ProbPerModel

• The quality of subsoil analyses has considerable consequences for the design of offshore foundation structures – the more reliable the models, the more precisely the foundations can be adapted.
• However, the methods currently applied are associated with considerable uncertainties, with foundation structures consequently being generously dimensioned, which results in higher costs.
• In the “ProbPerModel” research project, the partners are developing a 3D subsoil model integrating the geological measurements and synthetic data.
• Fraunhofer IWES, the project coordinator, is responsible in the “Data and models” subproject for the data collection, the application of novel interpolation methods for the creation of the models, and for the certification of the models among other things.

Before construction work can begin on an offshore wind farm, the subsoil must be investigated and a corresponding assessment performed. The data collected serve as the basis for the planning of the sites on the one hand, but also for the design of the foundation structures of the individual wind turbines. Conventional approaches for the subsoil assessment lead to foundations being very generously dimensioned due to the large safety factors involved.

As such, there is a need for innovative solutions in subsoil modeling on the basis of probabilistic methods. However, proven methods and a certified analysis tool for the calculated synthetic data are lacking.

This is where the “ProbPerModel” research project comes in. The aim is to develop a probabilistic 3D subsoil model for subsoil investigation and assessment. To do so, the scientists will quantify existing uncertainties such as the heterogeneity of the subsoil, transformation of parameters, and measuring uncertainties in such a way that the subsoil parameters relevant for measurement can be derived at every point of the model. The more accurate documentation of statistical dispersion of the parameters allows optimized foundation designs and thus leaner, more cost-effective foundations.

In the “Data and models” subproject, Fraunhofer IWES, the project coordinator, is taking over all the tasks related to data collection and modeling. The documentation and further processing of statistical uncertainties in the modeling plays a decisive role in this. For example, the researchers will develop methods for converting statistics of measured values into statistics of interpolated profiles and integrating them into a geological model as well as using neural networks to create a (semi)-automatic interpolation of geotechnical profiles in 3D space among other things. In addition, they are developing solutions for analyzing the statistical safety of the generated model and its certification.