Thermo-Hydro-Mechanical modelling of a reinjection operation in the North Alpine Foreland Basin

The use of geothermal energy inevitably causes changes in pore fluid pressure in the subsurface due to the production of fluids at one location and their reinjection at another location. In turn, this local change in pore pressure influences the undisturbed effective stress state. Depending on the stress state and changes in pore pressure, the rock's stability can be compromised. A resulting failure of the rock is perceived as induced seismicity. Since this is undesirable, the thermo-hydro-mechanical response of the rock to operations is often studied with the aim of better understanding the underlying mechanisms that lead to induced seismicity and potentially identifying ways to constrain its impact. However, the availability of direct data on the relevant parameters is usually sparse, if available at all. Thus, it is of interest to understand the magnitude of impact of different parameters on the stability. The ultimate goal is to enhance understanding of the parameters that are most decisive.

Here, we present and quantify the influence of various relevant parameters on the stability of the rock mass. Specifically, we examine the effect of different initial stress states, varying hydraulic and mechanical fault properties, and different rock stiffnesses on the stability of the rock mass. For quantification and comparability, we use the slip tendency as a measure of how close a rock mass is to failure. Differences in slip tendency due to different parameter sets enable us to assess the impact of uncertain information for a specific parameter on the eventual uncertainty of stability prediction.

We illustrate this approach with a case study from the North Alpine Foreland Basin. The geothermal power plant in Unterhaching has operated for almost two decades. Currently, it is used solely for district heating, but it was previously employed for power generation. During its operation, it has experienced several hundred microseismic events around the reinjection well that are attributed to the operation.

We set up a 3D thermo-hydro-mechanical model around the reinjection well to model the response of the stress field to ongoing fluid reinjections. The model geometry is based on a 3D seismic survey that includes six lithological units, each populated with corresponding rock properties. Additionally, pore fluid overpressures, as observed locally in the North Alpine Foreland Basin, are incorporated. Different stress states based on data records and model results are calibrated. Furthermore, the hydraulic properties of the faults are assumed to be either sealing or conducting. Several model scenarios allow us to eventually identify those parameter sets that are in agreement with observations of induced seismicity and reject those that do not align with them. Essentially, this enhances the quality of model predictions and facilitates a more accurate assessment of future operations.