This document discusses a project that aims to improve geological interpretation of seismic data and develop an interpretation system using artificial intelligence to predict hazardous structures during tunnel construction. It gives an overview of data mining conducted on geological and seismic properties of rocks from the Faido access tunnel in Switzerland. Statistical analysis found relationships between seismic properties (wave velocities, ratios, and gradients) and geological features (fracturing and fabric), which will help in developing the interpretation system to predict challenges for tunnel construction.
This document discusses a project that aims to improve geological interpretation of seismic data and develop an interpretation system using artificial intelligence to predict hazardous structures during tunnel construction. It gives an overview of data mining conducted on geological and seismic properties of rocks from the Faido access tunnel in Switzerland. Statistical analysis found relationships between seismic properties (wave velocities, ratios, and gradients) and geological features (fracturing and fabric), which will help in developing the interpretation system to predict challenges for tunnel construction.
This document discusses a project that aims to improve geological interpretation of seismic data and develop an interpretation system using artificial intelligence to predict hazardous structures during tunnel construction. It gives an overview of data mining conducted on geological and seismic properties of rocks from the Faido access tunnel in Switzerland. Statistical analysis found relationships between seismic properties (wave velocities, ratios, and gradients) and geological features (fracturing and fabric), which will help in developing the interpretation system to predict challenges for tunnel construction.
SYSTEM WITHIN THE ALPTRANSIT PROJECT C. D. Klose (1,2), R. Giese (1), S. Löw (2), G. Borm (1) (1) Geomechanics and Geotechnology, GFZ Potsdam, Potsdam 14473, Germany, (2) Geology Institute, ETH-Zürich, Zürich 8093, Switzerland Especially for deep underground excavations, the prediction of the locations of small- scale hazardous geotechnical structures is nearly impossible when exploration is re- stricted to surface based methods. Hence, for the AlpTransit base tunnels, exploration ahead has become an essential component of the excavation plan. The project de- scribed in this talk aims at improving the technology for the geological interpretation of reflection seismic data. The discovered geological-seismic relations will be used to develop an interpretation system based on artificial intelligence to predict hazardous geotechnical structures of the advancing tunnel face. This talk gives, at first, an overview about the data mining of geological and seismic properties of metamorphic rocks within the Penninic gneiss zone in Southern Switzer- land. The data results from measurements of a specific geophysical prediction system developed by the GFZ Potsdam, Germany, along the 2600 m long and 1400 m deep Faido access tunnel. The goal is to find those seismic features (i.e. compression and shear wave velocities, velocity ratios and velocity gradients) which show a significant relation to geological properties (i.e. fracturing and fabric features). The seismic properties were acquired from different tomograms, whereas the geolog- ical features derive from tunnel face maps. The features are statistically compared with the seismic rock properties taking into account the different methods used for the tunnel excavation (TBM and Drill/Blast). Fracturing and the mica content stay in a positive relation to the velocity values. Both, P- and S-wave velocities near the tunnel surface describe the petrology better, whereas in the interior of the rock mass they correlate to natural micro- and macro-scopic fractures surrounding tectonites, i.e. cataclasites. The latter lie outside of the excavation damage zone and the tunnel loos- ening zone. The shear wave velocities are better indicators for rock fracturing than compression wave velocities. The velocity ratios indicate the mica content and the water content of the rocks.