Jiang Tao Yi

Due to the lack of knowledge of site-specific data and corresponding pile test results, the reliability of random finite element analysis (RFEM) on monopile and its findings remain questionable. This paper performed an RFEM on monopile in a specific offshore wind farm. The spatial variability of soil parameters in each seabed layer was carefully characterized based on both laboratory test results and calibrated in situ CPTU data. Statistical results demonstrated substantial variability in undrained shear strengths of two silty clay layers and insignificant variability in friction angles of three sandy soil layers, and all soil properties following lognormal distribution. RFEM incorporating the spatial variability of two silty clay layers was conducted to examine the probability distribution of monopile's bearing capacities. The monopile was wished-in-place to overcome the convergence difficulty induced by the deep penetration of pipe piles in the layered seabed. Monte Carlo simulation findings suggest that the spatial variability of soil strength resulted in an overprediction of the vertical bearing capacity, comparison between the statistical results of monopile capacities from RFEM and 43 pile test data in the field proves the reliability of the numerical method. Finally, a quantitative relationship between the safety factor and failure probability was developed to improve monopile design.