| dc.description.abstract | CO2 storage in aquifers has the potential of storage capacities of up to 1.5 Gt in the Netherlands. In geomechanical modelling of the risk of seismicity upon CO2 injection in aquifers, currently the potential of chemical weakening of faults is often not taken in account. Laboratory experiments or observations form natural analogues of CO2 injection however suggest that the dissolution of CO2, which causes a lowering of the pH, can result in the dissolution of calcite which can cause weakening of fault strength or frictional strength. In this study, the effect on seismicity of chemically weakening static friction, dynamic friction, or cohesion of calcite or calcite bearing fault gouge in the presence of dissolved CO2 is tested. To test these effects, three existing codes are coupled to simulate CO2 flow, stress changes and corresponding fault rupture. Injection and flow through the sandstone reservoir is modelled using flow simulator TOUGH-ECO2M, from which the spatiotemporal relationships of temperature, pressure and CO2 saturation are taken. Corresponding stress changes are calculated from poro-elastic and thermo-elastic equations using inhouse software SRIMA (Seal and Reservoir Integrity Mechanical Analysis). The stress changes are then projected on a fault surface with predefined roughness using BSQuakeSim (Block-Spring earthQuake catalogues Simulator), which is a rupture model for induced seismicity following stress changes. CO2 concentration is coupled in several simulations to the value of either static friction, dynamic friction or cohesion following simple linear relations to individually test the influence of CO2 on the rupture timing, magnitude and frequency. This study found that poro-elastic stress changes, thermos-elastic stress changes and chemical weakening from CO2 saturation variation individually contribute to destabilisation of a fault following different spatio-temporal relations. Chemical weakening mainly lowers the required thermo-elastic stress change to reach criticality but can also act as the triggering mechanism. Pressure changes and CO2- calcite interaction affect the fault stability over a longer spatial extent than temperature changes. In the current simulation setup, faults are largely chemically weakened prior to the arrival of a cold front at the fault surface. It is found that lowering the static friction leads to earlier rupture events, a higher total number of rupture events but not necessarily of higher moment magnitude. Lowering the static friction increases the stress drop and thus stress release upon rupture. Therefore, rupture events of higher moment magnitude are found. Chemically lowering cohesion values leads to earlier rupture events, and a larger area of the fault surface along which slip occurs. Localized weakening of parts of a fault surface could affect the stability of the whole fault. In should be noted that to visualise changes in seismicity, initial criticality was required, therefore some unfavourable assumptions were made regarding injection rate and injection temperature. Also, the reactivity of calcite with the dissolved CO2 is most likely overestimated in this study. However, under certain site-specific reservoir and injection conditions, localized fault strength weakening following CO2 injection can contribute to the triggering of rupture events and magnitude of events and should therefore be taken in account in future seismic risk assessments of CCS projects in aquifers. | |
