Induced seismicity due to natural gas production is observed at different sites worldwide. Common understanding states that the pressure drop caused by gas production leads to compaction, which affects the stress field in the reservoir and the surrounding rock formations and hence reactivates preexisting faults and induces earthquakes. In this study, we show that the multiphase fluid flow involved in natural gas extraction activities should be included. We use a fully coupled fluid flow and geomechanics simulator, which accounts for stress‐dependent permeability and linear poroelasticity, to better determine the conditions leading to fault reactivation. In our model setup, gas is produced from a porous reservoir, divided into two compartments that are offset by a normal fault. Results show that fluid flow plays a major role in pore pressure and stress evolution within the fault. Fault strength is significantly reduced due to fluid flow into the fault zone from the neighboring reservoir compartment and other formations. We also analyze scenarios for minimizing seismicity after a period of production, such as (i) well shut‐in and (ii) gas reinjection. In the case of well shut‐in, a highly stressed fault zone can still be reactivated several decades after production has ceased, although on average the shut‐in results in a reduction in seismicity. In the case of gas reinjection, fault reactivation can be avoided if gas is injected directly into the compartment under depletion. However, gas reinjection into a neighboring compartment does not stop the fault from being reactivated.

In 2013, fluid injection during the St. Gallen deep geothermal project, Switzerland, induced hundreds of seismic events, including a ML 3.5 earthquake on a fault hundreds of meters away from the well. Recent studies have suggested the direct pressure effect through permeable hydraulic connections and poroelastic effects as possible mechanisms for inducing seismicity on distant faults. In St. Gallen, operational, seismic, and earthquake data are available to investigate the underlying physical mechanisms using a numerical model. The results show that Coulomb stress changes at the fault can be 3 orders of magnitude greater when a hydraulic connection is present. Combining this with several field observations, we conclude that the direct pressure effect was more likely the predominant mechanism behind the seismicity induced in St. Gallen. The detection of hydraulic connections may be important for future projects as pressure can be driven far from the well and reactivate remote faults.

Abstract. In July 2013, the city of St. Gallen conducted a deep geothermal project that aimed to exploit energy for district heating and generating power. A few days after an injection test and two acid stimulations that caused only minor seismicity, a gas kick forced the operators to inject drilling mud to combat the kick. Subsequently, multiple earthquakes were induced on a fault several hundred meters away from the well, including a ML 3.5 event that was felt throughout the nearby population centers. Given the occurrence of a gas kick and a felt seismic sequence with low total injected fluid volumes (∼1200 m3), the St. Gallen deep geothermal project represents a particularly interesting case study of induced seismicity. Here, we first present a conceptual model based on seismic, borehole, and seismological data suggesting a hydraulic connection between the well and the fault. The overpressurized gas, which is assumed to be initially sealed by the fault, may have been released due to the stimulations before entering the well via the hydraulic connection. We test this hypothesis with a numerical model calibrated against the borehole pressure of the injection test. We successfully reproduce the gas kick and spatiotemporal characteristics of the main seismicity sequence following the well control operation. The results indicate that the gas may have destabilized the fault during and after the injection operations and could have enhanced the resulting seismicity. This study may have implications for future deep hydrothermal projects conducted in similar geological conditions with potentially overpressurized in-place gas.

<p>We present a synthetic inversion study illustrating two approaches which help to deal with heterogeneous sensitivities in 3D frequency-domain controlled-source electromagnetic inverse problems.</p> <p>Using edge-based vector finite-element approximations on tetrahedral meshes and a preconditioned non-linear conjugate gradient algorithm, we invert for impedance tensor elements generated by a set of two coincident perpendicularly oriented horizontal electric or horizontal magnetic dipole sources. Depending on the number and locations of sources and the choice of impedance tensor components used for inversion, the sensitivity patterns can differ significantly. Measurement setups with a small number of sources, but many receiver stations at the surface covering near-field, transition zone and far-field, are often deployed for land-based controlled-source electromagnetic measurements. Such a setup can result in accumulated sensitivities close to the sources and receivers, which implies strongest model updates in these regions and can mislead the inverse algorithm to a search direction, where no physically meaningful model can be produced nor the data are fitted.<br /> <br />In order to mitigate the influence of strong sensitivities near sources and receivers on the inversion process, we apply an efficient preconditioner and customisable weights in the model regularisation matrix. The preconditioner is updated with the Broyden-Fletcher-Goldfarb-Shanno algorithm using the diagonal of the approximate Hessian matrix as start preconditioner. The latter is computationally expensive to obtain, but aims at finding a favourable search direction for the inverse algorithm already in early iterations and distributing the model update more evenly in the domain. To account for the sensitivity loss with depth, we implemented a depth weighting functional in the model regularisation term. The approach is based on counteracting the exponential and geometrical decay of the electromagnetic fields with depth and distance from the sources. In practical, we increase the smoothing in the shallow part of the model close to the source locations, where no structure is expected. We present synthetic examples indicating that this approach is an efficient way of helping the inversion to converge, obtaining a reliable model and resolving structure at depth.</p>

In the development of deep geological repositories (DGRs), performance assessment modeling is used to evaluate the integrity and performance of the engineered and geological barriers for thousands or millions of years of evolution of the disposal system. To evaluate the suitability of a site for a DGR, geoscientific data from dedicated site investigation programs are integrated into site-specific assessments. This paper presents the development and implementation of a modeling workflow aimed at comparing three potential siting areas for a DGR in Switzerland from the viewpoint of long-term safety and technical feasibility. The workflow follows the guidelines of the national regulator addressing safety relevant criteria such as the barrier efficiency of the host rock and its mechanical and chemical integrity in response to repository-induced influences and the long-term stability of the repository site over geological scales. In the regulatory requirements, the role of parametric, conceptual, and scenario uncertainty has been identified as an issue of special importance in the site selection process. The assessment approach comprises a portfolio of numerical models for the simulation of solute, gas and heat transport in the repository nearfield. The modeling was performed with deterministic as well as probabilistic variants integrated in an indicator-based approach that allows the consistent comparison of the candidate sites using quantitative dimensionless performance indices. The model-based assessment of the sites allows a traceable, transparent, and verifiable implementation of the site selection process.