Post-scour reliability assessment of offshore monopiles incorporating field monitoring data and multivariate Gaussian modelling

Scour around monopile foundations can significantly affect the performance and reliability of offshore wind turbines. Owing to the complex sediment-fluid-structure interactions, scour depth exhibits substantial variability and should therefore be treated as a random variable rather than a deterministic quantity as assumed in most existing studies. Based on high-quality scour inspection data across an offshore wind farm, this paper proposes a probabilistic framework to assess the post-scour reliability of monopile foundations and to determine the design safety factors required to achieve specified target reliability levels. First, a multivariate Gaussian distribution is employed to characterise the joint distribution of correlated scour depths measured at four inspection times. A beam-spring model with an effective depth correction is then combined with Monte Carlo simulation to model the stochastic pile responses considering the uncertainty in local scour depths. Reliability assessments for both ultimate and serviceability limit states are subsequently carried out for each inspection time, and the required safety factors accounting for scour depth variability are determined for target reliability levels. Results show that the scour depth at each inspection time can be reasonably modelled by a normal distribution. For the ultimate limit state, the required design safety factors remain within the code-recommended range of 1.2-2.0 when the target probability of failure is 10⁻³ or 10⁻⁴, but increase to 2.05 and 2.09 at 6 and 8 months after installation, respectively, when the target probability of failure is 10⁻⁵. For the serviceability limit state, the required safety factors exceed 4.0 for all target probabilities of failure considered. These results indicate that traditional design approaches may be unconservative when scour depth variability is explicitly considered, particularly with respect to serviceability requirements.