankment using Bayesian updating

A trial embankment constructed on soft clay has been back analysed using Bayesian updating of parameters in a one-dimensional geotechnical model. The purpose of the back analysis was to demonstrate that Bayesian updating can be used to fit the data and to predict the final settlement. There was a limited amount of laboratory test data and site investigation data available, as is normal commercial practice. The initial material parameters were deemed to be biased due to sample disturbance and its effect on laboratory test data. The monitoring data comprised surface settlement measurements at the centreline of the embankment. The embankment was constructed in two stages over a period of 7 years. Bayesian updating of parameters using settlement was able to fit the entire time-settlement history.