WP 2 – Ground modelling and clustering strategies

In a series of internally financed R&D projects, a so-called Intelligent Ground Model (IGM) has been developed at NGI, where interpreted geophysical and geotechnical data are combined, in order to describe the site conditions and major soil units. Based on these, generic geotechnical soil profiles can be predicted by means of interpolation and extrapolation from actual investigated locations, using the seismic data as guides for the soil stratigraphy across the site.

This generation, however, is inherent uncertain, and spatial and statistical variation needs to be considered. The uncertainty has a direct impact on the foundation concept selection and design. Hence the quantification of the uncertainties is crucial for the optimisation process.

The primary goal of this WP is to introduce statistical procedures, seismic attributes as well as machine-learning techniques in the IGM, enabling the user to quantify the impact of the variation of soil profiles and soil properties. Deliverable of this WP is a module which can generate generic N-dimensional profiles including ranges and distributions of the interpolated or extrapolated parameters. This will enable to perform the actual foundation design including a reliability assessment, that means quantifying the uncertainties with respect to the considered design limit state.

A secondary goal is to advance the IGM to also include metocean data, on which basis – together with the geotechnical and geophysical data – perform a clustering of the OWF. This allows to optimise the park-layout, and assist to identify critical locations which need particular attention.