Faults in geomechanical models – Necessary, nice, or nonsense?

Faults are an important factor for geoenergy applications due to either their sealing or conducting properties or their mechanical behaviour. Consequently, (thermo-hydro-) mechanical numerical investigation of geoenergy applications often include faults in their modelled rock volume. It is often assumed, that faults can significantly alter the far-field stresses, impacting both magnitudes and the orientation. In contrast to the far-field, stress rotations in the vicinity of faults are clearly observed in numerous borehole stress analyses across the world.

While an impact of faults on the stress field is expected, the representation of faults in (thermo-hydro-) mechanical numerical models is technically highly diverse. We investigate different methods to incorporate faults in geomechanical-numerical models and the relationship between faults and the stress state on two different spatial scales.

(1) The impact of faults on the stress state at distances of several hundred meters to a few kilometres (far-field) is tested. Therefore, faults are modelled with different numerical representations, material properties, fault orientations w.r.t. the stress field, fault width, extent, and boundary conditions. The results show that the impact of faults on the far-field is negligible in terms of the principal stress magnitudes and orientations. Only in extreme cases, stress changes in the far-field (>1km) can be observed, but these are not significant considering the general uncertainties in stress field observations.

(2) Stress changes within the fault zone are investigated, too. Particularly, the material contrast between the intact rock and the damage zone and fault core is regarded. This contrast can be responsible for a dramatic change in the stress tensor, observed as a rotation of the principal stress axes. In general, the change in the stress field increases with increasing stiffness contrast. The orientation of the fault w.r.t. the background stress field and the relative stress magnitudes, particularly the differential stress, lead to further stress changes. A small angle between the fault and the maximum principal stress axis and a small differential stress promote stress changes.

The study indicates that the impact of faults on the stress field is mostly limited to the fault’s near-field. These models provide an upper limit of stress changes, as several factor which alter stress changes (joints, viscosity etc.) are not included. However, the stress changes depend on the acting processes and material properties. Furthermore, for models used for site investigation, the implementation method and the mesh resolution can play an important role. All these factors need to be considered when planning the setup of a model with faults and their implementation.

The work was partly funded by BGE SpannEnD 2.0 project, the Bavarian State Ministry of Education and Culture (Science and Arts) within the framework of the “Geothermal-Alliance Bavaria” (GAB), and the DFG (grant PHYSALIS 523456847).