000004888 001__ 4888
000004888 005__ 20200526162249.0
000004888 037__ $$aTHESIS-2020-0003
000004888 100__ $$aDe Meyer, Cédric
000004888 245__ $$aTheoretical model for the analysis of induced seismicity at geothermal projects in the Campine Basin (Belgium)
000004888 260__ $$c2020
000004888 300__ $$a75
000004888 502__ $$bBachelor$$cKU Leuven
000004888 520__ $$aIn recent years, small earthquakes occurred near the injection well of the Balmatt geothermal facility in Mol. The main reason for this induced seismicity is the injection of water in the limestone reservoir of the Dinantian and the reactivation of (a) fault(s) that is (are) presumed to be present underneath the facility. It is necessary to reduce the risk of seismic events, to ensure the future of other geothermal projects, such as the facility of Janssen Pharmaceutics at Beerse. This reduction is partially accomplished by using accurate values for geotechnical parameters in the calculation of the stress state of the geological setting. In this study, a Python model is developed that summarizes the equations for the calculation of the stress state in Mohr space. The resulting model is an enhanced version of the model made by Van Baelen (2019) and improved by Van Noten (2019). The primary enhancement that is made in this study is the incorporation of calculations for the stress state of faults with all possible orientations in 3D space. The model plots the stress state of a fault for a given set of geotechnical parameters and calculates the maximum allowed injection fluid pressure before an optimally-oriented fault would be reactivated. All calculations use estimated parameter values obtained from a literature analysis. The model also makes sensitivity analyses for the used geotechnical parameters to investigate the influence of parameter value changes on the predicted maximum injection fluid pressure. The model confirms the occurrence of induced seismicity, as observed at the Balmatt plant, when the maximum injection pressure that is reported for this facility is used. This result shows that the constructed Python codes have predictive power. For an optimally-oriented plane, the calculated maximum allowed injection pressure equals 1.8 MPa, which is much lower than the injection pressures used at the plant. A maximum allowed injection pressure for the reactivation of an optimally-oriented fault is also proposed for the facility of Janssen Pharmaceutics, to try to increase this predictive power of the model. The sensitivity analysis shows a multitude of relationships between the parameters and the resulting injection pressure. No relationship is observed for the Poisson’s ratio (for an optimally-oriented plane), the rock density and the depth of the water table. An increase in calculated injection pressure is observed for an increase in the cohesion of the fault plane, water density and static friction angle. A decrease in calculated injection pressure is observed for decreasing values of the depth of the target aquifer and the borehole fluid density. If the dip of the fault differs from the dip of the optimally-oriented plane, the maximum injection pressure increases. Lastly, a decrease in the azimuthal angle between the fault’s trace and σ3 increases the value of the maximum allowed injection pressure.
000004888 594__ $$aNO
000004888 6531_ $$ainduced seismicity
000004888 6531_ $$aMohr circle
000004888 6531_ $$afault reactivation
000004888 701__ $$aSintubin, Manuel (KULeuven)
000004888 701__ $$aVan Noten, Koen (ROB)
000004888 701__ $$aVan Baelen, Hervé (NIRAS)
000004888 8560_ $$fkoen.vannoten@observatoire.be
000004888 980__ $$aTHESIS