Search

In areas of induced seismicity, earthquakes can be triggered by stress changes due to fluid injection and static deformation from fault slip. Here we present a method to distinguish between injection‐driven and earthquake‐driven triggering of induced seismicity by combining a calibrated, fully coupled, poroelastic stress model of wastewater injection with interpretation of a machine learning algorithm trained on both earthquake catalog and modeled stress features. We investigate seismicity from Paradox Valley, Colorado as an ideal test case: a single, high‐pressure injector that has induced thousands of earthquakes since 1991. Using feature importance analysis, we find that injection‐driven earthquakes are approximately 22± $\pm $5% of the total catalog but act as background events that can trigger subsequent aftershocks. Injection‐driven events also have distinct spatiotemporal clustering properties with a larger b‐value, closer proximity to the well, and earlier occurrence in the injection history. Generalization of our technique can help characterize triggering processes in other regions where induced seismicity occurs. Plain Language Summary: The Paradox Valley Unit, Colorado in the central United States has had a remarkable increase in seismicity coincident with over 8 million cubic meters of brine fluid injection since 1991, inducing thousands of earthquakes within an aquifer 4.5 km below the surface. We use a physics‐based model of the Earth combined with statistical and machine learning techniques to help discern which earthquakes are triggered by other earthquakes and which earthquakes are directly triggered by the stress changes from the well as well as their comparative characteristics. Discerning which earthquakes are directly caused from pressure changes due to the fluid injected by the well can inform our understanding of earthquake physics and provide useful information to operators of energy production sites. Key Points: Physics‐based models with interpretable machine learning can identify likely triggering mechanism of earthquakes within an induced sequenceWith this technique, we show that only 22土5% of earthquakes in Paradox Valley are primarily injection‐drivenInjection‐driven events have a larger b‐value, are closer to the well, and occur earlier in the injection history [ABSTRACT FROM AUTHOR]

Aged hydrocarbon wells, if proper care is not ensured, can crack, get corroded, and leak subsurface fluids. Permian Basin in Texas, home to thousands of such wells, has seen numerous blowouts and wastewater leaks. Our study employs surface deformation derived from satellite observations, and injection well records to investigate these events. The results reveal an over‐pressurized wastewater aquifer producing a surface uplift of 20 cm/yr, likely due to wastewater being injected tens of kilometers away. Focusing on a January 2022 blowout resulting in 3 cm subsidence in 2 weeks, our geophysical model suggests aquifer over‐pressurization as the cause. With an excess pressure of over 3 MPa in the aquifer, several more such blowouts are possible in the near future. This research highlights the urgent need to better understand the impact of subsurface fluid injection and calls for prompt action to mitigate the environmental effects of oil and gas production. Plain Language Summary: Wastewater, a byproduct during oil extraction, is generally injected back into the ground. Permian Basin has witnessed several incidents of leakage of wastewater in the last 3 years. Recently in January 2022, huge amounts of wastewater were expelled at a high pressure from an old well. We used satellite data and wastewater injection data to understand the cause of these events. We found that the wastewater injection happening nearly several kilometers away is responsible for these leakages. We also discovered a highly pressurized wastewater lake below the surface in this region using geophysical modeling. Due to high pressures, the land in this region rose by 40 cm in just 2 years. Meanwhile, a part of this region sunk by 3 cm because of the leakage in January 2022. Our findings raise concerns about the potential for more leakages in the near future if action is not taken. Key Points: We utilized satellite observations of land surface deformation to explore the subsurface anomalies in the Permian BasinWe established a significant link between wastewater injection and oil well blowouts in the Permian BasinWe discovered an over‐pressurized aquifer responsible for the reported blowouts, with a potential for more blowouts in the near future [ABSTRACT FROM AUTHOR]

Laumontite is a common and potentially frictionally unstable hydrothermal alteration product present in deep faults of the Gonghe EGS reservoir. We characterize the friction‐stability characteristics of synthetic laumontite gouge under in situ reservoir conditions. The pure laumontite gouge is frictionally strong (μ = 0.73–0.98) and the quartz/laumontite mixture (1:1) is generally less strong (μ = 0.73–0.78) under experimental conditions (Pc = 95 MPa, T = 90–250°C, Pf = 0–90 MPa). The shear velocity was stepped between 6.1, 0.61, then 0.061 μm/s for our experiments. For both gouges, the friction coefficient is independent of temperature and increases with elevated pore pressures. The pure gouge and mixture are strongly velocity‐weakening over a broad range in temperatures (∼90–220°C) and excess pore pressures (0–90 MPa) relevant to the Gonghe stimulation. Microearthquakes (MEQs) observed during stimulation are confined to within the broad depth range of inferred frictional instability—although fluid overpressures are also limited to this region. The observation that laumontite mixtures are frictionally unstable over a broad range of pressures and especially temperatures representative of EGS reservoirs and insensitive to the presence of the coexisting mineral phase (quartz) suggests its presence is a strong indicator of potential seismic hazard. Plain Language Summary: Laumontite is a very low‐grade altered mineral that can easily occur in fractures or faults in granite, basalt, or sandstone. Laumontite is widely developed in the Gonghe geothermal reservoir of Western China. Fluid injection into deep geothermal rock mass may reactivate subsurface faults containing altered minerals and cause earthquakes. Hence, we conducted laboratory experimental analysis on the frictional characteristics of simulated laumontite gouge to further understand the impact of fluid injection on the triggering of deep fault earthquakes. These experiments were performed at conditions reflecting the temperature and pressure of the water injection depth of the Gonghe geothermal reservoir. The results showed that the fault's strength and friction stability strongly depend on pore pressure and temperature. Our study emphasizes the significant role of the altered mineral laumontite in controlling fault strength and stability, as well as its potential for inducing earthquakes. A possible implication of this work is that when selecting geothermal resource targets that require fluid‐injection operations, it is best to avoid laumontite‐rich sites or reservoir sections. Key Points: We report the first evaluations of frictional stability properties for laumontite gouge and mixtures under hydrothermal conditionsGouges are strongly velocity‐weakening over a broad range of temperatures and insensitive to pressures and relative proportionsThis instability field is coincident with P‐T conditions typical for EGS and hence an indicator of ubiquitous seismic hazard [ABSTRACT FROM AUTHOR]

La Fossa Caldera at Vulcano (Italy) has been showing signs of unrest since September 2021. To investigate this phenomenon, we conducted an analysis of geodetic and seismological data from July to December 2021. In particular, we analyzed Multi Temporal Interferometric Synthetic Aperture Radar and Global Navigation Satellite System data, showing a pronounced elliptical uplift signal, which we elaborated using analytical source modeling. Additionally, seismic data were used to identify seismicity associated with hydrothermal system activity and assess its temporal evolution. The results indicate that the observed deformation is consistent with the expansion of the hydrothermal system within the La Fossa Caldera. These findings align with the analysis of seismic data, revealing signals indicative of hydrothermal activity, such as Very Long Period events. The results suggest that the ongoing phenomenon since 2021 represents a hydrothermal unrest, similar to the one observed during the late 1970s to early 1990s. Plain Language Summary: La Fossa Caldera at Vulcano Island, part of the Aeolian Islands archipelago in Italy, has shown an increased volcanic activity since September 2021. This activity is characterized by an increase in fumarole temperatures, massive gas emissions, as well as a marked uplift of the crater area, accompanied by an increase in seismicity. To investigate the nature of these phenomena, an analysis of ground deformation data obtained from Multi Temporal Interferometric Synthetic Aperture Radar and Global Navigation Satellite System measurements is presented. Additionally, a detailed analysis of data recorded by the seismic network on Vulcano Island has been conducted. The results indicate that these anomalies can be attributed to the expansion of the hydrothermal system, a phenomenon previously observed in the late 1970s and early 1990s. Key Points: Multi Temporal Interferometric Synthetic Aperture Radar enabled investigating localized ground deformation in the La Fossa CalderaThe analysis of local seismicity indicates it is associated with the injection of fluids into conduit‐like structuresThe modeled source of ground deformation associated with the 2021 unrest is consistent with the pressurization of the hydrothermal system [ABSTRACT FROM AUTHOR]

We present a series of controlled fluid injection experiments in the laboratory on a pre‐stressed natural rough fracture with a high initial permeability (∼10−13 m2) in granite using different fluid pressurization rates. Our results show that fluid injection on a fracture with a slight velocity‐strengthening frictional behavior exhibits dilatant slow slip in association with a permeability increase up to ∼41 times attained at the maximum slip velocity of 0.085 mm/s for the highest‐rate injection case. Under these conditions, the slip velocity‐dependent change in hydraulic aperture is a dominant process to explain the transient evolution of fracture permeability, which is modulated by fluid pressurization rate and fracture surface asperities. This leads to the conclusion that permeability evolution can be engineered for subsurface geoenergy applications by controlling the fluid pressurization rate on slowly slipping fractures. Plain Language Summary: Understanding the evolution of fracture permeability during hydraulic stimulation of subsurface reservoirs is the key to characterizing fluid transport and formulating strategies to limit induced seismicity. Accordingly, there is a significant interest in deciphering how the fluid pressurization rate, a constitutive operational parameter during injection, influences the transient permeability change during fracture slip. We conducted a series of experiments in the laboratory using different fluid pressurization rates on a natural rough fracture in granite under a pre‐stressed state. The fracture had a high initial permeability. Our findings show that when fluid is injected into a fracture with a slight velocity‐strengthening frictional behavior, it causes slow slipping with significant permeability enhancement. The change in hydraulic aperture caused by slip velocity is the main reason for the temporary change in permeability, and this effect is modulated by fluid pressurization rate and fracture surface irregularities. Our results suggest that we can modulate the permeability of subsurface geoenergy reservoirs by controlling the fluid pressurization rate on slowly slipping fractures. Key Points: We conducted fluid injection experiments on a pre‐stressed natural rough fracture in granite at different pressurization ratesThe velocity‐strengthening fracture exhibits slow slip accompanied by a significant increase in permeability during fluid injectionTransient fracture permeability is controlled by injection‐induced slip velocity, modulated by pressurization rate and surface asperities [ABSTRACT FROM AUTHOR]

The factors controlling earthquake swarm duration are remain unclear, especially in the long‐living ones. A severe earthquake swarm struck the tip of the Noto peninsula, Japan. Ten M > 4.0 earthquakes occurred, and the sequence has continued more than 4 years. We investigated the spatiotemporal characteristics of the swarm using relocated hypocenters to elucidate the factors causing this long duration. The swarm consists of four seismic clusters—northern, northeastern, western, and southern—the latter of which began first. Diffusive hypocenter migrations were observed in the western, northern, and northeastern clusters with moderate to low diffusivities, implying a low‐permeability environment. Rapid diffusive migration associated with intermittent seismicity deep within the southern cluster suggests the presence of a highly pressurized fluid supply. We conclude that the nature of this fluid supply combined with intermittent seismicity from the southern cluster and a low‐permeability environment are the key causes of this long‐living swarm. Plain Language Summary: Earthquake swarms are sequences of several earthquakes occurring in a concentrated area over a given period. Unlike other major earthquakes, which have one main shock and several subsequent aftershocks, swarms lack a clear mainshock event. The causes of long‐lasting earthquake swarms are not sufficiently understood. In the northeastern tip of the Noto Peninsula in Japan, more than 20,000 earthquakes occurred between May 2018 and June 2022, including 10 events over magnitude 4.0. To understand the controlling factors of this long‐living earthquake swarm, we investigated the spatiotemporal characteristics of the swarm using high‐resolution relocated hypocenter locations. The hypocenters of the swarm are spatially separated in four clusters and initiated from the southern cluster. We also observed a diffusive pattern in hypocenter distribution, which is typical of earthquake swarms surrounding volcanoes or fluid injection wells, implying the existence of fluid as a driving factor of the swarm. In the southern cluster specifically, we found several intermittent seismic activities with rapid diffusive changes in hypocenter distribution, suggesting the presence of a highly pressurized, deep‐source fluid supply. The intermittent fluid supply from the southern cluster toward the others and the relatively low‐permeability environment are key factors in the longevity of this earthquake swarm. Key Points: An energic and long‐living earthquake swarm has been observed in the northeastern tip of the Noto peninsula, JapanObserved diffusive hypocenter migrations suggest that pore fluid pressure migration is a driving factor of the swarmIntermittent seismicity at the bottom of the initial cluster suggests that a geyser‐like fluid supply is a key factor in swarm longevity [ABSTRACT FROM AUTHOR]

Earthquakes induced by human activities can impede the development of underground resources. Significant induced events (M5) have caused both economic and human losses. The recent ML 5.6 (MW 5.1) event near Peace River, Alberta occurs in a region of in situ bitumen recovery. We find that 3.4 cm of ground deformation was caused by reverse fault slip (∼29 cm), possibly related to Peace River Arch faulting. Events are located within the shallow basement, nearby to significant wastewater injection into Paleozoic strata. We find a statistical relationship between earthquakes and injection operations. These events were likely related to the in situ bitumen development: dominantly from wastewater disposal induced pore pressure increases, with smaller poroelastic contributions from bitumen recovery. The assessment of this earthquake as induced will likely have implications for future energy development, management, and regulation—including carbon capture and blue hydrogen. Plain Language Summary: Earthquakes can be caused by underground fluid injection; cases of M5 induced events have caused damage and harm. One of the largest recorded earthquakes in Alberta (ML 5.6) occurred in a region of underground oil sand development. Here, ground shaking and deformation information are combined into an interpreted result: that ancient faults were reactivated with reverse slip. The fault slip is largely within the crystalline basement, with a small portion extending into basal sediments. Nearby injection operations dispose of petroleum‐related wastewater in these basal sediments. This earthquake was likely triggered by the injection process: injection increases pore pressure, which diffuses laterally along permeable sediments, until encountering fractured rock, which channelizes flow into the crystalline basement—the increase of pore pressure within the fault continues until reaching a critical point for slip initiation. This event likely being induced will have important implications for future operations. Key Points: On 30 November 2022 one of Alberta's largest recorded events (ML 5.6, MW 5.1) occurred in a region of in situ bitumen recoveryA well oriented fault was reactivated with reverse slip (29 cm), causing up to 3.4 cm of ground deformationThe event was likely induced by pore pressure from disposal, with smaller poroelastic stress changes from bitumen production [ABSTRACT FROM AUTHOR]

Predicting the behavior, geometry, and flow properties of subsurface fractures remains a challenging problem. Seismic models that can characterize fractures usually suffer from low spatiotemporal resolution. Here, we develop a correlative double‐difference time‐lapse full waveform inversion of continuous active source seismic monitoring data for determining high‐spatiotemporal‐resolution time‐lapse Vp models of in‐situ fracture evolution at a shallow contamination site in Wyoming, USA. Assisted by rock physics modeling, we find that (a) rapidly increasing pore pressure initializes and grows the fracture, increasing the porosity slightly (from ∼13.7% to ∼14.6%) in the tight clay formation, thus decreasing Vp (∼50 m/s); (b) the fluid injection continues decreasing Vp, likely through the introduction of gas bubbles in the injectate; and (c) final Vp reductions reach over ∼150 m/s due to a posited ∼4.5% gas saturation. Our results demonstrate that high‐resolution Vp changes are indicative of mechanical and fluid changes within the fracture zone during hydrofracturing. Plain Language Summary: Induced fractures serve as the primary flow paths for subsurface fluids in many engineered systems involving injection, extraction, or long‐term storage (e.g., groundwater, carbon storage, and enhanced geothermal exploitation). A detailed understanding of the coupled thermal‐hydro‐mechanical‐chemical processes present in fractured systems requires real‐time monitoring of the fracturing process (initiation and growth), a challenging problem. In this study, we develop an approach of time‐lapse full waveform inversion (TLFWI) of continuous active source seismic monitoring data for estimating accurate seismic velocity changes induced by fracture evolution. The localized P‐wave velocity (Vp) reductions by our TLFWI are closely consistent with various fracturing stages. These dynamic reduced Vp zones are the result of fracturing‐induced subsurface porosity in the fracture opening and then introducing gas‐bubble during the fluid injection. Key Points: Seismic velocity changes caused by near‐surface hydrofracturing are obtained from the correlative double‐difference time‐lapse full waveform inversionWe observed small velocity reduction (∼2%) due to fracture growth and larger velocity reduction (∼7%) potentially caused by gas bubblesRock physics modeling supports quantitative interpretations and allows critical insight to understand hydrofracturing dynamics [ABSTRACT FROM AUTHOR]

Microseismicity associated with fluid pressurization in the subsurface occurs during fluid injection but can also be triggered after injection shut‐in. Understanding the extent and duration of the post‐injection microseismicity is critical to limit the risk of fluid‐induced seismicity and insure the safe utilization of the subsurface. Using theoretical and numerical techniques, we investigated how aseismic slip on a fault plane evolves and stops after a fluid pressurization event. We found that the locking mechanisms controlling the arrest of aseismic slip highly depend on the initial fault stress criticality and the pressurization duration. The absolute arrest time of fault aseismic slip after injection shut‐in is proportional to the pressurization duration and increases significantly with the initial fault stress criticality. Given that microseismicity can be triggered by aseismic slip, these results provide insights into the mechanics controlling the arrest of microseismicity after fluid pressurization as a milestone toward induced seismicity mitigation strategies. Plain Language Summary: Injection of fluid in the subsurface for energy and storage applications can lead to the onset of microseismicity, and possibly to major induced seismic events. Fluid pressurization decreases the shear strength of surrounding faults and slip occurs when the in situ shear stress on a fault reaches its shear strength. The nature of slip (aseismic or seismic) depends on the rate at which it occurs and thus on the stability of the deformation. Understanding the mechanics controlling the onset and arrest of aseismic slip and the transition to seismic slip is therefore key to design mitigation strategies for the safe utilization of the subsurface. In this contribution, we investigate using theoretical and numerical techniques how aseismic slip on a fault plane nucleates and evolves in response to fluid injection and how it stops after injection shut‐in when fluid pressure relaxes. We demonstrate that critically stressed faults prior to injection can slip for a longer time after injection shut‐in than during injection and that the extent of rupture can double in size after the end of injection. These results help to quantify the duration and sphere of influence of fluid injection where microseismicity can occur during and after injection. Key Points: Quantification of fault aseismic slip that continues to propagate away from the former injection point after injection shut‐inArrest time of aseismic slip is proportional to the pressurization durationCritically stressed faults slip significantly longer and at a greater distance than marginally pressurized faults after injection shut‐in [ABSTRACT FROM AUTHOR]

To illuminate brittle and ductile fracturing of magma, we investigated bubble expansion and fracturing in two contrasting fluids: a Maxwell‐type viscoelastic fluid and a Bingham‐type yield‐strength fluid. Measurements of the complex shear modulus, G′ + iG″ (i is the imaginary unit), under small‐strain oscillation showed that both fluids are elastic (G′ > G″) with similar rigidity. Viscous behavior (G′ < G″) appeared at lower frequency in the Maxwell fluid but at larger strain in the Bingham fluid. When we injected air into the Maxwell fluid, bubbles expanded viscously at low flux, fractured in a brittle manner at high flux, and behaved transitionally at intermediate flux. In contrast, we observed no fracturing in the Bingham fluid. This demonstrates that the G′ > G″ condition is insufficient to infer that brittle fracturing can occur. Brittle fracturing of the Maxwell fluid occurred not at a critical strain rate but under decreasing strain rate and increasing stress. Plain Language Summary: The flow‐to‐fracture transition of magma is an essential process controlling eruption explosivity. This transition has been explained based on the linear Maxwell viscoelastic model: magma can flow slowly but breaks in a brittle manner under rapid deformation. The Bingham model represents another type of fluid‐solid transition: a material flows under a sufficiently large force but behaves elastically below this force. The present study investigates bubble growth and fracture by air injection in Maxwell‐type and Bingham‐type fluids, both having similar elasticity. We demonstrate that brittle fracture occurs in the Maxwell fluid at high air flux but never in the Bingham fluid. Detailed observations of the deformation until brittle fracture reveals a stress increase and strain rate decrease. The fracture onset is not determined by the critical strain rate, contrary to conventional brittle fragmentation criteria widely used in volcanology. These results will help improve brittle‐ductile transition models for complex fluids in volcanic eruptions. Key Points: Rapid air injection generated brittle fracture in a Maxwell fluid but not in a Bingham fluid, although they have similar elasticityThe observation of elasticity in small‐amplitude oscillation tests is not sufficient to infer that the fluid can fractureThe strain rate around a bubble decreased toward fracture, contrary to the conventional critical strain‐rate fracture criterion [ABSTRACT FROM AUTHOR]

We conducted fluid injection experiments on cylindrical low‐permeability granite samples with a critically stressed sawcut fault at local injection rates of 0.2 and 0.8 mL/min and confining pressures of 31 and 61 MPa. A local array of six strain gauges attached close to the faults allows us to estimate the nucleation length of each injection‐induced dynamic slip event (i.e., laboratory earthquake). We find nucleation lengths decrease from approximately 90% to <15% of the fault length with higher injection rate and increased effective normal stress. Injection‐induced laboratory earthquakes with smaller nucleation lengths show generally higher peak slip rates and larger fault slip displacements, signifying an intensified seismic hazard. Our results also indicate that initially stable fault patches may be reactivated to slip seismically by increasing injection rates. This study systematically demonstrates that higher injection rates constitute dynamic loading, which increase the seismic hazard by shrinking the earthquake nucleation length. Plain Language Summary: Earthquakes associated with fluid injection in various geo‐energy settings, such as shale gas and deep geothermal energy, have shelved many projects with great potential. However, the injection‐rate dependence of earthquake nucleation length, which characterizes the length of a slowly slipping (creeping) fault patch in preparation for a subsequent earthquake, remains to be studied. To this end, we performed fluid injection experiments on low‐permeability granite samples containing a critically stressed sawcut fault. We demonstrate that the nucleation length of injection‐induced laboratory earthquakes shrinks with higher effective normal stress and larger injection rate. We show that laboratory earthquakes induced by higher fluid injection rates are characterized by smaller nucleation lengths, faster peak slip rates, and larger fault slip displacements. Our results also suggest that a stable fault patch subject to low‐rate injection may be induced to slip seismically by increasing injection rates. Our findings may explain the higher seismic hazards associated with higher injection rates as has been frequently documented in field cases. Key Points: The nucleation length of injection‐induced laboratory earthquakes reduces with higher injection rate and greater effective normal stressInjection‐induced laboratory earthquakes with smaller nucleation lengths exhibit higher peak slip rates and larger fault slip displacementsFast‐rate fluid injection may cause initially stable fault patches of smaller sizes to slip seismically, leading to higher seismic hazards [ABSTRACT FROM AUTHOR]

Estimates of intruded magma volume are critical for forecasting volcanic unrest. Geodetic modeling can provide such estimates but is of limited use in submarine and highly vegetated settings. A complementary approach could be to use estimates of seismic moment release. In this study, we examine the moment‐volume scaling of several proximal volcanic earthquake sequences and compare it to that of injection‐induced seismicity. We find a notable similarity in scaling between the volcanic sequences, which contrasts with the broad range of responses exhibited by anthropogenic injection‐induced sequences. This may imply an underlying similarity in the geologic conditions for volcanoes that is distinct from induced seismicity settings. It could also allow for estimates of intruded volume to be made without geodetic information. This provides further insight into the factors controlling seismogenesis in these different settings and has implications for volcano seismology and injection‐induced seismicity hazard estimation. Plain Language Summary: Knowing how much magma is beneath a volcano before or during an eruption can help forecast its behavior and aid evacuation or relief efforts. Currently, satellite images or sensitive measurements of the ground inflation are used to model the amount of magma, but this isn't always possible if the volcano is underwater or shrouded by tree cover. A proposed alternative to this is to use the size and number of earthquakes as a gauge on the amount of volume change occurring underground. The relationship between the total amount of energy released by earthquakes and volume change has been a key research area of earthquakes triggered in the injection of fluids by industries like geothermal energy or hydraulic fracturing (i.e., fracking). We study several well‐recorded volcanoes and see that there is a similarity in the observed volume change and the amount of energy released from associated earthquakes. This is very different from the many examples of anthropogenic injection‐induced seismicity, which show a very broad spread. This implies that the similar conditions under volcanoes lead to similar responses to volume changes. This helps us better understand and thus potentially mitigate the hazards of both volcanoes and industrial fluid injection. Key Points: We observe a similar moment‐volume scaling for volcano seismicity, contrasting with the range exhibited by injection‐induced examplesThe results show that scaling between moment release and volume change can be used to estimate intruded volumesThis also provides insights into the geologic factors which control the crust's response to volume changes [ABSTRACT FROM AUTHOR]

Abstract Microseismicity associated with fluid pressurization in the subsurface occurs during fluid injection but can also be triggered after injection shut‐in. Understanding the extent and duration of the post‐injection microseismicity is critical to limit the risk of fluid‐induced seismicity and insure the safe utilization of the subsurface. Using theoretical and numerical techniques, we investigated how aseismic slip on a fault plane evolves and stops after a fluid pressurization event. We found that the locking mechanisms controlling the arrest of aseismic slip highly depend on the initial fault stress criticality and the pressurization duration. The absolute arrest time of fault aseismic slip after injection shut‐in is proportional to the pressurization duration and increases significantly with the initial fault stress criticality. Given that microseismicity can be triggered by aseismic slip, these results provide insights into the mechanics controlling the arrest of microseismicity after fluid pressurization as a milestone toward induced seismicity mitigation strategies.

This letter compares the predictions of two expressions proposed for the porosity evolution in the context of rate and state friction. One (Segall & Rice, 1995, https://doi.org/10.1029/95jb02403) depends only on the sliding velocity; the other (Sleep, 1995, https://doi.org/10.1029/94jb03340) depends only on the state variable. Simulations of both are similar for velocity stepping and slide‐hold‐slide experiments. They differ significantly for normal effective stress jumps at constant sliding velocity. Segall and Rice (1995, https://doi.org/10.1029/95jb02403) predicts no change in the porosity; Sleep (1995, https://doi.org/10.1029/94jb03340) does. Simulation with a spring‐block model indicates that the magnitude of rapid slip events is essentially the same for the two formulations. Variations of porosity and induced pore pressure near rapid slip events are similar and consistent with experimental observations. Predicted porosity variations during slow slip intervals and the time at which rapid slip events occur are significantly different. The simulation indicates that changes in friction stress due to pore pressure changes exceed those due to rate and state effects. Plain Language Summary: The interaction of pore fluid and deformation is important for many geological processes and for technological applications involving fluid injection, such as carbon sequestration, disposal of waste fluids from hydraulic fracturing, and geothermal stimulation. Often these applications have induced earthquakes that have raised cause for concern. An important element of the interaction of pore fluid and deformation is the evolution of porosity, that is, the ratio of volume of voids to the total volume, with the slip on a fault surface. If the porosity increases more rapidly than pore fluid can diffuse out of pores, the pore pressure decreases and increases the frictional resistance to slip; conversely, porosity decreases can increase the pore pressure and reduce the frictional resistance to slip. This paper compares two proposed descriptions for porosity evolution with slip. Both have similar predictions for standard laboratory experiments but differ significantly for changes of pore fluid pressure at constant sliding velocity. In simulations with a simple spring ‐ block model both predict changes in pore fluid pressure with time near rapid slip events that are consistent with experiments. The predicted changes in porosity differ in the intervals of slow slip between rapid slip events. Key Points: Two formulations that have been suggested for porosity evolution give similar results for simulated velocity stepping and slide‐hold‐slide testsThe two formulations give significantly different results for effective normal stress changes at constant slip velocityA spring‐block simulation shows that both formulations predict pore pressure near rapid slip events consistent with laboratory observations [ABSTRACT FROM AUTHOR]

Large‐scale CO2 sequestration into geological formations has been suggested to reduce CO2 emissions from industrial activities. However, much like enhanced geothermal stimulation and wastewater injection, CO2 sequestration has a potential to induce earthquake along weak faults, which can be considered a negative impact on safety and public opinion. This study shows the physical mechanisms of potential seismic hazards along basement faults driven by CO2 sequestration under variation in geological and operational constraints. Specifically we compare the poroelastic behaviors between multiphase flow and single‐phase flow cases, highlighting specific needs of evaluating induced seismicity associated with CO2 sequestration. In contrast to single‐phase injection scenario, slower migration of the CO2 plume than pressure pulse may delay accumulation of pressure and stress along basement faults that may not be mitigated immediately by shut‐in of injection. The impact of multiphase flow system, therefore, needs to be considered for proper monitoring and mitigation strategies. Plain Language Summary: Large‐scale geological CO2 storage has been considered one of key mitigation strategies to reduce greenhouse gas emission. Geological formation is evaluated to decide the presence of preexisting faults prone to failure prior to CO2 injection, but the hidden and/or undetected faults are sometimes observed due to induced seismicity detected after fluid injection. Hence, comprehensive mechanistic study is required to study the impact of CO2 injection into geological formations in the presence of faults. This study demonstrates that earthquakes can strike unexpectedly along the faults at large distances and depths even after terminating well operations, caused by stress transfer through rock pores or interaction among injected CO2, preexisting pore fluids and rock matrix. Contrast of multiphase fluid properties between CO2 and brine can generate larger and delayed perturbations in pressure and stress fields along faults that may maintain the potential of seismicity after shut‐in. Elastic energy can transfer to low‐permeability or hydraulically isolated basement faults, such that geomechanical simulations based on comprehensive geophysical data are required to detect deformation‐driven signals and mitigate potential seismic hazards associated with geological CO2 sequestration. Key Points: Hydro‐mechanical deformation of the CO2‐saturated reservoir can destabilize basement faults, even hydraulically isolated from the reservoirTwo‐phase flow system can generate prolonged and large stress changes along the faults due to contrast of flow and transport characteristicsTerminating injection may not mitigate elastic energy in a basement immediately that can induce post shut‐in earthquakes along deep faults [ABSTRACT FROM AUTHOR]

One of the challenges associated with subsurface high‐pressure fluid injections is the estimation of the seismic hazard and its spatial footprint. Field data have shown that the spatial footprint typically varies significantly between injections into the basement and injections above basement. Here, we show that varying degrees of spatial correlations in porosity or log(permeability) can explain this observation. Using high‐resolution well‐log data, we first show that porosity within the basement tends to follow a power‐law scaling, S(k) ∼ 1/kβ, with β ≈ 0.9, while above basement β > 1.4. Using this in a novel conceptual model, we show that β controls the spatial footprint of fluid‐induced seismicity such that large values of β lead to more seismic activity at large distances and a higher variability in the spatio‐temporal migration of seismic events, hence, explaining the field observations. Our findings indicate that correlations in log(permeability) need to be incorporated in the seismic hazard assessment. Plain Language Summary: Fluid‐induced earthquakes are a side effect of industrial operations such as hydraulic fracturing and enhanced geothermal systems, where high‐pressure fluids are pumped into the earth's crust to increase oil and/or gas flow to a well from petroleum‐bearing rock formations or to improve permeability in underground geothermal reservoirs. The induced seismicity has become a concern for industry and nearby residents. One of the challenges associated with this seismic hazard is answering the question of how far away from the injection site seismicity can be induced. Previous observations showed that if the fluids are pumped into the crystalline basement, the occurrence of the seismicity is typically very localized in space, while otherwise it is not. Here, we introduce a novel model that explains this behavior as a consequence of a varying degree of correlations in permeability. By analyzing high‐quality permeability data, we directly show that the degree of correlations in the crystalline basement is indeed significantly different from formations above the basement. Our findings can be directly incorporated in any seismic hazard assessment to improve the safety of workers and residents. Key Points: Spatial correlations in permeability can control the spatial localization of fluid‐induced seismicitySpatial correlations in permeability tend to be different within the basement and above basementThe variability in the migration profile of fluid‐induced seismicity is controlled by the permeability field [ABSTRACT FROM AUTHOR]

Earthquakes resulting from hydraulic fracturing (HF) can have delayed triggering relative to injection commencement over a varied range of time scales, with the majority of M ≥ 4 mainshocks occurring near/after well completion. This poses serious challenges for risk mitigation and hazard assessment. Here, we document a high‐resolution, three‐dimensional source migration process with delayed mainshock triggering that is controlled by local hydrogeological conditions near Fox Creek, Alberta, Canada. Our results reveal that poroelastic effects might contribute to induced seismicity, but are probably insufficient to activate a large fault segment not critically stressed. The rapid pore‐pressure build‐up from HF can be very localized and capable of producing large, felt earthquakes if adequate hydrological paths exist. We interpret the delayed triggering as a manifestation of pore‐pressure build‐up along pre‐existing faults needed to facilitate seismic failure. Our findings can explain why so few injection operations are seismogenic. Plain Language Summary: Fluid injection‐induced earthquakes (IIE), especially those M ≥ 4 mainshocks, are often observed to occur near or after well completion. Such delayed triggering relative to injection commencement poses serious challenges for both regulators and the energy industry to establish an effective mitigation strategy for the potential seismic risk. In this study, we reveal a high‐resolution, complex three‐dimensional pattern of IIE migration near Fox Creek, Alberta, Canada. The observed first‐outward‐then‐inward IIE sequence highlights the significance of hydrogeological networks in facilitating fluid pressure migration and the associated seismic failure. The detailed spatiotemporal distribution of IIE suggests that the effect of pore‐pressure build‐up from hydraulic fracturing (HF) can be very localized. The delayed triggering is a combined result of the fluid pressure migration and the current stress state of the hosting fault system away from the HF wells. The findings from this study also provide plausible explanations on why only a very limited number of fluid injections are seismogenic. Key Points: We document a complex 3D source migration process with delayed mainshock triggering that is controlled by a local hydrogeological settingPoroelastic effects contribute to induced events but are probably insufficient to activate a large fault segment not critically stressedRapid pore‐pressure build‐up can be very localized and lead to large earthquakes if adequate hydrological paths exist [ABSTRACT FROM AUTHOR]

The 2011 Mw 5.7 Prague earthquake is the second largest induced earthquake in Oklahoma, and occurred after decades of wastewater disposal. The local geological structure that led to the triggering of this large earthquake is not well understood. In this study, tomographic inversion of seismic data recorded by a dense local seismic network resulted in a high‐resolution 3D velocity model with three major layers. The model clearly illuminates the geometry and characteristics of the Meeker‐Prague Fault that hosted the 2011 Prague sequence. A conceptual model is proposed to link the tomographic structure to the triggering process of the sequence. The low‐permeability second layer at ∼1.5–3.5 km may be the key that delays the occurrence of the first sizeable earthquake after decades of wastewater injection. However, a low‐shear‐velocity zone within this layer at the intersection of two major faults could have provided a fluid pathway to facilitate downward fluid propagation. Plain Language Summary: The recent surge of seismicity in the central United States since 2008 has largely been attributed to wastewater disposal from oil and gas production. The 6 November 2011 Mw 5.7 earthquake near Prague, is the second largest earthquake ever recorded instrumentally in the state of Oklahoma. Previous studies mainly focused on the characterization of seismicity in the 2011 Prague earthquake sequence and found that the majority of earthquakes occurred along the Meeker‐Prague fault, which is a ∼20 km splay fault off the Wilzetta fault zone. In this study, we apply the seismic traveltime tomography technique on data recorded by a dense local seismograph network to image the fine upper crustal velocity structure in the Prague fault zone. Our results clearly mapped zones with significant velocity anomalies in the fault zone compared to the surrounding area. And the velocity anomalies coincide with the clusters of seismicity, indicating that fluid injection probably played an important role in altering the properties of the fault zone and therefore promoting the occurrence of earthquakes. This study highlights that it is critical to investigate the fault zone structure to advance our understanding of the process of pore pressure diffusion and fluid migration in triggering of induced seismicity. Key Points: We develop a fine fault zone velocity structure for the 2011 Mw 5.7 Prague earthquake and relocate the earthquakes using the 3D modelsThe results depict the Meeker‐Prague Fault that hosted the majority of the larger events in the Prague sequenceWe propose a conceptual model that links velocity structure to the triggering process of the Prague sequence due to wastewater injection [ABSTRACT FROM AUTHOR]

The injection of reactive fluid into fractured reservoirs is relevant to many subsurface processes. During injection, fracture dissolution forms various patterns. Despite its importance, the evolution of geometries in radial fractures remains unexplored experimentally. Here, by flow‐visualization experiments, we characterize dissolution patterns and show that the patterns shift from compact to wormhole to uniform as flow rate increases. For the first time, we observe dissolution hotspots with maximum local dissolution rate in radial fractures, occurring at a distance from the inlet. We elucidate the key role of gravity for the occurrence of hotspots. The locations of hotspots correspond to the state that the effects of transport in the horizontal and vertical directions are comparable. We establish a theoretical model for predicting locations of hotspots in good agreement with experiments. This work demonstrates transitions of dissolution patterns in fractures and reveals the underlying mechanism for dissolution hotspots previously unidentified. Plain Language Summary: Rock fractures commonly serve as dominant pathways for the fluids flow in the Earth's crust. If the flowing fluid is reactive, such fluid injection into fractured reservoirs expands the fracture aperture and produces complex dissolution patterns. It then significantly affects flow pathways and is critical for many subsurface processes, including geological carbon sequestration, acid‐injection enhanced oil recovery, and hydraulic fracturing of the shale formation. Here we perform flow‐through dissolution experiments in radial fractures. We report that the dissolution morphology shifts from compact to wormhole to uniform patterns as the flow rate increases. For the first time, we observe dissolution hotspots occurring at a distance from the inlet. We elucidate that even for the horizontal fractures, gravity plays a key role in the occurrence of dissolution hotspots. We show that due to the gravity effect and the stratification, the locations of hotspots correspond to the state that the effects of transport in the horizontal and vertical directions are comparable. We present a theoretical model to capture this state, and the model can reasonably predict the locations of hotspots. Our work improves understanding on how the flow rate affects the dissolution patterns and further reveals the underlying mechanism for dissolution hotspots previously unidentified. Key Points: We report the direct observation that dissolution regimes shift from the compact to wormhole to uniform patterns in radial rough fracturesWe observe dissolution hotspots for the first time in radial flow and elucidate the key role of gravity even for horizontal conditionsWe establish a model that can predict well the locations of dissolution hotspots and reveal the underlying mechanism for the hotspots [ABSTRACT FROM AUTHOR]

The possibility of controlling induced earthquake magnitude through managed metering of water injection has yet to be rationalized. Mechanisms of reducing magnitudes of induced events through cycled fluid injection remain unclear. To explore such mechanisms and this possibility, we report experiments with water injection into laboratory faults. Water injection results in early triggering for both single and cycled injection. However, the maximum moment magnitude and total energy of the repeating induced earthquakes during cycled water injection are both lower than those of induced earthquakes induced with continuous injection and for natural earthquakes without water injection. Higher permeability of the host reduces the number of injection‐induced earthquakes but increases their moment. With an increasing number of water injection cycles, both maximum moment and total energy decrease, particularly as permeability decreases, while the number of induced events increases. The moment magnitude of induced events can thereby be controlled through cycled fluid injection. Plain Language Summary: Fluid injection‐triggered earthquakes have been documented worldwide. Although it is clear that fluid pressure triggers fault slip and results in induced earthquakes, the possibility of controlling induced earthquakes through managed metering of water injection has not been rigorously examined. We performed experiments with single and cycled injection into laboratory faults. Our results show that cycled water injection into a fault could successfully reduce both the maximum magnitude of the individual induced seismic events and the total radiated energy—by setting appropriate water injection schedules. The magnitude and number of induced events are dependent on the number of water injection cycles and rock permeability. These results may provide a better understanding of managing and preventing large induced earthquakes. Key Points: Water injection results in early occurrence of the fault seismic slip for both single and cycled injectionCycled water injection could reduce both the maximum magnitude of the individual induced seismic events and the total released energyBoth maximum moment and total energy decrease with an increasing number of water injection cycles [ABSTRACT FROM AUTHOR]

We investigate the relative importance of injection and production on the spatial‐temporal distribution of induced seismicity at the Raft River geothermal field. We use time‐series of InSAR measurements to document surface deformation and calibrate a hydro‐mechanical model to estimate effective stress changes imparted by injection and production. Seismicity, located predominantly in the basement, is induced primarily by poroelastic stresses from cold water reinjection into a shallower reservoir. The poroelastic effect of production from a deeper reservoir is minimal and inconsistent with observed seismicity, as is pore‐pressure‐diffusion in the basement and along reactivated faults. We estimate an initial strength excess of ∼20 kPa in the basement and sedimentary cover, but the seismicity rate in the sedimentary cover is four times lower, reflecting lower density of seed‐points for earthquake nucleation. Our modeling workflow could be used to assess the impact of fluid extraction or injection on seismicity and help design or guide operations. Plain Language Summary: It is important to understand the mechanisms behind human‐induced earthquakes, whether they are beneficial in the context of generating fractures for effective geothermal energy systems, or hazardous in the case of large earthquakes that may cause structural damage. Here, we present a case study of the Raft River geothermal field, which has operated since 2007 and generated earthquakes since 2010. Our objective is to understand how hot water, which is extracted from a deep sedimentary layer, and cold water, which is injected into a shallower layer, contribute to the observed seismicity, which have primarily occurred underneath the extraction layer. We construct a model to determine how the injection and extraction affect underground fluid pressures, and how these pressures impart stresses throughout the subsurface. This model is calibrated using InSAR, a satellite‐based technique which provides precise measurements of the Earth's surface deformation in time. The results show that the shallow cold water injection is the main culprit behind deep earthquakes, because of the stresses imparted by the fluid pressure increase, whereas the effects of deep extraction is negligible. Our modeling workflow could be used to assess the impact of fluid extraction or injection on seismicity and help guide operations. Key Points: A hydro‐mechanical model of effective stress changes from injection and production is calibrated using surface deformation from InSARSeismicity, located predominantly in the basement, is primarily induced by poroelastic stress changes from cold water re‐injectionSimilar stress changes trigger four times more earthquakes per unit volume in the basement than in the sedimentary cover [ABSTRACT FROM AUTHOR]

Aftershock cascades play an important role in forecasting seismicity in natural and human‐made situations. While their behavior including the spatial aftershock zone has been the focus of many studies in tectonic settings, this is not the case when fluid flows are involved. Using high‐quality seismic catalogs, we probe aftershocks dynamics in five settings influenced by fluids: (a) induced seismicity in Oklahoma and Kansas, (b) natural swarms in California and Nevada, and (c) suspected swarms in the Yuha Desert (California). All settings exhibit significant aftershock behavior highlighting the importance of event‐event triggering processes. The spatial aftershock zones scale with mainshock magnitude as expected based on the rupture length. While (a) and (b) show a rapid decay beyond their rupture length, (c) exhibits long‐range behavior suggesting that fluid migration might not be the dominant mechanism. We also find that the scaling of aftershock productivity with mainshock magnitude together with the Gutenberg‐Richter b‐value might allow to distinguish between natural swarms and induced seismicity. Plain Language Summary: While it is known that fluid injection operations can induce seismic activity, it has remained unclear how this activity compares to their natural counterpart, seismic swarms driven by natural fluid flows. The latter are typically characterized by the absence of a dominant event within the seismic sequence, while exhibiting other characteristics consistent with tectonic sequences including aftershock triggering. Our analysis of high‐quality seismic catalogs for both types of fluid‐driven seismicity shows that both exhibit a significant amount of aftershocks arising from "secondary" processes (i.e. stress‐based event‐event triggering as an indirect consequence of fluid injections) leading to spatially localized aftershocks zones. Yet, the trade‐off between the seismic productivity relation, which refers to the average increase in the number of aftershocks with the magnitude of their trigger, and the distribution of earthquake magnitudes controls the relative role of small compared to large triggers and we find that aftershock triggering is much more dominated by smaller events in the induced setting. Both findings are of direct importance for earthquake forecasting and seismic hazard assessment. Key Points: Significant event‐event triggering is present in both natural swarms and induced seismicityBoth fluid‐driven settings are characterized by narrow aftershock zonesAftershock triggering is dominated by smaller triggers in induced seismicity but much less so for natural swarms [ABSTRACT FROM AUTHOR]

Fluid pressurization of critically stressed sheared zones can trigger slip mechanisms at work in many geological processes. Using discrete element modeling, we simulate pore‐pressure‐step creep test experiments on a sheared granular layer under a sub‐critical stress state to investigate the micromechanical processes at stake during fluid induced reactivation. The global response is consistent with available experiments. The progressive increase of pore pressure promotes slow steady creep at sub‐critical stress states, and fast accelerated dynamic slip once the critical strength is overcome. Our multi‐scale analyses show that these two emergent behaviors correlate to characteristic deformation modes: diffuse deformation during creep, and highly localized deformation during rupture. Creep corresponds to bulk deformation while rupture results from grain rotations initiating from overpressure induced unlocking of contacts located within the shear band which, consequently, acts as a roller bearing for the surrounding bulk. Plain Language Summary: Fluids can be at the origin of catastrophic disasters, e.g., earthquakes related to deep subsurface fluid injections or lanslides triggered by short‐term changes of hydrological conditions. It is now well assumed that these phenomena originate from mechanisms taking place in critically stressed shear zones found along tectonic faults, rock mass fractures or localized deformation bands. The increase of pore pressure promotes slip along these shear zones as confirmed by numerous experimental and numerical studies. In this work, we present computer simulations that reproduce the progressive reactivation of a granular shear zone as a result of fluid pressurization. Our simulations provide grain‐scale information that improves understanding of fluid induced slip behaviors and illuminate micromechanical details of phenomenological, macroscale observations. Key Points: Fluid induced reactivation can be either stable or unstable depending on the deformation modeSlow steady creep is accommodated through distributed bulk deformation at sub‐critical stress statesAccelerated dynamic slip results from intense grain rearrangements localized within the shear band [ABSTRACT FROM AUTHOR]

Earthquake swarms are manifestations of aseismic driving processes deep in the crust. We examine the spatiotemporal distribution of aseismic processes in Southern California using a 12‐years catalog of swarms derived with deep learning algorithms. In a core portion of the plate boundary region, which is not associated with elevated heat flow, we identify 92 long‐duration swarms ranging from 6 months to 7 years that constitute 26.4% of the total seismicity. We find that 53% of the swarms exhibit ultra‐slow diffusive patterns with propagating backfronts, consistent with expectations for natural fluid injection processes. The chronology of the swarms indicates that the aseismic driving processes were active at all times during 2008–2020. The observations challenge common views about the nature of swarms, which would characterize any one of these sequences as anomalous. The regional prevalence of these sequences suggests that transient fluid injection processes play a key role in crustal fluid transport. Plain Language Summary: Earthquake swarms are one of the main types of earthquake sequences. In contrast to mainshock‐aftershock sequences, swarms are believed to be driven predominantly by processes other than tectonic loading or earthquake interactions. We build a high resolution catalog of earthquakes using machine learning algorithms to study long‐duration swarms and their physical driving processes. We find nearly 100 swarms lasting 6 months to 7 years that have spatial and temporal patterns consistent with natural fluid injections in the crust. Key Points: We use deep learning algorithms to identify nearly 100 swarms in Southern California lasting 6 months to 7 years53% of the swarms have ultra‐slow diffusive patterns with propagating backfronts, suggesting they are fluid drivenThe chronology of the swarms shows that these aseismic driving processes were occurring at all times during 2008–2020 [ABSTRACT FROM AUTHOR]

Fluid injection into rock formations can either produce complex branched hydraulic fractures, create simple planar fractures, or be dominated by porous diffusion. Currently, the optimum injection parameters to create branched fractures are unknown. We conducted repeatable hydraulic fracturing experiments using analog‐rock samples with controlled heterogeneity to quantify the fluid parameters that promote fracture branching. A large range of injection rates and fluid viscosities were used to investigate their effects on induced fracture patterns. Paired with a simple analytical model, our results identify the threshold at which fracture transitions from an isolated planar crack to branched cracks when closed natural fractures exist. These results demonstrate that this transition can be controlled by injection rate and fluid viscosity. In relation to the field practices, the present model predicts slickwater and lower viscosity fluid injections promote fracture branching, with the Marcellus shale used as an example. Plain Language Summary: Hydraulic fracturing involves injecting fluid under high pressure into wells to create fractures. This is a key technique that enables economic hydrocarbon production from tight petroleum‐bearing formations whose ability to produce is otherwise too low. The same technique has also been researched for geothermal energy exploitation. Previous studies demonstrate that hydraulic fractures can either grow as planar or bifurcate into multiple branches. Branched fractures provide a larger surface area which is beneficial for increasing gas production from wells. Except for a 2019 incomplete theoretical study by the authors, the conditions that cause branching were until now experimentally unverified, and so one had to rely on intuitive guessing how to achieve it and optimize it. To address this problem, we conducted laboratory hydraulic fracturing experiments to quantify the injection rates and fluid viscosities that will cause branching. Our results indicate that fractures can transition from a single planar crack to branched cracks when the fluid viscosity is in an optimum range for a given injection rate. We also propose a theory to predict our experimental results and apply our results to field applications. This provides a useful new capability to improve the control of hydraulic fracturing. Key Points: The transition from planar to branched hydraulic fractures was observed and quantifiedBranching requires permeable, pre‐existing natural weak layers with microcracks as well as a favorable stress stateFracture branching is promoted by optimized injection rates and fluid viscosity [ABSTRACT FROM AUTHOR]

Rock materials show dramatic dynamic weakening in large-displacement (m), high-velocity (~1 m/s) friction experiments, providing a mechanism for the generation of large, natural earthquakes. However, whether such weakening occurs during induced M3-4 earthquakes (dm displacements) is unknown. We performed rotary-shear experiments on simulated fault gouges prepared from the source-, reservoir- and caprock formations present in the seismogenic Groningen gas field (Netherlands). Watersaturated gouges were subjected to a slip pulse reaching a peak circumferential velocity of 1.2-1.7 m/s and total displacements of 13-20 cm, at 2.5-20 MPa normal stress. The results show 22%-81% dynamic weakening within 5-12 cm of slip, depending on normal stress and gouge composition. At 20 MPa normal stress, dynamic weakening from peak friction coefficients of 0.4-0.9 to 0.19-0.27 was observed, probably through thermal pressurization. We infer that similar effects play a key role during induced seismic slip on faults in the Groningen and other reservoir systems. Plain Language Summary During large, natural earthquakes, rapid sliding motion on faults through several meters of displacement generates large amounts of frictional heat. This triggers mechanisms that cause crushed fault rocks to weaken drastically, allowing the rupture to propagate, and the earthquake to grow. However, whether this occurs during relatively small, human-induced seismicity, caused by hydrocarbon production or fluid injection in the subsurface, for example, is unknown. Here we simulate seismic fault slip during magnitude 3 to 4 earthquakes, using crushed rock materials derived from the seismogenic Groningen gas field (Netherlands). Our experiments show that even for shortdisplacement (dm), representative for such induced earthquakes, the fault rocks weaken significantly (22%-81%), depending on mineralogy and on the normal stress acting on the fault. These findings provide insights for the larger induced events recorded in Groningen (magnitude up to 3.6), and improve hazard assessment capabilities using numerical models. [ABSTRACT FROM AUTHOR]

While the notion that injecting fluids into the subsurface can reactivate faults by reducing frictional resistance is well established, the ensuing evolution of the slip is still poorly understood. What controls whether the induced slip remains stable and confined to the fluid‐affected zone or accelerates into a runaway earthquake? Are there observable indicators of the propensity to earthquakes before they happen? Here, we investigate these questions by modeling a unique fluid‐injection experiment on a natural fault with laboratory‐derived friction laws. We show that a range of fault models with diverging stability with sustained injection reproduce the slip measured during pressurization. Upon depressurization, however, the most unstable scenario departs from the observations, suggesting that the fault is relatively stable. The models could be further distinguished with optimized depressurization tests or spatially distributed monitoring. Our findings indicate that avoiding injection near low‐residual‐friction faults and depressurizing during slip acceleration could help prevent large‐scale earthquakes. Plain Language Summary: Fluid injections into the Earth's crust are common practice in the exploitation of subsurface energy resources such as geothermal energy, shale gas, and conventional hydrocarbons. These injections can perturb nearby fault structures and hence induce earthquakes and transient slow slip. Understanding what controls the stability (i.e., the propensity to generate earthquakes) and spatial extent of the fault response as well as identifying precarious faults is crucial to minimize the seismic hazard associated with these industrial practices. Here, we take a step toward this goal by modeling a unique experiment, in which water was injected into a natural fault and the resulting slip measured directly at depth. We first show that multiple models can explain the observations equally well, while pressure is increased in the experiment. In these models, how stable the fault response is with further injection and how large of a zone is reactivated compared to the fluid‐affected region depends on frictional properties. We then demonstrate that the slow slip response to a decrease in injection pressure further constrains the range of admissible models. Our work suggests that it may be possible to identify potentially hazardous faults with optimally designed injection tests without inducing damaging earthquakes. Key Points: Multiple frictional models with different stability reproduce the slip observed during the pressurization stage of a field experimentThe depressurization phase provides additional constraints on hydromechanical parameters and hence fault stabilityFault stability and the spatial extent of slip relative to the pressurized region depend on residual friction versus initial stress levels [ABSTRACT FROM AUTHOR]

Geoenergy and geoengineering applications usually involve fluid injection into and production from fractured media. Accounting for fractures is important because of the strong poromechanical coupling that ties pore pressure changes and deformation. A possible approach to the problem uses equivalent porous media to reduce the computational cost and model complexity instead of explicitly including fractures in the models. We investigate the validity of this simplification by comparing these two approaches. Simulation results show that pore pressure distribution significantly differs between the two approaches even when both are calibrated to predict identical values at the injection and production wells. Additionally, changes in fracture stability are not well captured with the equivalent porous medium. We conclude that explicitly accounting for fractures in numerical models may be necessary under some circumstances to perform reliable coupled thermohydromechanical simulations, which could be used in conjunction with other tools for induced seismicity forecasting. Plain Language Summary: The subsurface will play an important role in decarbonizing the economy. The transition to carbon neutrality can be accelerated by utilizing geothermal energy, returning carbon underground, and storing energy in the subsurface to offset the fluctuations in production of renewables. These low‐carbon geoenergy technologies oftentimes deal with fractured rock masses. To minimize the risks related to geoenergy projects, numerical simulations are performed to predict the response of the subsurface to fluid injection and production. Given the complexity and high computational cost of simulating fractures, fractured rock is usually treated as an equivalent porous medium. Here we investigate the validity of this simplification by comparing these two approaches. We find that even though an equivalent porous medium can reproduce the fluid pressure changes at wells, the pressure distribution within the fractured media significantly differs between the two approaches. Equivalent porous media fail to reliably predict the rock behavior when fracture spacing is within the order of the reservoir size and computer simulations should explicitly include fractures to provide reliable forecasts and reduce the risks of induced seismicity. The latter is necessary to enable a successful deployment of geoenergies to mitigate the climate crisis. Key Points: Equivalent porous media can reproduce pressure and temperature evolution at wells but not their distributions within the fractured mediaConsideration of explicit fractures affects stability and induced seismicity in a way that cannot be represented by equivalent mediaSimulating thermo‐hydro‐mechanical processes requires explicit inclusion of fractures whenever they add inhomogeneities at reservoir scale [ABSTRACT FROM AUTHOR]

Despite the variety of studies that have investigated the development of fracture networks during kerogen maturation in organic‐rich shale, the fracture interaction modes and generation mechanisms in three‐dimensions are not yet fully understood. In this study, we introduce a novel numerical approach to model the evolution of fracture swarms with three‐dimensional nonplanar geometries. This model enables precise simulation of the propagation, interplay, and coalescence of the fracture swarms with variable apertures and geometries via solving fluid flow, fracture growth, and stress interference. Our results suggest five basic fracture interaction modes between neighboring fractures. The evolving fracture swarms exhibit simultaneous, alternant, and differential growth characteristics at different development phases. We also elucidate the mechanical mechanisms that determine the evolution of three‐dimensional curved fracture swarms. This work yields an improved understanding of fluid‐driven fracture swarms' development in organic‐rich shale due to rapid fluid generation. Plain Language Summary: Rapid pressure increase due to fluid generation or fluid injection in pores of rock may promote the extension of multiple fractures, which is a common issue involved in natural geological processes and geo‐resources development. However, how the grown fractures interact with each other in three‐dimensions and why these fractures tend to develop into curved, interconnected geometries remain poorly understood. Here, we introduce a novel computer algorithm to accurately simulate the growing process of three‐dimensional fracture arrays with evolving fracture apertures and geometries. It is found that there are five basic fracture interaction modes that support the fracture arrays to form complex geometry and topology in three‐dimensions during fluid generation in shale rock. Through modeling, we also demonstrate how the stress in shale rock governs the growth of fracture arrays. This study results in a better understanding of fracture generation and oil/gas transport in unobservable subsurface rock. Key Points: Development of three‐dimensional nonplanar fracture swarms due to kerogen maturation in organic‐rich shale is numerically simulatedFive basic fracture interaction modes are proposed to understand the patterns of geomechanically grown fracture swarms in three‐dimensionsMechanical mechanisms that determine the simultaneous, alternant, and differential growth of curved fracture swarms are elucidated [ABSTRACT FROM AUTHOR]

Like natural seismicity, induced seismicity caused by injection also shows a power law size distribution, and its gradient b‐value is used for seismic hazard analysis. Despite the well‐known result that b‐value is negatively correlated with differential stress for natural earthquakes, there is no similar correlation for b‐value variations because the differential stress is nearly constant for injection‐induced seismicity, we thus investigate the b‐value dependence on the relative shear stress acting on existing fractures, using the fault orientation and in situ stress information. The seismicity occurring along existing fractures having high shear stress has significantly lower b‐values than does that associated with lower shear stress fractures. We examine the b‐value dependency on slightly changing differential stress, but the relationship is not clear. The b‐value for injection‐induced seismicity is dependent on relative shear stress on faults, which is a novel physical explanation for the b‐value variations of induced seismicity. Plain Language Summary: Frequency magnitude distribution of a series of natural earthquakes and laboratory earthquakes correlates with the applied stress. However, the variation of the frequency magnitude distribution for injection‐induced seismicity has not been understood well since the stress state is rather constant in reservoir scale. This study investigates the stress state of the faults that caused the injection‐induced seismicity. We found that the events that occurred from the fault that oriented to have relatively higher shear stress caused the b‐value reduction. This finding provides the new perspective of the scaling law of frequency magnitude distribution for induced seismicity. This insight leads to seismic hazard mitigation due to fluid injection. Key Points: We estimate shear stresses on the faults from in situ stress and fault orientation, and study the correlation to magnitude and b‐valueb‐value negatively correlates with the shear stress rather than differential stressShear stress is the cause of low b‐value for injection‐induced seismicity case [ABSTRACT FROM AUTHOR]

The offshore Castor Underground Gas Storage (UGS) project had to be halted after gas injection triggered three M4 earthquakes, each larger than any ever induced by UGS. The mechanisms that induced seismicity in the crystalline basement at 5–10 km depth after gas injection at 1.7 km depth remain unknown. Here, we propose a combination of mechanisms to explain the observed seismicity. First, the critically stressed Amposta fault, bounding the storage formation, crept by the superposition of well‐known overpressure effects and buoyancy of the relatively light injected gas. This aseismic slip brought an unmapped critically stressed fault in the hydraulically disconnected crystalline basement to failure. We attribute the delay between induced earthquakes to the pressure drop associated to expansion of areas where earthquakes slips cause further instabilities. Earthquakes occur only after these pressure drops have dissipated. Understanding triggering mechanisms is key to forecast induced seismicity and successfully design deep underground operations. Plain Language Summary: Underground fluid injection for energy‐related activities usually induces microseismicity (not felt earthquakes). But felt earthquakes are sometimes induced, which may cause public concern and project cancellation. This was the case of the offshore Castor underground gas storage (UGS), Spain. Gas injection induced numerous seismic events, including three with magnitudes around 4, larger than any other earthquake ever induced by UGS. These earthquakes occurred after the stop of injection in the crystalline basement, which is hydraulically disconnected from, and significantly deeper than the storage formation. To explain this seismicity, we propose a combination of mechanisms. First, the Amposta fault, bounding the site, crept (failed slowly, aseismically) because of pore pressure buildup induced by injection, which reduces the effective compression stabilizing faults, and stress variations caused by gas buoyancy. Aseismic slip of the Amposta fault provoked underground displacements, which reactivated a critically stressed unmapped fault in the crystalline basement. We attribute the delay between earthquakes of the sequence to rock expansion, and stabilizing pore pressure drop, in areas where slip displacement might cause new earthquakes. Earthquakes occur only after the pressure drop dissipates, which takes some time. Assessing fault stability prior to gas injection would have identified the risk of inducing seismicity. Key Points: Gas injection, through both pore pressure buildup and buoyancy, reactivated mainly aseismically the critically stressed Amposta faultShear slip stress transfer due to creep of the Amposta fault reactivated an unmapped critically stressed fault in the crystalline basementFelt earthquakes did not occur until pore pressure reductions had dissipated where shear slip stress transfer led to worsened stability [ABSTRACT FROM AUTHOR]

High‐pressure fluid injections cause transient pore pressure changes over large distances, which may induce seismicity. The zone of influence for such an injection was studied at high spatial resolutions in six decameter‐scaled fluid injection experiments in crystalline rock. Pore pressure time series revealed two distinct responses based on the lag time and magnitude of pressure change, namely, a near‐ and far‐field response. The near‐field response is due to pressure diffusion. In the far‐field, the fast response time and decay of pressure changes are produced by effective stress changes in the anisotropic stress field. Our experiments confirm that fracture fluid pressure perturbations around the injection point are not limited to the near field and can extend beyond the pressurized zone. Plain Language Summary: The far‐field pore pressure response in geological reservoirs due to high pressure fluid injection is not clarified yet. Direct observations of far‐field pore pressure changes were analyzed in the framework of the In‐Situ Stimulation and Circulation project executed at the Grimsel Test Site. The findings show two distinct behavior, one related to pore pressure diffusion in the near‐field of the injection and another one related to poro‐elastic effects. Key Points: Pore pressure time series reveal a near‐field response dominated by pressure diffusionThe far‐field is dominated by a quasi‐instantaneous poro‐elastic response due to the static anisotropic stress fieldInjection data showed pressure changes can extend 3–5 times farther than the pressurized volume because of poroelasticity [ABSTRACT FROM AUTHOR]

Earthquakes nucleate when large amounts of elastic energy, stored in the earth's crust, are suddenly released due to abrupt sliding over a fault. Fluid injections can reactivate existing seismogenic faults and induce/trigger earthquakes by increasing fluid pressure. Here we develop an analogous experimental system of simultaneously loaded and wetted absorbent porous paper to quantify theoretically the process of wetting‐induced earthquakes. This strategy allows us to gradually release the stored energy by provoking low intensity tremors. We identify the key parameters that control the outcome of the applied injection strategy, which include the initial stress state, fault segmentation, and segment‐activation rate. Subsequent injections, initiated at high stress levels, can drive the system faster toward its instability point, nucleating a large earthquake. Starting at low stress levels, however, they can reduce the magnitude of the natural event by at least one unit. Plain Language Summary: Understanding natural and anthropogenic seismicity is a major scientific challenge. Here we present a novel analogue fault model using absorbent porous paper, which gives new insights on earthquake mitigation. When scaled to in‐situ conditions, the porous paper model represents a natural seismic rupture of magnitude Mw = 5.9. By progressively wetting it, we simulate fluid injections in the earth's crust and draw analogies to large‐scale industrial projects. In our experiments, each injection is accompanied by tremors, which progressively release energy and modify the energy budget of the system. Without precise knowledge of the fault properties, we risk driving the system faster toward an unexpected large seismic event. However, provided that the model's key parameters–fault segmentation, segment‐activation rate, and stress state–are well known or controlled, the natural rupture can be mitigated by at least one unit. We expect that these results will facilitate risk reduction in current fluid injection projects and inspire earthquake mitigation strategies for real tectonic faults. Key Points: From an energetic point of view, absorbent porous paper can be an ideal, low‐cost surrogate rock material for studying induced seismicitySegmentation of faults and sequential fluid injection in each segment can mitigate potential earthquakes by at least one order of magnitudeWe show that fault segmentation, segment‐activation rate and stress state predominantly control the result of applied injection strategies [ABSTRACT FROM AUTHOR]

Understanding the relation between injection‐induced seismic moment release and operational parameters is crucial for early identification of possible seismic hazards associated with fluid‐injection projects. We conducted laboratory fluid‐injection experiments on permeable sandstone samples containing a critically stressed fault at different fluid pressurization rates. The observed fluid‐induced fault deformation is dominantly aseismic. Fluid‐induced stick‐slip and fault creep reveal that total seismic moment release of acoustic emission (AE) events is related to total injected volume, independent of respective fault slip behavior. Seismic moment release rate of AE scales with measured fault slip velocity. For injection‐induced fault slip in a homogeneous pressurized region, released moment shows a linear scaling with injected volume for stable slip (steady slip and fault creep), while we find a cubic relation for dynamic slip. Our results highlight that monitoring evolution of seismic moment release with injected volume in some cases may assist in discriminating between stable slip and unstable runaway ruptures. Plain Language Summary: Anthropogenic earthquakes caused by fluid injection have been reported worldwide to occur in the frame of waste‐water disposal, CO2 sequestration, and stimulation of hydrocarbon or deep geothermal reservoirs. To study the dynamics of injection‐induced seismic energy release in a controlled environment, we performed laboratory fluid injection experiments on critically stressed high‐permeability sandstone samples with a prefabricated fault. We monitored acoustic emission occurring during injection‐induced fault sliding. We find that the total seismic deformation (expressed as total seismic moment) is related to total injected volume, independent of fault slip modes (i.e., dynamic slip, steady slip, and fault creep). Seismic moment release rate roughly scales with fault slip velocity. In our experiments, the fluid pressure front migrates faster than the rupture front by about 5 orders of magnitude, resulting in fault slip within a zone of homogeneous fluid overpressure. We find that cumulative seismic moment scales linearly with the injected volume for stable slip (steady slip and fault creep), while it follows a cubic relation for dynamic slip. Our experimental results suggest that the deviation of cumulative moment release with injected volume from a linear trend in practice might be a sign for potential seismic risk. This may be considered in modifying current injection strategies. Key Points: Injection‐induced fault deformation is dominantly aseismicTotal moment release depends on total injected volume, independent of fault slip behaviorMoment‐injected volume scaling is linear for stable slip but shows a cubic relation for dynamic slip [ABSTRACT FROM AUTHOR]

This paper analyzes the effects of pore pressure rate for a spring‐block system that is a simple model of a laboratory experiment. Pore pressure is increased at a constant rate in a remote reservoir, and slip is governed by rate and state friction. The frequency of rapid slip events increases with the increase of a nondimensional pressure rate that is the ratio of the time scale of frictional sliding to that for pressure increase. As the pressure rate increases, the more rapid increase of pore pressure on the slip surface quickly stabilizes slip events due to rate and state friction. Rate and state and pressure rate effects interact in a limited range of pressure rate and diffusivity. This range includes pressure rates and diffusivities representative of recent laboratory experiments. Plain Language Summary: Recent field observations have identified fluid injection as an important factor in causing the dramatic increase of earthquakes in the central United States, and recent laboratory experiments have observed effects of fluid pressure rate on frictional sliding. This paper studies a simple model of a laboratory experiment: a block resting on a frictional surface and pulled by a spring. The frictional resistance to sliding depends on the rate and history of sliding. Fluid pressure is increased at a constant rate at a distance remote from the surface. The paper calculates the types and characteristics of rapid slip events and their dependence on the pressure rate and how fast fluid can diffuse from the reservoir to the frictional surface. Key Points: Rapid slip events occur during the duration of a representative experiment in a limited range of pressure rate and diffusivityAbove a certain pressure rate slip events are strongly damped by a rapid decrease of effective stressInteraction between fluid diffusion and pressure rate affects the type, frequency, and magnitude of slip events [ABSTRACT FROM AUTHOR]

Induced earthquakes from waste disposal operations in otherwise tectonically stable regions significantly increases seismic hazard. It remains unclear why injections induce large earthquakes on non‐optimally oriented faults kilometers below the injection horizon, particularly since fluids are not injected under pressure, but rather poured, into the well as observed in the Milan, Kansas area. Here we propose a mechanism for induced earthquakes whereby the karstic lower Arbuckle provides the short‐circuit that establishes a tens of MPa stepwise fluid pressure increase within the basement upon arrival of the hydraulic connection to the free surface and ultimately induce slip on the deeper fault. We investigate this scenario through modeling and mechanical analysis and show that earthquakes near Milan are likely induced by large (and sudden) fluid pressure changes when the karst network links two previously isolated hydrological systems. Plan Language Summary: We used numerical models to simulate the coupled hydrological and mechanical processes inducing seismicity in southern Kansas. We find that the presence of an extensive karst reservoir, the Arbuckle group, was a necessary condition to produce the M4.9 Milan earthquake, the largest earthquake to occur in over 100 years in Kansas. The results of these models also suggest that a significant percentage of the induced seismicity in Oklahoma would not have occurred without the presence of the Arbuckle. As demonstrated by this work, coupled hydromechanical models are critical to help understanding fluid behaviors in injection reservoirs and can be used to further understand the spatiotemporal distribution of induced earthquakes. Key Points: Earthquakes near Milan are likely induced by large and sudden fluid pressure changesHydrogeology of karst networks is the dominant factor controlling high pressure changesLarge pore pressure changes are required to induce earthquakes on nonoptimally oriented faults [ABSTRACT FROM AUTHOR]

Here we used Sentinel‐1 interferometric synthetic aperture radar (InSAR) data acquired between November 2014 to January 2019 to map how the basin's surface has deformed in response to fluid injection and extraction. While our stacking approach has low complexity, its accuracy increases with the Sentinel‐1 data volume. With an automated outlier removal algorithm, we achieved ∼2 mm/year accuracy across the basin in the presence of up to ±15 cm tropospheric noise. We observed numerous subsidence and uplift features near active production and disposal wells, with the maximum deformation rate occurring in 2018 when production peaked. The most important deformation signatures are linear patterns that extend tens of kilometers near Pecos, TX, where a cluster of increased seismic events was cataloged by the Texas Seismological Network (TexNet). Our elastic modeling results demonstrate that fluid extraction and dip slip along normal faults are potential causes for the observed seismicity and deformation patterns. Plain Language Summary: Over the past decade, breakthroughs in horizontal drilling and hydraulic fracturing have made the Permian Basin one of the most productive oil fields in the world. Using spaceborne interferometric synthetic aperture radar (InSAR), we mapped how the Permian Basin's land surface has deformed from oil and gas production activities. We developed a new processing technique to mitigate tropospheric noise associated with turbulent variations, which allows us measure ground changes with millimeter‐level accuracy. We observed numerous subsidence and uplift features near active production and disposal wells. The observed deformation rate is the highest in 2018 when the largest volume of oil and gas was produced in the basin. The InSAR‐observed subsidence patterns over the Pecos area can be modeled as dip slip over multiple normal faults and discretized cylindrical reservoir compaction. The implication for the scientific community, as well as a broader sector of stakeholders, is that the increase in high‐quality satellite‐based data now allows us to monitor vast areas for subsurface stress and pore pressure changes in oil‐producing regions. Key Points: We developed an InSAR processing strategy that achieves basin‐wide ∼2 mm/year accuracy in the presence of up to 15 cm of tropospheric noiseSentinel InSAR results show that the Permian Basin's sharp increase in shale oil production has led to numerous surface deformation featuresInSAR subsidence patterns near Pecos can be modeled as dip slip over multiple normal faults and discretized cylindrical reservoir compaction [ABSTRACT FROM AUTHOR]

Over 100 small‐ to moderate‐sized earthquakes, including an Mw 5.0 event, were detected during September 2015 to November 2016 near the town of Cushing, Oklahoma. The seismic sequence was spatial‐temporally linked to four wastewater disposal wells within 4 km. We calculate pore pressure and stress perturbations caused by fluid injection at multiple wells and analyze seismic risk in a Coulomb failure stress framework. Despite being more than an order of magnitude smaller than the pore pressure perturbation, the sign of shear stress change, in the sense of assumed right‐lateral fault motion, dictates where earthquakes are induced. Most of the relocated earthquakes are located within areas of positive shear stress changes. Our results suggest that poroelastic stress changes also play an essential role in the wastewater disposal environment, and a strategic design of well locations with respect to fault orientation and direction of motion can help mitigate induced seismic hazard. Plain Language Summary: Fluid injected into the subsurface is known to induce earthquakes. This study analyzes the relationship between the 2015/2016 Cushing earthquake sequence and wastewater disposal at four wells within 4 km of the sequence. Our results reveal that while pore pressure increase due to fluid diffusion makes the dominant contribution to promote fault slip toward instability, shear stress change in the sense of motion on a preexisting fault is a critical factor that determines where earthquakes can occur. This study suggests that a strategic design of disposal wells locations, with respect to mapped faults, may be an effective way to mitigate injection‐induced seismicity. Key Points: We calculate poroelastic stress changes to understand earthquake triggering due to injection at multidisposal wells near Cushing, OklahomaThe earthquakes occurred on a preexisting fault where shear stress change is small in amplitude but positive in right‐lateral strike‐slip motionShear stress change determined by relative well‐to‐fault location can also affect induced seismicity occurrence and hazard assessment [ABSTRACT FROM AUTHOR]

The manuscript analyzes the stimulation for an Enhanced Geothermal System development in Acoculco, Mexico. Using an analytical penny‐shaped hydraulic fracture model covering different propagation regimes, we computed the final fracture length and width by varying fluid properties with temperature. Our analysis indicates that for the given scenario, the fluid viscosity plays a minor role and instead flow rate and time of the stimulation are the controlling variables. We computed the fracture reopening as a consequence of water reinjection in the second stage of the stimulation through numerical computations based on the enriched discontinuity method. The computation shows that a single isolated fracture will not provide sufficient permeability, as the continuous injection will quickly fill and pressurize the crack. We demonstrate that the fracture needs to be connected to a permeable network to avoid excessive pressurization and achieve a commercially exploitable reservoir for Enhanced Geothermal System. Plain Language Summary: At Acoculco, Mexico, the temperature of the rock at 2 km depth is approximately 300°C, which can be potentially exploited for the production of geothermal energy. To produce deep geothermal energy, water must be extracted from the rock, which is currently not possible in Acoculco because the host rock, a granite, is practically impermeable. A possible solution is to inject water into the rock formation and create one or multiple large fractures: Through the fractures, the fluid will be able to circulate and geothermal energy can be extracted. In this work, we show, through numerical calculations, that it will be relatively easy to create a large fracture by injecting high‐pressure fluid, but if the fracture is not connected to a network of preexisting fractures, harvesting geothermal energy would be unlikely. The pressure will rapidly increase during the reinjection, which could trigger induced earthquakes if it exceeds sufficiently large values. We suggest instead to create fractures in multistage and maximize a chance of intersecting existing networks through zonal isolation. Key Points: Hydraulic stimulation can achieve fracture lengths of hundreds of meters in an EGS siteReinjection pressure quickly increases if the fracture is unconnected, making stimulation ineffectiveProlonged fluid injection does not improve permeability in unconnected fractures and zonal isolation should be employed instead [ABSTRACT FROM AUTHOR]

The Amundsen Sea low (ASL) is one of the key components of the Antarctic surface climate. Here, we assess the subseasonal to seasonal (S2S) predictive skill of the ASL in two state‐of‐the‐art forecasting systems based on the anomaly correlation coefficient (ACC). It is found that the ASL predictability during austral spring is higher than in the other seasons with lead times ranging from 1 to 4 weeks. The spring ASL predictability has an ACC of about 0.6 with lead times of 1–2 weeks and about 0.44 with lead times of 3–4 weeks. The enhanced ASL predictability during austral spring is mainly due to the strengthening of stratosphere‐troposphere coupling, coincident with a weakened stratospheric polar night jet and increased wave‐mean flow interactions in the stratosphere. This study demonstrates that the stratosphere‐troposphere coupling provides an important source of predictability for the Antarctic surface weather and climate on S2S timescale. Plain Language Summary: The Amundsen Sea low (ASL) is a persistent low‐pressure system located in the Southern Ocean between the Antarctic Peninsula and the Ross Sea. The presence of the ASL strongly modifies local weather systems (e.g., storms and rainfall), as well as Antarctic sea ice. As shipping activities are increasing over the Antarctic Ocean, the predictability of the ASL is becoming increasingly important. This study assesses the ASL predictability in two state‐of‐the‐art forecasting systems and identifies a significant enhancement of subseasonal forecast skill during austral spring. Further analysis reveals that the enhancement of the ASL predictability during austral spring is related to the timing of generally increased stratosphere‐troposphere coupling, suggesting that the atmospheric subseasonal predictability may be significantly improved in forecasting systems that capture the stratosphere‐troposphere coupling well. Key Points: The Amundsen Sea low (ASL) predictability in the forecasting systems has a seasonal dependency with high skill only during austral springThe skillful ASL predictions during austral spring are found to be coincident with the timing of increased stratosphere‐troposphere couplingThe atmospheric subseasonal predictability may be significantly improved in forecasting systems that capture stratosphere‐troposphere coupling well [ABSTRACT FROM AUTHOR]

In June 2019, an earthquake sequence comprising five M > 5 events occurred in a region of southwest China with fluid injection for both hydraulic fracturing and salt mining, which raised an extensive controversy on the cause. Here we use interferometric synthetic aperture radar (InSAR) observations to determine the source parameters of the sequence and to investigate the relationship with local injection activities. Both Sentinel‐1 and Advanced Land Observing Satellite 2 SAR images are collected to measure coseismic and preseismic surface deformation. Geodetic inversions with coseismic observations show that the sequence ruptured a previously unmapped southwest‐dipping thrust fault above 3 km depth, which intersects with the open‐hole sections of wells for solution mining of salt. In the 4 months before the sequence, cumulative line‐of‐sight displacements near the wells are around 1–2 cm after correcting seasonal‐like deformation. Our results indicate that water injection likely enhanced pore pressure within the fault zone and thus contributed to inducing the sequence. Plain Language Summary: Earthquakes are seldom reported in relation to fluid injection for salt mining. Here, we present a unique case of an earthquake sequence probably induced by such human activity in Changning (southwest China). The sequence began with a Ms 6.0 earthquake, which is an unprecedented magnitude and larger than the historical records of earthquakes induced by wastewater disposal and hydraulic fracturing. Due to the sparse distribution of seismic stations and the failure of a key station closest to injection wells for salt mining before the Ms 6.0 earthquake, available seismic data are not enough to constrain the spatial locations of events in the earthquake sequence. Here we instead use high spatial‐resolution interferometric synthetic aperture radar observations of coseismic surface deformation to image the causative fault of the sequence. We find that the open‐hole sections of injection wells intersect the seismogenic fault. Moreover, the cumulative surface displacements near the salt mine in the ~4 months before the sequence are 1–2 cm, which was mainly caused by ground uplift. Accordingly, we hypothesize that fluid injection increased pore pressure within the fault zone and brought it to rupture. Our interferometric synthetic aperture radar observations and inversion results present significant evidence toward the causality between salt mining and the earthquake sequence. Key Points: InSAR observations show significant ground deformation due to the Changning earthquake sequence near the salt mineInSAR inversions reveal that the Changning earthquake sequence ruptured a southwest‐dipping thrust fault above the depth of 3 kmThe open‐hole sections of wells injecting for solution mining of salt intersect the InSAR‐inferred seismogenic fault [ABSTRACT FROM AUTHOR]

Fault/fracture slip and seismicity caused by fluid injection are of major interest to subsurface science and engineering. However, some fundamental aspects of the temporal and spatial evolutions of induced seismicity remain unresolved. In this paper, we present the results of a laboratory injection‐induced shear test conducted on a rough granite fracture with concurrent acoustic emission (AE) monitoring. The results demonstrate a sequence of aseismic‐seismic‐aseismic fracture motion during fluid injection. It is observed that the temporal evolution of AE/microseismic activities was accompanied by the changes of slip velocity, stress drop, and friction coefficient. The seismic slip phase consists of three subphases, namely, a first quasi‐static slip, a dynamic slip, and a second quasi‐static slip. The dynamic slip occurred with an AE mainshock, simulating earthquake instability triggered by injection in deep crystalline rocks. In addition, slip heterogeneity controlled by the fracture surface roughness is directly evidenced by the spatially heterogeneous distribution of AE hypocenters and fracture surface topography. Plain Language Summary: Inducing fracture slip, dilation, and propagation to create a highly conductive fracture network by injection has been viewed as an effective reservoir stimulation approach for subsurface energy extraction. Fault/fracture slip can cause seismic/microseismic events, which can be used to estimate the stimulated volume. There are also some concerns regarding seismic hazard related to fluid injection. In order to better understand the underlying physics of fault slip and induced seismicity, we performed laboratory experiments. Here, we report results from an injection‐induced shear test conducted on a rough granite fracture, in which the fracture deformation, fluid flow, and acoustic emission were all concurrently measured. Our results demonstrate that faults/fractures may slip in different modes (from creep, slow quasi‐static slip, to fast dynamic slip) during fluid injection and each slip mode is reflected by a different AE/microseismic response and frictional behavior (slip weakening or slip strengthening). Moreover, it is observed that the spatially heterogeneous distribution of induced AE events is directly linked with the fracture surface topography and the shear‐induced asperity damage, indicating the controlling effect of surface topography on fault slip heterogeneity. Key Points: A laboratory injection‐induced shear test has been conducted on a rough granite fracture with concurrent acoustic emission monitoringAseismic‐seismic transition accompanied by temporal evolution of slip velocity, stress drop, and friction coefficient has been observedThe link between slip heterogeneity and fracture roughness is revealed by AE hypocenters distribution and fracture surface topography [ABSTRACT FROM AUTHOR]

Hydrofracturing is a routine industrial technique whose safety depends on fractures remaining confined within the target rock volume. Both observations and theoretical models show that, if the fluid volume is larger than a critical value, pockets of fluid can propagate large distances in the Earth's crust in a self‐sustained, uncontrolled manner. Existing models for such critical volumes are unsatisfactory; most are two‐dimensional and depend on poorly constrained parameters (typically the fracture length). Here we derive both analytically and numerically in three‐dimensional scale‐independent critical volumes as a function of only rock and fluid properties. We apply our model to gas, water, and magma injections in laboratory, industrial, and natural settings, showing that our critical volumes are consistent with observations and can be used as conservative estimates. We discuss competing mechanisms promoting fracture arrest, whose quantitative study could help to assess more comprehensively the safety of hydrofracturing operations. Plain Language Summary: Fractures in rocks can act as channels for fluids. Fracking, or hydrofracturing, involves injection of fluids at high pressure in order to grow fractures within the rock and increase its permeability. Fluid volumes need to be kept below a threshold value: If the fluid volume is larger, then the stresses at the tips of the fluid pocket will be large enough for the fluids to force their way around by fracturing the rock ahead of them. Previous theoretical models for the critical volumes are unsatisfactory as they are two‐dimensional and based on poorly constrained parameters. We derive and test a new three‐dimensional equation that uses only rock and fluid parameters. We find that typical volumes injected in hydrofracturing operations are over the limit we define. We argue they are still mostly safe as additional processes often hinder fracture growth. Further work is needed to comprehensively quantify mechanisms that hinder hydrofracture arrest. Key Points: We define critical volumes for fractures under the influence of a stress gradients that cause self‐sustaining propagationWe define such critical volumes analytically such that they are independent of scale. These are verified numericallyThe equation predicts the correct scale in natural and analog examples but for hydro‐fracturing its too low [ABSTRACT FROM AUTHOR]

An outstanding question for induced seismicity is whether the volume of injected fluid and/or the spatial extent of the resulting pore pressure and stress perturbations limit rupture size. We simulate ruptures with and without injection‐induced pore pressure perturbations, using 2‐D dynamic rupture simulations on rough faults. Ruptures are not necessarily limited by pressure perturbations when (1) background shear stress is above a critical value, or (2) pore pressure is high. Both conditions depend on fault roughness. Stress heterogeneity from fault roughness primarily determines where ruptures stop; pore pressure has a secondary effect. Ruptures may be limited by fluid volume or pressure perturbation extent when background stress and fault roughness are low, and the maximum pore pressure perturbation is less than 10% of the background effective normal stress. Future work should combine our methodology with simulation of the loading, injection, and nucleation phases to improve understanding of injection‐induced ruptures. Plain Language Summary: Earthquakes can be induced or triggered by fluid injected deep underground, if the fluid encounters faults. Previous studies of induced seismicity at different injection sites around the world have empirically found that in many cases the maximum magnitude earthquake is found to scale with total volume of injected fluid. However, this is not always the case, and the level and heterogeneity of preexisting stress on faults likely plays an important role in determining the final earthquake size. In this paper, we use numerical simulations of earthquakes to quantify one source of stress heterogeneity—that arising from geometric roughness—and study how changes in pore pressure and stress from fluid injection interact with preexisting stress to influence earthquake size. We find that earthquakes are not limited by the injected volume, except under specific conditions. Instead, earthquakes stop where preexisting conditions are unfavorable for continued rupture; in our case because of bends in the fault geometry. Earthquakes can well exceed the predicted maximum magnitude, depending on the preexisting stress on the fault, how rough it is, and the magnitude and extent of the perturbation from injection. Key Points: Rupture size is not necessarily limited by the volume of injected fluid in earthquake simulations with imposed pore pressure perturbationsStress heterogeneity arising from geometric roughness may be the primary cause for rupture termination on rough faultsHigh initial shear stress or pore pressure can trigger a rupture larger than the volume‐based magnitude limit on faults with low roughness [ABSTRACT FROM AUTHOR]

Understanding the physical mechanisms governing fluid‐induced fault slip is important for improved mitigation of seismic risks associated with large‐scale fluid injection. We conducted fluid‐induced fault slip experiments in the laboratory on critically stressed saw‐cut sandstone samples with high permeability using different fluid pressurization rates. Our experimental results demonstrate that fault slip behavior is governed by fluid pressurization rate rather than injection pressure. Slow stick‐slip episodes (peak slip velocity < 4 μm/s) are induced by fast fluid injection rate, whereas fault creep with slip velocity < 0.4 μm/s mainly occurs in response to slow fluid injection rate. Fluid‐induced fault slip may remain mechanically stable for loading stiffness larger than fault stiffness. Independent of fault slip mode, we observed dynamic frictional weakening of the artificial fault at elevated pore pressure. Our observations highlight that varying fluid injection rates may assist in reducing potential seismic hazards of field‐scale fluid injection projects. Plain Language Summary: Human‐induced earthquakes from field‐scale fluid injection projects including enhanced geothermal system and deep wastewater injection have been documented worldwide. Although it is clear that fluid pressure plays a crucial role in triggering fault slip, the physical mechanism behind induced seismicity still remains poorly understood. We performed laboratory tests, and here we present two fluid‐induced slip experiments conducted on permeable Bentheim sandstone samples crosscut by a fault that is critically stressed. Fault slip is then triggered by pumping the water from the bottom end of the sample at different fluid injection rates. Our results show that fault slip is controlled by fluid pressure increase rate rather than by the absolute magnitude of fluid pressure. In contrast to episodes of relatively rapid but stable sliding events caused by a fast fluid injection rate, fault creep is observed during slow fluid injection. Strong weakening of the dynamic friction coefficient of the experimental fault is observed at elevated pore pressure, independent of fault slip mode. These results may provide a better understanding of the complex behavior of fluid‐induced fault slip on the field scale. Key Points: Fluid‐induced fault slip experiments have been conducted on critically stressed sandstone samples with saw‐cut fracturesFluid‐induced fault slip mode is controlled by fluid pressurization rate rather than by the magnitude of fluid pressureIndependent of fault slip mode, we observed dynamic frictional weakening of the artificial fault at elevated pore pressure [ABSTRACT FROM AUTHOR]

Induced seismicity has expanded into south‐central Kansas, an area with rare damaging natural earthquakes, leading to the second‐highest seismicity rate in the central United States after Oklahoma. Here we assess the mechanical effects of large‐scale injection in the combined area of western Oklahoma and southern Kansas during 2010–2018 and its link to the observed seismicity using physics‐based hydromechanical and seismicity rate models. Such models link injection operations to seismic hazards and allow solving for the spatially variable distribution of background seismic productivity that yields an acceptable match between the observed and modeled seismicity. We show that injection in Oklahoma amplifies the total Coulomb stress change and seismicity rate by 1.5‐fold and threefold, respectively, in south‐central Kansas. This cross‐border interaction modulates the annual earthquake probability in Kansas. We conclude that the issue of induced seismicity is not a local problem due to the far‐reaching effects of fluid diffusion. Key Points: Unified hydromechanical model to reproduce and forecast induced seismicity for western Oklahoma and southern KansasThe state of stress and seismicity rate in Kansas are modulated by injection in OklahomaThe probability of large earthquakes in Kansas is exacerbated by the hydromechanical interaction with fluid injection in Oklahoma [ABSTRACT FROM AUTHOR]

Analysis of past and present stimulation projects reveals that the temporal evolution and growth of maximum observed moment magnitudes may be linked directly to the injected fluid volume and hydraulic energy. Overall evolution of seismic moment seems independent of the tectonic stress regime and is most likely governed by reservoir specific parameters, such as the preexisting structural inventory. Data suggest that magnitudes can grow either in a stable way, indicating the constant propagation of self‐arrested ruptures, or unbound, for which the maximum magnitude is only limited by the size of tectonic faults and fault connectivity. Transition between the two states may occur at any time during injection or not at all. Monitoring and traffic light systems used during stimulations need to account for the possibility of unstable rupture propagation from the very beginning of injection by observing the entire seismicity evolution in near‐real time and at high resolution for an immediate reaction in injection strategy. Plain Language Summary: Predicting and controlling the size of earthquakes caused by fluid injection is currently the major concern of many projects associated with geothermal energy production. Here, we analyze the magnitude and seismic moment evolution with injection parameters for prominent geothermal and scientific projects to date. Evolution of seismicity seems to be largely independent of the tectonic stress background and seemingly depends on reservoir specific characteristics. We find that the maximum observed magnitudes relate linearly to the injected volume or hydraulic energy. A linear relation suggests stable growth of induced ruptures, as predicted by current models, or rupture growth may no longer depend on the stimulated volume but on tectonics. A system may change between the two states during the course of fluid injection. Close‐by and high‐resolution monitoring of seismic and hydraulic parameters in near‐real time may help identify these fundamental changes in ample time to change injection strategy and manage maximum magnitudes. Key Points: Seismic moment during stimulation grows mostly linearly with injected fluid volumeClose‐by near‐real‐time monitoring of temporal seismic moment evolution may help identifying critical changes in the stimulated systemEvolution of seismicity seems largely independent of the tectonic regime and background stress [ABSTRACT FROM AUTHOR]

We monitored the seismohydromechanical rock mass response to high‐pressure fluid injection during a decameter‐scale hydraulic stimulation experiment in crystalline rock at the Grimsel Test Site, Switzerland. Time series recorded at two pressure monitoring locations show abrupt pressure increases that change in amplitude and appearance between subsequent stimulation cycles. Induced seismicity correlates with the propagation of one of these pressure fronts. Deformation data along the same shear zone shows permanent fracture dislocation preceded by strong transient fracture opening. We interpret these observations as nonlinear pressure diffusion along flow channels that reorganize in response to hydromechanical effects during stimulation. Combining these observations with the in situ stress field estimated before stimulation, we argue that the underlying hydromechanical processes involve mixed‐mode stimulation with both Mode I and II/III fracture dislocation. Plain Language Summary: Hydraulic shearing of preexisting fractures and hydraulic fracturing are commonly used to enhance the hydraulic conductivity and connectivity within in situ fracture networks, for example, in the framework of geothermal heat exploitation. Both treatments are based on high‐pressure fluid injections that induce deformation in the overall rock mass and dislocations along the fractures therein. It was previously observed that characteristic responses of both treatments can appear simultaneously during high‐pressure fluid injection. It is beneficial to locally resolve the different response behaviors in the rock mass, as they yield different connectivity enhancements. During a 20‐m‐scale stimulation experiment, we observed pressure signals within the target rock mass indicative of both shearing and fracturing behavior. These signals were monitored at ~7 m distance to injection location. The aseismic/seismic character of these pressure signals is in agreement with the different response behavior. Including previously conducted in situ stress characterization data, we argue that our experiment provides insight into the simultaneous occurrence of shearing and fracturing behavior. The results of this study are important for the interpretation of reservoir‐scale stimulation experiments and predictive numerical modeling of rock mass responses to high‐pressure fluid injections. Key Points: Behavior of directly monitored high‐pressure pulses indicates nonlinear diffusion induced by fracture dilationHigh fracture fluid pressure signals change direction during stimulation and are associated with seismic and aseismic deformationSeismohydromechanical monitoring indicates a combination of Mode I and II/III deformation during high‐pressure fluid injections [ABSTRACT FROM AUTHOR]

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. Plain Language Summary: In 2013, a deep geothermal project carried out in St. Gallen, Switzerland, aimed to generate heat and electricity for that city. But fluid pumped into the 4 km deep borehole caused hundreds of small earthquakes and one larger one that was distinctly felt by the local population. Although it is known that injecting fluid can change the conditions underground and trigger earthquakes, in St. Gallen the tremors occurred several hundred meters away from the well, necessitating a more detailed analysis. Consequently, we performed computer simulations designed to reveal otherwise hidden processes in the subsurface. The results indicate that the injected fluids rapidly flowed through permeable cracks toward the site where the earthquakes occurred, leading to significant changes in subsurface conditions. Knowledge of such highly permeable underground channels may help to improve the control over man‐made earthquakes in future industrial projects. Key Points: We model stress changes on a fault several hundred meters away from the St. Gallen injection wellA hydraulic connection yields stress changes 3 orders of magnitude greater than poroelastic effectsThe direct pressure effect is more likely to be the dominant mechanism inducing the seismicity [ABSTRACT FROM AUTHOR]

The increasing number of damaging earthquakes, induced by underground injection of fluids, demonstrates the need of a structured methodology for critical thinking about and clear communication of induced seismic risk. Using the Pohang, South Korea, geothermal project and the damaging (US$75M) local magnitude ML 5.35 induced earthquake in 2017, we present a straightforward methodology that summarizes induced seismic risk in a single time‐dependent number, the Value at Induced Risk (VaIR). VaIR quantifies the amount of economic losses that will not be exceeded at a given confidence level. Our method reveals that the Pohang project was exposed to a 2% chance of causing economic losses of US$140M or more (1% of US$3.2B or more). The high risk could have been identified during the sequence of reservoir stimulation, lasting almost 2 years. VaIR opens new ways to assess and communicate risk to all stakeholders for informed risk management decision making. Plain Language Summary: High‐pressure injection at a geothermal reservoir close to the city of Pohang, South Korea, triggered a damaging (US$75M) ML 5.35 earthquake in November 2017. We present a straightforward methodology for seismic risk assessment, communication, and management for underground injection of fluids. Our method, Value at Induced Risk (VaIR), combines preoperation loss scenario modeling and magnitude‐frequency statistics of the evolving induced seismicity catalog to monitor the evolution of induced seismic risk in real time. Retrospectively, we apply VaIR to the Pohang case. Operators, regulators, and nonexpert stakeholders could have been well‐informed about the elevating risk caused by underground injection in real time. Application of VaIR to future underground injection projects will help to exclude unacceptable losses with high confidence. Key Points: Value at Induced Risk (VaIR) is a straightforward method to communicate and manage seismic risk caused by underground injection of fluidsVaIR combines preoperation loss scenario modeling and magnitude statistics of the evolving induced seismicity catalog to monitor riskVaIR helps to exclude unacceptable losses with high confidence and to inform operators, regulators and nonexpert stakeholders at risk [ABSTRACT FROM AUTHOR]