Seismic characterisation of claystone as potential host rock for nuclear waste: applied studies in underground research laboratories

The search for a suitable repository represents an interdisciplinary challenge, whereby the investigation of potential host rocks is of central importance. While rock salt and crystalline rock have been the subject of intensive research since the 1970s, the investigation of claystone only began in the 1990s. Claystone is a promising host rock in Germany, but research here has so far focussed primarily on rock salt and crystalline rock as well. It is therefore essential to analyse the properties of claystone under real conditions and to further develop methods for investigating it in an underground infrastructure, such as a underground research laboratory (URL). This work is dedicated to the elastic properties of claystone, in particular the seismic velocities and the anisotropy. The latter, as a direction-dependent parameter, plays a central role in the creation of velocity models. In addition, the temperature-dependent change in seismic velocity is investigated in order to evaluate possibilities for seismic monitoring. At the beginning, the influence of anisotropy on the traveltimes of seismic measurements is illustrated using forward modelling. The Devito programme is used, which is able to take anisotropy into account efficiently using symbolic algorithms. To illustrate this, the qP-wave equation is solved and the approach is generalised so that it can also be applied to other approaches. An example from the geology of Mont Terri demonstrates the difference between isotropic and anisotropic velocities using seismic waveforms. The first case study was carried out in the URL Mont Terri, where a seismic survey was conducted along several galleries and niches to investigate the different facies of the Opalinus Clay (shaly and carbonate-rich sandy facies). For this purpose, 3-component geophones were installed in 2 m deep rock anchors and positioned at regular intervals around the survey area. The seismic waves were generated using a pneumatic impact source. The first arrivals for P- and S-wave were analysed in two ways: firstly to determine the anisotropy of the respective facies and secondly for 2D travel-time tomography. The shaly facies exhibit an anisotropy of 23% for P-waves and 32% for S-waves. The carbonate-rich sandy facies with 20% (P-waves) and 30% (S-waves) shows similar values. The traveltime tomography revealed an average velocity of 2.9 km/s for the P-waves in the shaly facies and 3.5 km/s in the carbonate-rich sandy facies. For the S-waves, mean velocities of 1.3 km/s (shaly facies) and 1.6 km/s (carbonate-rich sandy facies) were determined. The velocity contrast made it possible to precisely localise the two facies, which corresponds exactly to the mapped outcrop line of the geology. The second case study was also carried out in the URL Mont Terri, focusing on the laboratory’s Main Fault. For this purpose, the measurement geometry was extended with additional geophones and source points to enable a more precise determination of the anisotropy. This extended geometry also served as the basis for calculating the amount of carbonate-rich sandy facies from mixed ray paths of the two facies and achieving improved azimuthal coverage for this facies. The analysis showed that the shaly facies had a significantly higher S-wave anisotropy compared to the previous study, with 22% for P-waves and 36% for S-waves. In the carbonate-rich sandy facies, both values also increased to 24% (P-waves) and 36% (S-waves). The traveltime tomography showed that niche MA within the study area has a significant disturbance influence. This makes it necessary to extend the traveltime tomography into the third-dimension in order to avoid unrealistic ray paths. In addition, the ray coverage is one-sided, especially with long ray paths, which leads to averaging effects and prevents a detailed resolution of the fault zone. As part of the third study, a baseline survey for an injection experiment was carried out in the URL Tournemire and a fault zone was analysed for this purpose. This measurement serves as a reference for analysing changes in the elastic parameters. The fault zone was successfully resolved using traveltime tomography, which showed that the choice of seismic source has a significant influence. An impact source was again used for the experiment, supplemented by a magnetostrictive vibration source. The latter provided a better match between the measured velocities and the mapped fault zone. The influence of temperature changes on the seismic velocities was also investigated. For this purpose, a temperature field was simulated, generated by two heating elements, and converted into velocity changes. These were in turn transformed into synthetic traveltimes in order to check the recoverability of the generated anomaly. The results showed that the anomaly can be reliably reconstructed for both P- and S-waves. The results of the studies show that both the anisotropic and the temperature-dependent velocity changes are essential parameters for the seismic characterisation of claystone. They also show the difficulties arising from the infrastructure of an URL. These findings contribute to the further development of the geophysical investigation of claystone as a final rock and enable improved methods for monitoring and modelling the rock properties under real conditions.