Numerical simulation of mixed aseismic/seismic fault-slip induced by fluid injection using coupled X-FEM analysis

Fluid injection into a rock mass from industrial processes can cause perceivable seismic events that may raise public concern. This seismicity can be caused by injection-induced fluid pressure in the rock mass causing slip on faults. Here we provide a method to distinguish between aseismic and seismic mobilisation and to predict fault movement due to anthropogenic fluid injection. This was achieved by extending a two-dimensional fully coupled fluid and mechanical loading extended finite element model (X-FEM) via development of a dynamic analysis module as a standalone code in Matlab. This code considers fluid flow along the fault as well as into the rock mass and uses a directly proportional equivalent injected flow rate into the fault as the input. This model was validated by comparing the resultant pressure and normal and shear displacements calculated at the centre of the fault against observations from a decametre-scale in-situ experiment. The main results were that not only the mechanics of the fault could be simulated using this approach, but that the simulation correctly predicted the onset of seismicity and transition to dynamic analysis and at similar seismic magnitudes to observations. Parametric studies investigated the influence of the flow rate (when injecting a constant volume of water) and the effect of rate and state frictional parameters in representing modes of seismicity. The main conclusion is that this modelling technique using X-FEM provides an accurate method in accurately predicting modes, location and timing of fault remobilisation due to fluid injection inclusive of important precursory aseismic fault movements. These results are important, since they demonstrate the applicability of this X-FEM approach in accurately predicting the mechanics of fault reactivation and the resultant seismicity, aiding in the design and scheduling of fluid injection operations and in the optimisation of operational parameters.