This report concerns a study which was initially conducted for the Swedish Nuclear Power Inspectorate (SKI), which is now merged into the Swedish Radiation Safety Authority (SSM). The conclusions and viewpoints presented in the report are those of the authors and do not necessarily coincide with those of the SSM.
The Swedish Nuclear Fuel and Waste Management Co (SKB) has published Site Descriptive Models (SDM) of two sites, Forsmark and Laxemar, where SKB recently has finished site investigations for a repository for spent nuclear fuel. These models form the base for the conceptual understanding of the past and future evolution of the site and constitute one of the fundament in the forthcoming SKB application for building the repository. The approach in geological/structural modelling may vary. Essential in all modelling is that the data used is representative, the treatment of the data is transparent and systematic and that the applied approach is appropriate in a performance assessment.
The classification of geophysical borehole data, in order to identify and characterize structures, which intercepts the borehole, is an important part of 3D modelling of the structural pattern in the bedrock.
The base data used in this study, have been provided and delivered by the SKB and consist of borehole logging data from cored borehole KFM03A at SKB’s Forsmark site. The statistical method chosen for this study, Cluster Analysis using K-means method, groups data into a pre-defined number of clusters with the goal to minimize the variance for data within the group and to maximize the variance between the clusters.