Earthquakes caused by fluid injection into deep underground reservoirs constitute an increasingly recognized risk to populations and infrastructure. Quantitative assessment of induced seismic hazard, ...however, requires estimating the maximum possible magnitude earthquake that may be induced during fluid injection. Here I seek constraints on an upper limit for the largest possible earthquake using source‐physics simulations that consider rate‐and‐state friction and hydromechanical interaction along a straight homogeneous fault. Depending on the orientation of the pressurized fault in the ambient stress field, different rupture behaviors can occur: (1) uncontrolled rupture‐front propagation beyond the pressure front or (2) rupture‐front propagation arresting at the pressure front. In the first case, fault properties determine the earthquake magnitude, and the upper magnitude limit may be similar to natural earthquakes. In the second case, the maximum magnitude can be controlled by carefully designing and monitoring injection and thus restricting the pressurized fault area.
Key Points
Maximum possible magnitude during fluid injection depends on fault orientation
Along critically stressed faults the rupture‐front may propagate beyond the pressure front
Fault orientations and stress field can be included in induced seismic hazard analysis
Earthquake-induced landslides constitute a critical component of seismic hazard in mountainous regions. While many seismic slope stability analysis methods exist with varying degrees of complexity, ...details of interactions between seismic waves and incipient landslides are not well understood and rarely incorporated, in particular for deep-seated slope instabilities. We present a series of 2D distinct-element numerical models aimed at clarifying interactions between earthquakes and large rock slope instabilities. The study has two main goals: 1) to explore the role of amplification in enhancing co-seismic slope deformation — a relationship widely discussed in literature but rarely tested quantitatively; and 2) to compare our numerical results with the well-established Newmark-method, which is commonly used in seismic slope stability analysis. We focus on three amplification phenomena: 1) geometric (topographic) amplification, 2) amplification related to material contrasts, and 3) amplification related to compliant fractures. Slope height, topography, seismic velocity contrasts, and internal strength and damage history were varied systematically in a series of models with a relatively simple, scalable geometry. For each model, we compute the spatial amplification patterns and displacement induced by real earthquake ground motions. We find that material contrasts and internal fracturing create both the largest amplification factors and induced displacements, while the effect of geometry is comparably small. Newmark-type sliding block methods underestimate displacements by not accounting for material contrasts and internal fracturing within the landslide body — both common phenomena in deep-seated slope instabilities. Although larger amplification factors tend to be associated with greater displacements, we did not identify a clear link between ground motion frequency content, spectral amplification, and induced displacement. Nevertheless, observation of amplification patterns can play an important role in seismic slope stability analyses, as: 1) strong amplification (related to material contrasts or compliant fractures) is an indicator of potentially large co-seismic displacements; and 2) amplification patterns can be used to constrain geological and numerical models used for seismic stability analysis. The complexity of wave–slope interactions, as well as the potential to severely underestimate hazard using Newmark-type methods, motivates use of rigorous numerical modeling for quantitative seismic hazard and risk assessment of large landslides.
Display omitted
•Method to investigate seismic wave amplification+co-seismic displacement of landslides is presented.•Material contrasts and internal fracturing lead to strongest amplification and displacements.•Newmark's analysis underestimates displacements for large deep seated landslides.•Strong amplification indicative of high internal rock mass damage and history of slope deformation.
In this paper, we describe the investigations and actions taken to reduce risk and prevent casualties from a catastrophic 210,000 m
3
rockslope failure, which occurred near the village of Preonzo in ...the Swiss Alps on May 15, 2012. We describe the geological predisposition and displacement history before and during the accelerated creep stage as well as the development and operation of an efficient early warning system. The failure of May 15, 2012, occurred from a large and retrogressive instability in gneisses and amphibolites with a total volume of about 350,000 m
3
, which formed an alpine meadow 1250 m above the valley floor. About 140,000 m
3
of unstable rock mass remained in place and might collapse partially or completely in the future. The instability showed clearly visible signs of movements along a tension crack since 1989 and accelerated creep with significant hydromechanical forcing since about 2006. Because the active rockslide at Preonzo threatened a large industrial facility and important transport routes located directly at the toe of the slope, an early warning system was installed in 2010. The thresholds for prealarm, general public alarm, and evacuation were derived from crack meter and total station monitoring data covering a period of about 10 years, supplemented with information from past failure events with similar predisposition. These thresholds were successfully applied to evacuate the industrial facility and to close important roads a few days before the catastrophic slope failure of May 15, 2012. The rock slope failure occurred in two events, exposing a compound rupture plane dipping 42° and generating deposits in the midslope portion with a travel angle of 39°. Three hours after the second rockslide, the fresh deposits became reactivated in a devastating debris avalanche that reached the foot of the slope but did not destroy any infrastructure. The final run-out distance of this combined rock collapse–debris avalanche corresponded to the predictions made in the year 2004.
Fluid pressure within the Earth's crust is a key driver for triggering natural and human‐induced seismicity. Measuring fluid pressure evolution would be highly beneficial for understanding the ...underlying driving mechanisms and supporting seismic hazard assessment. Here we show that seismic velocities monitored on the 20‐m scale respond directly to changes in fluid pressure. Our data show that volumetric strain resulting from effective stress changes is sensed by seismic velocity, while shear dislocation is not. We are able to calibrate seismic velocity evolution against fluid pressure and strain with in situ measurements during a decameter‐scale fluid injection experiment in crystalline rock. Thus, our 4‐D seismic tomograms enable tracking of fluid pressure and strain evolution. Our findings demonstrate a strong potential toward monitoring transient fluid pressure variations and stress changes for well‐instrumented field sites and could be extended to monitoring hydraulic stimulations in deep reservoirs.
Plain Language Summary
The pressure of fluids in the subsurface is generally a function of depth as well as the regional geological history. Changes to the subsurface fluid pressure—be it natural or human induced—disturb the stress field and are known to drive volcanic eruptions, as well as to trigger earthquakes. For example, pressure increase by fluid injection for hydraulic stimulation and wastewater disposal has been linked to earthquake activity. Unfortunately, pressure measurements need direct access through boreholes, so that pressure data are only available for few locations. A method for estimating the spatial distribution of fluid pressure remotely would thus be highly beneficial. From measurements in a 20‐m‐scale experiment in granite, we find that fluid pressure propagation can be predicted from observed seismic velocity variations, based on a strong correlation between observed changes in seismic velocities and fluid pressure measured within the rock. As seismic velocities can be readily measured on the reservoir scale, our results demonstrate a strong potential of seismic velocity monitoring for remotely estimating fluid pressure changes in deep reservoirs, along faults, or in volcanic systems. The estimated pressure and stress changes could be an important input to real‐time risk analysis of fault reactivation and volcanic eruptions.
Key Points
Seismic velocities measured in 20‐m‐scale in situ experiment respond directly to high‐pressure fluid injections
In situ measurements of fluid pressure allow validation of seismic velocity measurements as proxy for field‐scale pressure monitoring
Three‐dimensional time‐lapse seismic velocity tomography allows monitoring of fluid pressure propagation through its relationship to effective stress
We explore the role of earthquake interactions during an injection‐induced seismic sequence. We propose a model, which considers both a transient pressure and static stress redistribution due to ...event interactions as triggering mechanisms. By calibrating the model against observations at the Enhanced Geothermal System of Basel, Switzerland, we are able to reproduce the time behavior of the seismicity rate. We observe that considering earthquake interactions in the modeling leads to a larger number of expected seismic events (24% more) if compared to a pressure‐induced seismicity only. The increase of the number of events is particularly evident after the end of the injection. We conclude that implementing a model for estimating the static stress changes due to mutual event interactions increases significantly the understanding of the process and the behavior of induced seismicity.
Key Points
We model synthetic catalogues for induced seismicity accounting for earthquake interactions in terms of static stress transfer
Static stress interactions affect the number of seismic events, which increase (24% more) with respect to a pressure‐induced only case
Static stress transfer has a significant role in understanding the rate and total number of events, as well as their spatial distribution
We performed a series of 12 hydraulic stimulation experiments in a 20m×20m×20m foliated, crystalline rock volume intersected by two distinct fault sets at the Grimsel Test Site, Switzerland. The goal ...of these experiments was to improve our understanding of stimulation processes associated with high-pressure fluid injection used for reservoir creation in enhanced or engineered geothermal systems. In the first six experiments, pre-existing fractures were stimulated to induce shear dilation and enhance permeability. Two types of shear zones were targeted for these hydroshearing experiments: (i) ductile ones with intense foliation and (ii) brittle–ductile ones associated with a fractured zone. The second series of six stimulations were performed in borehole intervals without natural fractures to initiate and propagate hydraulic fractures that connect the wellbore to the existing fracture network. The same injection protocol was used for all experiments within each stimulation series so that the differences observed will give insights into the effect of geology on the seismo-hydromechanical response rather than differences due to the injection protocols. Deformations and fluid pressure were monitored using a dense sensor network in boreholes surrounding the injection locations. Seismicity was recorded with sensitive in situ acoustic emission sensors both in boreholes and at the tunnel walls. We observed high variability in the seismic response in terms of seismogenic indices, b values, and spatial and temporal evolution during both hydroshearing and hydrofracturing experiments, which we attribute to local geological heterogeneities. Seismicity was most pronounced for injections into the highly conductive brittle–ductile shear zones, while the injectivity increase on these structures was only marginal. No significant differences between the seismic response of hydroshearing and hydrofracturing was identified, possibly because the hydrofractures interact with the same pre-existing fracture network that is reactivated during the hydroshearing experiments. Fault slip during the hydroshearing experiments was predominantly aseismic. The results of our hydraulic stimulations indicate that stimulation of short borehole intervals with limited fluid volumes (i.e., the concept of zonal insulation) may be an effective approach to limit induced seismic hazard if highly seismogenic structures can be avoided.
Deformation monitoring between 2004 and 2011 at the rock slope instability above Randa (Switzerland) has revealed an intriguing seasonal trend. Relative dislocation rates across active fractures ...increase when near‐surface rock temperatures drop in the fall and decrease after snowmelt as temperatures rise. This temporal pattern was observed with different monitoring systems at the ground surface and at depths up to 68 m, and represents the behavior of the entire instability. In this paper, the second of two companion pieces, we interpret this seasonal deformation trend as being controlled by thermomechanical (TM) effects driven by near‐surface temperature cycles. While Part 1 of this work demonstrated in a conceptual manner how TM effects can drive deep rock slope deformation and progressive failure, we present here in Part 2 a case study where temperature‐controlled deformation trends were observed in a natural setting. A 2D discrete‐element numerical model is employed, which allows failure along discontinuities and successfully reproduces the observed kinematics of the Randa instability. By implementing simplified ground surface temperature forcing, model results were able to reproduce the observed deformation pattern, and TM‐induced displacement rates and seasonal amplitudes in the model are of the same order of magnitude as measured values. Model results, however, exhibit spatial variation in displacement onset times while field measurements show more synchronous change. Additional heat transfer mechanisms, such as fracture ventilation, likely create deviations from the purely transient‐conductive temperature field modeled. We suggest that TM effects are especially important at Randa due to the absence of significant groundwater within the unstable rock mass.
Key Points
Displacement monitoring at Randa exhibit increased rates in winter
Temporal behavior at Randa is not related to ground water
Thermomechanical effects drive progressive failure at Randa
To characterize the stress field at the Grimsel Test Site (GTS) underground rock laboratory, a series of hydrofracturing and overcoring tests were performed. Hydrofracturing was accompanied by ...seismic monitoring using a network of highly sensitive piezosensors and accelerometers that were able to record small seismic events associated with metre-sized fractures. Due to potential discrepancies between the hydrofracture orientation and stress field estimates from overcoring, it was essential to obtain high-precision hypocentre locations that reliably illuminate fracture growth. Absolute locations were improved using a transverse isotropic P-wave velocity model and by applying joint hypocentre determination that allowed for the computation of station corrections. We further exploited the high degree of waveform similarity of events by applying cluster analysis and relative relocation. Resulting clouds of absolute and relative located seismicity showed a consistent east–west strike and 70° dip for all hydrofractures. The fracture growth direction from microseismicity is consistent with the principal stress orientations from the overcoring stress tests, provided that an anisotropic elastic model for the rock mass is used in the data inversions. The σ1 stress is significantly larger than the other two principal stresses and has a reasonably well-defined orientation that is subparallel to the fracture plane; σ2 and σ3 are almost equal in magnitude and thus lie on a circle defined by the standard errors of the solutions. The poles of the microseismicity planes also lie on this circle towards the north. Analysis of P-wave polarizations suggested double-couple focal mechanisms with both thrust and normal faulting mechanisms present, whereas strike-slip and thrust mechanisms would be expected from the overcoring-derived stress solution. The reasons for these discrepancies can be explained by pressure leak-off, but possibly may also involve stress field rotation around the propagating hydrofracture. Our study demonstrates that microseismicity monitoring along with high-resolution event locations provides valuable information for interpreting stress characterization measurements.
Cycles of glaciation impose mechanical stresses on underlying bedrock as glaciers advance, erode, and retreat. Fracture initiation and propagation constitute rock mass damage and act as preparatory ...factors for slope failures; however, the mechanics of paraglacial rock slope damage remain poorly characterized. Using conceptual numerical models closely based on the Aletsch Glacier region of Switzerland, we explore how in situ stress changes associated with fluctuating ice thickness can drive progressive rock mass failure preparing future slope instabilities. Our simulations reveal that glacial cycles as purely mechanical loading and unloading phenomena produce relatively limited new damage. However, ice fluctuations can increase the criticality of fractures in adjacent slopes, which may in turn increase the efficacy of fatigue processes. Bedrock erosion during glaciation promotes significant new damage during first deglaciation. An already weakened rock slope is more susceptible to damage from glacier loading and unloading and may fail completely. We find that damage kinematics are controlled by discontinuity geometry and the relative position of the glacier; ice advance and retreat both generate damage. We correlate model results with mapped landslides around the Great Aletsch Glacier. Our result that most damage occurs during first deglaciation agrees with the relative age of the majority of identified landslides. The kinematics and dimensions of a slope failure produced in our models are also in good agreement with characteristics of instabilities observed in the field. Our results extend simplified assumptions of glacial debuttressing, demonstrating in detail how cycles of ice loading, erosion, and unloading drive paraglacial rock slope damage.
Key Points
Simply adding then removing glacier ice from an alpine valley has little net effect on rock wall damage
Glacial erosion, i.e., rock debuttressing, creates significant new rock slope damage during first deglaciation
Damage kinematics vary during a glacial cycle: ice advance favors toppling, while retreat promotes sliding
Temporal changes in groundwater chemistry can reveal information about the evolution of flow path connectivity during crustal deformation. Here, we report transient helium and argon concentration ...anomalies monitored during a series of hydraulic reservoir stimulation experiments measured with an in situ gas equilibrium membrane inlet mass spectrometer. Geodetic and seismic analyses revealed that the applied stimulation treatments led to the formation of new fractures (hydraulic fracturing) and the reactivation of natural fractures (hydraulic shearing), both of which remobilized (He, Ar)-enriched fluids trapped in the rock mass. Our results demonstrate that integrating geochemical information with geodetic and seismic data provides critical insights to understanding dynamic changes in fracture network connectivity during reservoir stimulation. The results of this study also shed light on the linkages between fluid migration, rock deformation and seismicity at the decameter scale.
PDF
