Hierarchical Bayesian Inference of Soil Resistance to Offshore Monopile Driving

CPT-based pile drivability predictions have been widely applied in recent years for offshore monopiles. These predictions rely on well-established empirical relations for the (semi-static) shaft- and end-bearing resistance of the monopile. However, significant uncertainties remain, which originate from geospatial and operational uncertainties and limitations of the empirical relations and the pile-soil interface. Some effects are clustered: differences in pile geometry, shape or material could affect predictions, such as microstructure, geological history, rate-effects and transitional soils. A hierarchical Bayesian approach is proposed for parameter inference for groups of installed monopiles, while also inferring a global probability distribution.

Keywords: Pile drivability, Bayesian, Hierarchical, CPT.