No abstract

SUMMARY It is suggested that fluid injection in normal faulting stress regimes can stabilize a reservoir if the stress path is high enough. This stabilization is not seen when the reservoir is significantly cooled as a result of injection. Further, a new strategy is suggested for stimulating reservoirs in shear with a reduced chance of inducing a large magnitude seismic event. The version of this methodology presented here is applicable for reverse faulting stress regimes and involves an initial stress preconditioning stage where the reservoir is cooled and the pressure increase is limited. This process reduces the horizontal total stress and thereby also the differential stress. Next, the reservoir is stimulated with a rapid increase in pore pressure, resulting in shear failure at a lower differential stress than was initially present in the reservoir. Due to the connection seen between the Gutenberg–Richter b-value and differential stress, it is suggested that reservoirs stimulated in this fashion will exhibit higher b-values and thereby also have a reduced chance of hosting a large magnitude event. It is suggested that adaptations of this methodology are applicable to both normal and strike-slip faulting stress regimes.

An earthquake occurs when rupture propagation and slip develop on fault surfaces, such that the understanding of friction and fault geometry is crucial to the understanding of the mechanics of earthquakes. A fault's reaction to stress perturbations can be characterized in a variety of ways depending on its stability (e.g., experimentally (Spagnuolo et al., 2016) and numerically (Cattania & Segall, 2021;Lapusta et al., 2000)): stage 1, the fault remains locked; stage 2, the fault undergoes slow and stable sliding; stage 3, the fault exhibits short-lived local instabilities; stage 4, the fault accelerates and runaway seismic slip occurs, often with the activation of dynamic weakening mechanisms. The transitions between three first stages can be described through a combination of Mohr-Coulomb failure and rate-and-state friction laws (Barton, 1976;Dieterich, 1979Dieterich, , 1981Ruina, 1983), such that the frictional response of the fault is dependent on the slip rate and a state variable which accounts for the evolution of the sliding surface. However, stages 3 and 4 are difficult to explore experimentally (e.g., Spagnuolo et al., 2016;Wu & McLaskey, 2019); granted, there is a significant body of numerical work concerning this topic related to fault complexity (e.g.,

Depletion‐induced seismicity can pose a problem in some fluid‐producing subsurface reservoirs, in some cases requiring production rate control in order to limit the seismicity. This study investigates the use of reservoir stimulation to reduce the depletion‐induced seismicity rate. Depletion‐induced stress and pore pressure changes are evaluated in a shale cap rock, sandstone reservoir, and shale underburden system, which contains a horizontal well, all modeled in plane strain conditions. The seismicity rate is then predicted based on an existing seismicity model and is found to be dependent on the direction the well is drilled in with respect to the principal stresses. The case where the reservoir has first been stimulated is compared to the case where stimulation has not been performed (using the same production rates) for normal, reverse, and strike‐slip faulting stress regimes. Seismicity is reduced in the case of reservoir stimulation for both reverse and strike‐slip faulting stress regimes. The seismicity rate is only slightly reduced for the normal faulting stress regime. Stimulation also increases the distance that changes in pore pressure dominate over poroelastic stress changes in the reservoir. Further, it is found that the optimal orientation of a horizontal well, in terms of induced seismicity, is parallel to the minimum principal stress in normal faulting stress regimes and parallel to the maximum principal stress in reverse faulting stress regimes. The orientation of a horizontal well determines where the seismicity is located in a strike‐slip faulting stress regime.

<p>The stress state in the subsurface has been shown to be a hugely important parameter for a wide variety of considerations related to seismicity, both natural and anthropogenic. Industrial operations have been shown to be capable of influencing this subsurface state of stress, as most notably evidenced by instances of induced seismicity related to mining, reservoir impoundment, and reservoir-engineering applications such as production- and injection-induced seismicity related to pore pressure increase. The recognized significance of the stress state for many industrial operations as well as operators' proven ability to influence it, has led to the notion that the stress state can be intentionally preconditioned prior to an operation to that operation’s eventual benefit. The idea of preconditioning was first introduced by the mining industry in the late 1950's as a way to improve rockburst conditions in mines, by blasting to relieve stress in near-face regions. This idea of stress preconditioning has since been extended to the petroleum industry, beginning in the 1970’s and typically focused around hydraulic fracturing. Enhanced Geothermal Systems (EGSs) have been plagued by a number of instances of high-profile induced seismicity, most notably in Basel, Switzerland. This has led to the realization that the development of new reservoir stimulation techniques is crucial for the development of EGS. Here, we propose that the effective stress along a fault intersected by an EGS well be preconditioned prior to stimulation through an extended period of fluid production. Following this production phase, the fault is stimulated through high-pressure injection. Through analytical models related to pressure diffusion, earthquake nucleation, and earthquake rupture, it is suggested that this methodology would result in the halting of near-well nucleated events as they rupture towards the zone of reduced pore pressure. These models assume a constant permeability, linear slip weakening, and a near-critically stressed fault. The investigation is supported by a scaling analysis, shedding light on the suggested required magnitude of the preconditioning phase.</p><p><img src="https://contentmanager.copernicus.org/fileStorageProxy.php?f=gnp.336cb93f40e167920902461/sdaolpUECMynit/22UGE&app=m&a=0&c=c8f6623422eb35c9bb95aa37562056a1&ct=x&pn=gnp.elif&d=1" alt=""></p><p>Figure: A schematic illustrating the proposed strategy. On the left, a well (A) is drilled into a fault, shown in plane-view. Fluid is produced from this well, reducing the pore pressure. This production is continued for a significant amount of time, allowing the reduction of pore pressure to be significant and far reaching (D). The well is then stimulated with a short high pressure burst of injection (B). The stimulated zone shears near the well in this high-pressure zone, but is halted by the low-pressure zone (C). The corresponding pore pressure as a function of the radial distance is plotted on the right.</p>