View article

Deterministic code-based designs proposed for wind turbine foundations are typically biased on the conservative side, and overestimate the probability of failure, which can lead to higher than necessary construction cost. In this study reliability analysis of a gravity-based foundation, concerning its bearing capacity, is used to calibrate a code-based design procedure. A probabilistic finite-element model is developed to analyse the bearing capacity of a surface footing on soil with spatially variable undrained strength. Monte Carlo simulation is combined with a re-sampling simulation technique to perform the reliability analysis. The calibrated code-based design approach leads to savings of up to 20% in the concrete foundation volume, depending on the target annual reliability level. The study can form the basis for future optimisation on deterministic-based designs for wind turbine foundations.