Probabilistic Design Analysis of Plate Anchors in Spatially Variable Clay

Plate anchors play a vital role in various geotechnical structures, particularly in offshore settings where high pullout capacity is required. The current partial factor method accounts for uncertainties in loads and material properties separately using partial load and material factors, but it may not well address the effects of site investigation and soil spatial variability on the failure probability of design anchors. This study evaluated the reliability of offshore plate anchors designed using the partial factor method against the pullout limit state in clay. A Single-Random-Variable (SRV) method and a Random Finite Element Method (RFEM) were introduced and compared. Parametric studies were also performed to investigate the effects of load ratio, partial material factor, soil sampling location, and soil spatial correlation length on the estimated failure probability. The results highlight the importance of the soil sampling location and spatial variation in anchor design. The SRV method does not consider these factors, resulting in either over- or under-estimations of the failure probability compared to the RFEM. Practical implications from the findings were also discussed. Overall, this study aids in the probabilistic design of offshore plate anchors, the calibration of partial material factors, and in providing valuable guidance to their practical applications.