Numerous cases of induced/triggered seismicity have been reported in the last decades, directly or indirectly related to anthropogenic activity for the geo-resources exploration. Induced earthquakes felt by local population can often negatively affect public perception of geo-energies and may lead to the cancellation of important projects. Furthermore, large earthquakes may jeopardize wellbore stability and damage surface infrastructure. Thus, monitoring and modeling processes leading to fault reactivation, (seismic or aseismic) are critical to develop effective and reliable forecasting methodologies during deep underground exploitation. The complex interaction between injected fluids, subsurface geology, stress interactions, and resulting induced seismicity requires an interdisciplinary approach that accounts for coupled thermo-hydro-mechanical-chemical processes to understand the triggering mechanisms.
In this session, we invite contributions from research aimed at investigating the interaction of the above processes during exploitation of underground resources, including hydrocarbon extraction, wastewater disposal, geothermal-energy exploitation, hydraulic fracturing, gas storage and production, mining, and reservoir impoundment for hydro-energy. We particularly encourage novel contributions based on laboratory and underground near-fault experiments, numerical modeling, the spatio-temporal relationship between seismic properties, injection/extraction parameters, and/or geology, and fieldwork. Contributions covering both theoretical and experimental aspects of induced and triggered seismicity at multiple spatial and temporal scales are welcome.

Co-organized by EMRP1/SM6
Convener: Antonio Pio Rinaldi | Co-conveners: Léna CauchieECSECS, Rebecca M. Harrington, Marco Maria ScuderiECSECS, Victor Vilarrasa
| Attendance Thu, 07 May, 14:00–18:00 (CEST)

Files for download

Session materials Download all presentations (198MB)

Chat time: Thursday, 7 May 2020, 14:00–15:45

Chairperson: Antonio Pio Rinaldi, Lena Cauchie, Victor Vilarrasa
D917 |
Maria Bobrova, Egor Filev, Anna Shevtsova, Sergey Stanchits, Vladimir Stukachev, and Brice Lecampion

Understanding the processes of Hydraulic Fracturing (HF) initiation and propagation in different types of rocks is important for the design and optimization of HF during the exploitation of underground resources. The main goals were to study the dynamics of the process of hydraulic fracture growth and possible optimization of HF technology for both homogeneous and heterogeneous rocks. Laboratory experiments on HF with different injection parameters were carried out on natural limestone, dolomite and shale specimens. The dynamics of HF process was monitored by Acoustic Emission (AE) technique, on the analogy of induced microseismicity monitoring of HF in the field conditions. The shape of created HF and the size of leak-off zone were analyzed by X-Ray CT scanning technique after the testing.

Experiments on dolomite were conducted using fluids with different viscosities (1000-10000 cP) injected into the rock with a rate of 0.5 ml/min. In case of low viscosity, we observed low AE activity. After the test, the sample was cut in several pieces transverse to the expected fracture plane. We have found that HF has initiated, but did not reach the sample boundaries and leak-off was significant. The ten times increase of fluid viscosity resulted in significantly increased AE activity, smaller size of leak-off zone and higher breakdown pressure (21.8 against 18.7 MPa). The post-test 3D shape of HF surface obtained by X-Ray CT closely correlates with 3D shape of localized AE events, confirming that the fracture propagated in the direction of maximal stress, as expected. It means that viscosity of fracturing fluid had a significant effect on fracturing breakdown pressure and fracture behavior.

The influence of different rock types on hydraulic fracturing was studied with dolomite, limestone and shale samples. In case of dolomite and shale, sufficient number of Acoustic Emission events were recorded, which allowed tracing the direction and dynamics of fracture propagation. However, for the limestone, a very small number of AE events were localized with the same parameters of injected fluid. Comparison of dolomite and shale HFs shows that the crack in the shale had a more complex shape, deviating from the maximal stress direction, which was explained by rock heterogeneity, by the presence of natural cracks and inclined planes of weakness. It led us to conclusion that the rock fabric plays an important role in the behavior of hydraulic fracture in heterogeneous rock.

How to cite: Bobrova, M., Filev, E., Shevtsova, A., Stanchits, S., Stukachev, V., and Lecampion, B.: Influence of injection parameters on the dynamics of Hydraulic Fracturing monitored by Acoustic Emission, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-9423, https://doi.org/10.5194/egusphere-egu2020-9423, 2020

D918 |
Antoine L. Turquet, Renaud Toussaint, Fredrik K. Eriksen, Eirik Grude Flekkøy, and Knut Jørgen Måløy

An earthquake can happen due to many different phenomena such as sliding faults, fluid/gas injection into the subsurface or volcanic activities. Understanding the cause of earthquakes is one important step towards a better hazard assessment and better mitigation. In this study, we explore the physics behind different types of earthquakes by inducing similar mechanics in lab-scale experiments using an analogous model. Inside a transparent rectangular Hele-Shaw cell, we induce lab-scale microseismicity via pneumatic fracturing. An 80 x 40 cm transparent setup is prepared using a 1 mm thin layer of uncompacted granular medium having a fixed grain size is placed between two glass plates.
The seismic location results are compared with the image correlation results for displacement maps corresponding to the event times. Using air injection, this porous medium is compacted and fractured. This system is monitored using a camera recording 1000 images per second and accelerometers recording with 1 MHz sampling rate. Sources of earthquake-like vibrations are both located using acoustic recordings and image processing. We have observed that the deformation starts with compaction inside the medium; this compaction propagates toward the channel tips and causes the fingers to advance further inside the medium. We have observed (using optics and acoustics) that the movement starts inside the porous medium and progresses toward the channel tips, eventually causing channels to grow further. We also compared the characteristic patterns in these lab-scale events that are very similar to large scale correspondents, in particular with 2017 Mw 5.5 Pohang Earthquake. We reverse-engineered the signature of the recorded lab-scale signals to have a better understanding of this industrial hazard.

How to cite: Turquet, A. L., Toussaint, R., Eriksen, F. K., Flekkøy, E. G., and Måløy, K. J.: Reverse Engineering Earthquakes using Lab-scale Replicas: Application to Mw5.5 2017 Pohang Earthquake, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-7311, https://doi.org/10.5194/egusphere-egu2020-7311, 2020

D919 |
Yves Guglielmi, Jens Birkholzer, Jonathan Ajo-Franklin, Christophe Nussbaum, Frederic Cappa, PierPaolo Marchesini, Michelle Robertson, Martin Schoenball, Chett Hopp, Paul Cook, and Florian Soom

Understanding fault reactivation as a result of subsurface fluid injection in shales is critical in geologic CO2 sequestration and in assessing the performance of radioactive waste repositories in shale formations. Since 2015, two semi-controlled fault activation projects, called FS and FS-B, have been conducted in a fault zone intersecting a claystone formation at 300 m depth in the Mont Terri Underground Research Laboratory (Switzerland). In 2015, the FS project involved injection into 5 borehole intervals set at different locations within the fault zone. Detailed pressure and strain monitoring showed that injected fluids can only penetrate the fault when it is at or above the Coulomb failure criterion, highlighting complex mixed opening and slipping activation modes. Rupture modes were strongly driven by the structural complexity of the thick fault. An overall normal fault activation was observed. One key parameter affecting the reactivation behavior is the way the fault’s initial very low permeability dynamically increases at rupture. Such complexity may also explain a complex interplay between aseismic and seismic activation periods. Intact rock pore pressure variations were observed in a large volume around the rupture patch, producing pore pressure drops of ~4 10-4 MPa up to 20 m away from the ruptured fault patch. Fully coupled three-dimensional numerical analyses indicated that the observed pressure signals are in good accordance with a poro-elastic stress transfer triggered by the fault dislocation.


In 2019, the FS-B experiment started in the same fault, this time activating a larger fault zone volume of about 100 m extent near (and partially including) the initial FS testbed. In addition to the monitoring methods employed in the earlier experiment, FS-B features time-lapse geophysical imaging of long-term fluid flow and rupture processes. Five inclined holes were drilled parallel to the Main Fault dip at a distance of about 2-to-5m from the fault core “boundary”, with three boreholes drilled in the hanging wall and two boreholes drilled in the foot wall. An active seismic source-receiver array deployed in these five inclined boreholes allows tracking the variations of p- and s-wave velocities during fault leakage associated with rupture, post-rupture and eventually self-sealing behavior. The geophysical measurements are complemented by local three-dimensional displacements and pore pressures measurements distributed in three vertical boreholes drilled across the fault zone. DSS, DTS and DAS optical fibers cemented behind casing allow for the distributed strain monitoring in all the boreholes. Twelve acoustic emission sensors are cemented in two boreholes set across the fault zone and close to the injection borehole. Preliminary results from the new FS-B fault activation experiment will be discussed.

How to cite: Guglielmi, Y., Birkholzer, J., Ajo-Franklin, J., Nussbaum, C., Cappa, F., Marchesini, P., Robertson, M., Schoenball, M., Hopp, C., Cook, P., and Soom, F.: Imaging Poro-Elastic Effects induced by a Normal Fault Aseismic-to-Seismic Dislocation in Shales, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-6195, https://doi.org/10.5194/egusphere-egu2020-6195, 2020

D920 |
Carolin Boese, Grzegorz Kwiatek, Georg Dresen, Joerg Renner, Thomas Fischer, and Katrin Plenkers and the STIMTEC Team

Between early 2018 and late 2019 the STIMTEC hydraulic stimulation experiment was performed at ca.~130 m below surface at the Reiche Zeche research mine in Freiberg, Saxony/Germany. The project aims at gaining insight into the creation and growth of fractures in anisotropic and heterogeneous crystalline rock units, to develop and optimise hydraulic stimulation techniques for EGS applications and to control the associated induced seismicity under in situ conditions. A series of ten hydro-frac experiments were performed in a 63 m-long, 15°-inclined injection borehole and five mini-fracs for stress measurements in a sub-vertical borehole. These were monitored using a seismic monitoring system of twelve high-sensitivity Acoustic emission (AE) sensors, three accelerometers and one broadband sensor. More than 11,000 high-frequency AE events with source sizes on the cm-to-dm scale accompanied the hydraulic stimulation in five of ten stimulated intervals in the injection borehole. Several hundred AE events were recorded during the mini-fracs in the vertical borehole. We investigate the characteristics of induced AE events by combining information obtained from high-accuracy event locations using a transversely isotropic P-wave velocity model per station with station corrections, relative hypocentre locations, and focal mechanism solutions of selected events. The AE event clouds extend ca. 5 m radially from the injection points and show variying orientations and dips. The ca. 150 focal mechanism solutions obtained using P-wave polarisations display mixed-mode failure with a significant portion of them showing compaction. The orientation of the maximum principal stress inferred from the hydro-fracs in the injection and vertical boreholes has a trend of N348°E and a plunge of 20°, as typical for southeast Germany. However, discrepancies in the magnitudes of the principal stresses were measured between these boreholes ca. 15 m apart, resulting in different faulting regimes. We present stress orientations obtained from inverting focal mechanism solutions to provide additional information for interpreting stress-characterisation measurements.

How to cite: Boese, C., Kwiatek, G., Dresen, G., Renner, J., Fischer, T., and Plenkers, K. and the STIMTEC Team: Characterising induced acoustic emission activity observed during a mine-scale hydraulic-fracturing experiment in anisotropic crystalline rock, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-9681, https://doi.org/10.5194/egusphere-egu2020-9681, 2020

D921 |
Rebecca O. Salvage and David W. Eaton

On 30 November 2018, three felt earthquakes occurred in quick succession close to the city of Fort St. John, British Columbia, likely as a direct response to a hydraulic fracturing operation in the area. Events appear tightly clustered spatially within the upper 10 km of the crust. Hypocenters locate at the confluence between a large scale reverse faulting regime (in the north-west, probably due to the influence of the Rocky Mountain fold and thrust belt) and an oblique strike slip faulting regime (in the south-east, probably due to the influence of the Fort St. John Graben), resulting in a variety of focal mechanisms and a very complex local stress regime. Further analysis of the principal stresses suggests that σ1 is well constrained and close to horizontal, whereas σ2 and σ3 are poorly constrained, and can alternate between the horizontal and the vertical plane. Here, we present an overview of the temporal and spatial evolution of this seismic sequence and its relationship to hydraulic fracturing operations in the area, and examine the influence of large-scale regional tectonic structures on the generation of seismicity on this occasion.


How to cite: Salvage, R. O. and Eaton, D. W.: Hydraulic fracturing operations within an extremely complex stress regime: The case of Fort St. John, British Columbia, Canada, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-5702, https://doi.org/10.5194/egusphere-egu2020-5702, 2020

D922 |
| solicited
| Highlight
Xiaowei Chen, Guang Zhai, Manoochehr Shirzaei, and Yan Qin

In the last decade, Oklahoma has experienced significant changes in earthquake activities: earthquake rate dramatically increased since 2009, with a peak rate exceeding California, which has gradually decreased in recent years. This “accidental” large scale earthquake experiment provides us with rich datasets to further understand earthquake physics. Here, focusing on analyses of seismicity and accounting for the physics of earthquake nucleation, we link several studies to give a brief overview of Oklahoma earthquakes, and their implications for future studies in induced seismicity.  First, the analysis of spatiotemporal patterns of seismicity rate can help us infer the subsurface hydraulic parameters at both regional and local scales. At the regional level, the hydraulic diffusivities differ between Eastern and Western Oklahoma, separated by the Nemaha Fault, reflecting hydraulic properties of the Arbuckle Group (injection layer). At local scales within individual faults, the analysis suggested similar hydraulic diffusivities to crustal earthquake bursts from other tectonic regions, implying common properties of the crystalline basement.  Second, coupled poroelastic responses to injection on individual faults are essential, and produce seismicity rate forecast that more closely resemble observations. However, local stress tensor variations can significantly influence fault “criticality” and should be taken into account for modeling stress interactions. Third, in addition to injection-related stress changes, earthquake interactions and aseismic slip need to be considered in induced earthquake sequences, and detailed source modeling and statistical analysis are required to understand their roles in the evolutions of individual sequences further. Finally, Oklahoma seismicity offers opportunities to test short- to intermediate-term forecasting based on different physical models, and new windows into earthquake rupture initiations.

How to cite: Chen, X., Zhai, G., Shirzaei, M., and Qin, Y.: A decade of Oklahoma earthquakes: past, present and future , EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-10955, https://doi.org/10.5194/egusphere-egu2020-10955, 2020

D923 |
Man Zhang, Shemin Ge, Qiang Yang, and Xiaodong Ma

Xiluodu is currently the third largest hydropower station in the world and situates on the upper Yangtze River in Southwestern China. The 285.5 m-high dam lies in the center of a relatively intact and stable tectonic block, triangulated by three large fault zones. The seismicity in the region increased markedly since the reservoir was first impounded in 2013. Previous studies suggest a strong spatial-temporal link between the seismicity and reservoir impoundment. This study attempts at conducting a quantitative analysis integrating the geological and engineering data to constrain the link between the impoundment and the seismicity, which could inform the future seismic evolution in the area.

We first study the characters of the spatial activity of earthquakes in different periods to address the correlation between increased seismicity and reservoir impoundment. Since the impoundment, the earthquakes in this region can be plausibly separated spatially into two groups. The first group (including a ML5.4 and a ML5.5 event) is located within ~10km of reservoir, where a major fault zone is absent. Within this spatial range,  earthquakes > ML 2.0 are rare three years prior to the impoundment, but more than 1000 events were detected between the initial impoundment in 2013  and September 2014 when the reservoir reached its peak level. Thereafter, the fluctuations of water level were accompanied by continuous seismicity, albeit at a considerably lower rate. The seismicity in this region is strengthened again in 2019. The other group of earthquakes are clustered with several mapped major fault traces. Some of these events quickly followed the water level fluctuation, while some were observed after significant delays. In general, the distances between locations of delayed events and the reservoir gradually increase with time. 

To address the influence of impoundment on seismicity, we analyzed the hydrologic and mechanical effects of the impoundment, i.e., the fluid pressure diffusion and the reservoir loading. We computed the spatiotemporal changes of Coulomb stress on known faults resulting from these two effects. The sensitivity analysis of hydraulic and mechanical parameters shows that the changes of Coulomb stress in the area could increase to a level that is relevant to reactivation of faults. While the relationship between the impoundment and increase seismicity warrants further analysis, we hope to inform the regional seismic impact by integrating in-situ stress state, fault geometries, and the coupled hydro-mechanical stress changes.

How to cite: Zhang, M., Ge, S., Yang, Q., and Ma, X.: Reservoir impoundment, hydro-mechanical changes, and increased seismicity near the Xiluodu hydropower dam, Southwestern China, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-7618, https://doi.org/10.5194/egusphere-egu2020-7618, 2020

D924 |
Vala Hjörleifsdóttir, Gunnar Gunnarsson, Sigríður Kristjánsdóttir, Bergur Sigfússon, Halldór Geirsson, Kristín Jónsdóttir, Ingvi Gunnarsson, and Kristján Ágústsson

The 303 MW Hellisheiði, Iceland geothermal power plant was commissioned in 2006 and in early September 2011, reinjection of geothermal fluid was initiated in the second reinjection site of the plant; Húsmúli.  The site has 5 injection wells in operation, with depths of over 2000 m and a total of up to 500 l/s of fluid being reinjected into the site.  Seismicity had previously been observed in the region, including both natural seismicity before power plant operations started (e.g. Foulger et al., 1988) and induced seismicity during drilling of the injection wells (Ágústsson et al., 2015).  The reinjection caused severely increased level of seismicity within days, with two earthquakes of M 4.0 and M3.9 respectively, occurring a little over a month after the start of reinjection (Icelandic Meteorological Office catalog). The injection was also accompanied by uplift of approximately 2 cm (Juncu et al., 2018).  Due to the increased level of seismicity, a committee was formed and several measures on how to control it were suggested – including starting reinjection gradually after it has been stopped (Bessason et al., 2012).

In 2014, as a part of the Carbfix2 project, the reinjection fluid in Húsmúli was combined with gas, and CO2 and H2S, previously being released into the atmosphere, is now captured and reinjected into the basaltic formation (Matter et al., 2016, Gunnarsson et al., 2018).  It is estimated that the CO2 and H2S are crystalized into calcite and pyrite in under 2 years (Gunnarsson et al., 2018). This project has been very successful and is currently capturing and permanently storing an estimated 33% of the CO2 and 75% of the H2S extracted.

In this study we analyze seismicity data as reported by the Icelandic Meteorological Office Regional network, (1991-present) and the ON Power/ISOR local network (2016-present) and compare with operational parameters.  We show 1) how the seismicity responds to changes in flow, pressure and temperature of the injected fluid, 2) how individual wells seem to respond differently, 3) how the mitigation measures taken by the operator have worked and 4) look for changes in seismicity due to the CO2 sequestration.

This work has been funded by the European Union’s Horizon 2020 research and innovation Program projects Carbfix2 (grant agreement number 764760) and S4CE (grant agreement number 764810).

How to cite: Hjörleifsdóttir, V., Gunnarsson, G., Kristjánsdóttir, S., Sigfússon, B., Geirsson, H., Jónsdóttir, K., Gunnarsson, I., and Ágústsson, K.: Induced earthquakes at the carbon sequestration site Carbfix2, Hellisheiði, Iceland, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-19749, https://doi.org/10.5194/egusphere-egu2020-19749, 2020

D925 |
Rike Köpke, Olivier Lengliné, Jean Schmittbuhl, Emmanuel Gaucher, and Thomas Kohl

In a geothermal reservoir, seismicity may be induced due to changes in the subsurface as a result of drilling, stimulation or circulation operations. The induced seismic events are therefore strongly linked to the fluid flow, the mechanical state of the reservoir and the geological structures that impact the stress field and make this fluid flow possible. Here, the study is based on the monitoring of the development and operation of the deep geothermal site at Rittershoffen (Alsace, France) using different seismic networks covering various operational periods from September 2012 to present, including the drilling of the well doublet GRT1/GRT2, stimulation of GRT1 and well testing. The seismicity induced by these operations has the potential to give valuable insight into the geomechanical behaviour of the reservoir and the geometry of the fracture network. The present study gives an overview of the spatial and temporal development of the induced seismicity and the magnitudes of the events to provide insights into active structures in the reservoir.


To improve the level of detection, we first apply a template matching algorithm to the continuous waveforms recorded by the seismic networks. After running the detection with the template matching, the relative locations of all detected events are calculated as well as relative magnitudes. This workflow is applied to the whole time period from the start of the drilling in 2012 up to 2017. The spatial and temporal evolution of the events and their magnitudes shows how the different operations during reservoir development influence the seismogenic development of the reservoir and the seismic activity during continuous operation of the site. Further analysis like b-value computation, estimation of the best-fitting planes to the seismic clouds and evaluation of the waveform correlation between the seismic events give insight into the processes that induced the seismicity and the relation between different seismic intervals.


Focus of the present study is on the similarities and differences in the seismic response of the reservoir to the three subsequent stimulations of GRT1, called thermal, chemical and hydraulic stimulation. Results show that the seismicity induced during the hydraulic stimulation is much stronger in terms of seismicity rate and magnitudes than seismicity induced during thermal stimulation and migrates further into the reservoir. Noticeably, after a seismically quiet period of four days after the hydraulic stimulation a short burst of seismicity occurred unrelated to any operations on site. Seismicity during this delayed interval proved to have quite distinct characteristics from the seismicity induced during injection. While no significant seismicity was induced during chemical stimulation, the operation may have had an important influence on the seismic response of the reservoir during hydraulic stimulation by changing the state of the present fracture network.

How to cite: Köpke, R., Lengliné, O., Schmittbuhl, J., Gaucher, E., and Kohl, T.: Analysis of the seismic response to reservoir development and long-term operation at the Rittershoffen deep geothermal site, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-20317, https://doi.org/10.5194/egusphere-egu2020-20317, 2020

D926 |
Claudia Finger and Erik H. Saenger

Locating and characterizing the seismicity in a reservoir is crucial for any geothermal project. This study is the first time that the seismicity in a geothermal reservoir is characterized using time-reverse imaging (TRI). The spatio-temporal distribution of events in combination with focal mechanism solutions may enable the mapping of existing fault networks, the estimation of local stress regimes and the distinction between tectonic and induced events. Combining these results with results from other methodologies will in the future lead to an informed understanding of the physical processes occurring in reservoirs.

TRI is a method for locating and characterizing seismic events. TRI uses the whole time-reversed waveform and a seismic wave propagation solver to locate and characterize events. Therefore, it does not rely on the identification of seismic events and their onsets in the traces. In contrast to common tools that provide hypocenters and focal mechanism solutions for seismic events, TRI does not assume any a priori knowledge about the sources. Since events are not picked in the seismic traces, no assumption is made about the number of sources recorded in a certain time window. Similarly, the characterization of events does not exclude any source type or put any constraints or assumptions on the sources, such as them being only of double-couple nature. Therefore, TRI may be especially well-suited when the overall type of sources is not known or if it is suspected that common localization and characterization tools are not adequately depicting the physical processes in the subsurface.

In the first part of this study, seismic events, that occurred in the geothermal field of Los Humeros in Mexico, are located using TRI. So-called sensitivity maps are used to enhance the localization capabilities and to determine the spatial variation in source-location accuracy. In the second part of this study, the located events are characterized by determining the full time-dependent moment tensor. Since no assumption about the source type is made, these moment tensors complement results obtained from more standardized tools.

How to cite: Finger, C. and Saenger, E. H.: Locating and characterizing seismic events in Los Humeros (Mexico) using time-reverse imaging, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-13116, https://doi.org/10.5194/egusphere-egu2020-13116, 2020

D927 |
Georg Dresen, Stephan Bentz, Grzegorz Kwiatek, Patricia Martinez-Garzon, and Marco Bohnhoff

Recent results from an EGS project in Finland suggest a possibly successful physics-based approach in controlling stimulation-induced seismicity in geothermal projects. We analyzed the temporal evolution of seismicity and the growth of maximum observed moment magnitudes for a range of past and present stimulation projects. Our results show that the majority of the stimulation campaigns investigated  reveal a clear linear relation between  injected fluid volume, hydraulic energy and cumulative seismic moments. For most projects studied, the observations are in good agreement with existing physical models that predict a relation between injected fluid volume and maximum seismic moment of induced events. This suggest that seismicity results from a stable, pressure-controlled rupture process at least for an extended injection period. Overall evolution of seismicity is independent of tectonic stress regime and is most likely governed by reservoir specific parameters, such as the preexisting structural inventory. In contrast, there are few stimulations that reveal unbound increase in seismic moment suggesting that for these cases evolution of seismicity is mainly controlled by stress field, 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.

How to cite: Dresen, G., Bentz, S., Kwiatek, G., Martinez-Garzon, P., and Bohnhoff, M.: Seismic moment evolution during hydraulic stimulations in EGS projects, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-3392, https://doi.org/10.5194/egusphere-egu2020-3392, 2020

D928 |
| Highlight
Jean Schmittbuhl, Olivier Lengliné, Sophie Lambotte, Marc Grunberg, Cécile Doubre, Jérôme Vergne, François Cornet, and Frédéric Masson

On Nov 12, 2019, a Ml3.1 earthquake was felt by the whole population of the city of Strasbourg, France. It was located by the BCSF-RéNaSS (EOST) in the northwestern part of the town (Robertsau area) at a depth of 5.5km. Its location in the vicinity of the deep geothermal wells (GEOVEN), the temporal correlation with the injection activity on site, the similarity of the depth between the bottom of the wells and  the hypocenter of the event, the lack of local seismicity before the event occurrence, the known geological structures including crustal faults in the area, immediately questioned the possible triggering of the event by the deep geothermal activities despite the relatively large distance (4-5km). In order to assess the origin of the Ml3.1 event, we report here on the data analysis performed from the seismological monitoring of the local area using the catalog produced by BCSF-RéNaSS and the regional public seismic networks. The main result is that the event is part of a remote triggered swarm that was initiated at least six days before the main shock and lasted more than two months. Template matching has been applied and allowed for a significant improvement of the detections. Double-difference relocations evidenced a set of conjugated faults in the swarm area that extends over 800m. Focal mechanisms of the two main events are very consistent with the known regional fault in the area. The regional stress field in combination with the fault orientation and a Coulomb failure criterion shows that the swarm location is in an unstable domain if the cohesion of the fault is weak, particularly sensitive to stress perturbations.

How to cite: Schmittbuhl, J., Lengliné, O., Lambotte, S., Grunberg, M., Doubre, C., Vergne, J., Cornet, F., and Masson, F.: A triggered seismic swarm below the city of Strasbourg, France on Nov 2019, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-18712, https://doi.org/10.5194/egusphere-egu2020-18712, 2020

D929 |
Jer-Ming Chiu, Kwanghee Kim, Shu-Chioung Chiu, Suyoung Kang, Wooseok Seo, and Jongwon Han

The 2017 Pohang earthquake (ML 5.4) is the second largest earthquake occurred in an intraplate in modern Korea and is considered the largest induced earthquake from an EGS system around the world.  The mainshock was proceeded by a few foreshocks and followed by a few thousands of aftershocks.  Numerous densely distributed seismic stations in local and regional distances were deployed to monitor this earthquake sequence.  Original hypocenters in the Pohang region were located using HYPODD that is independent on crustal structures.  A comprehensive crustal Vp and Vs model was recently available from an invited committee of foreign experts based on well logs and regional seismic data.  This model is then revised, especially the uppermost few hundred meters, based on results from a study of S to P converted waves from shallow interfaces beneath various stations, from the traditional Wadati plots analysis, and from the interpretation of two short seismic reflection/refraction profiles.  From continuous data, 5 to 10 folds of additional earthquakes than the original manually picked events can be identified and located. P and S arrival times from all earthquakes are re-picked from continuous data and are relocated using the revised model and Hypoellipse program.  Temporal and spatial distribution of relocated seismicity at depths range from 3 to 7 km are more clustered and confined than that from the original catalog.  A few thin vertical cross-sectional views of hypocenters parallel and perpendicular to the seismicity reveal that seismicity propagates along multiple NE-SW trending faults beneath the Pohang basin and extending NE offshore into East Sea.  These fault system is sandwiched between the Yongshan fault and a few other secondary faults to the south.  The main shock (5.4) and the two largest aftershocks (4.3 and 4.6) as well as their associated aftershocks show predominantly NE-SW strike-slip with reverse faulting propagating along three different adjacent faults.  Geometry of active faults and their tectonic implications will be presented and discussed in the meeting.

How to cite: Chiu, J.-M., Kim, K., Chiu, S.-C., Kang, S., Seo, W., and Han, J.: Imaging of active faults from the temporal and spatial distributions of relocated seismicity induced beneath an EGS site in the Pohang region of southeastern Korea, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-8228, https://doi.org/10.5194/egusphere-egu2020-8228, 2020

D930 |
The fault that moved during the 2017 Mw 5.4 Pohang earthquake in SE Korea; implications for the EGS project and the generation of the Pohang earthquake
Toshihiko Shimamoto, Shengli Ma, Lu Yao, Tetsuhiro Togo, HyunJee Lim, and Moon Son
D931 |
Serge Shapiro and Jin-Han Ree

A strong earthquake of Mw5.5 occurred on 15 November 2017, shortly after finishing borehole fluid injections performed for the geothermal development of the Pohang Enhanced Geothermal System. With a high probability, the earthquake was triggered by these operations. In this work we consider the Pohang Earthquake in the frame of the Seismogenic Index Model. We attempt to estimate the triggering probability of this event as well as a general  probability of triggering of arbitrary-magnitude earthquakes at the Pohang site before and after the termination of the fluid injections. A fluid injection in a point of an infinite continuum is taken here as a prototype of the Pohang situation.

The seismogenic index of the Pohang site is approximately between -2 and -1. During the injection operations, one can observe  a tendency of the
seismogenic index to increase with time. This was possibly  an indication of a gradual involvement of seismically more active zones in the stimulated domain. Especially alarming was the event of Mw3.3 on April 15th of 2017. Probably, this event indicated a jump of the seismogenic index to -1. All injection operations in both boreholes should be stopped after this event.

Our estimate of the probability of the Pohang earthquake is approximately 15%. One of  decisive factors for  this relatively high probability was the low b value. A combination of a low b-value and a rather high seismogenic index made the probability of a hazardous event significant. A termination of all injection operations after the occurrence of the event of M_w3.3 would significantly reduce the probability of an M_w5.5 event down to approximately 3%. An injection termination at M_w2.3 would reduce it down to approximately 1%.

The Pohang earthquake has a clear character of a triggered event. A real-time well developed seismic observation system permitting a precise 3-D event location and a monitoring of the temporal evolution of the geometry of the stimulated volume and of the seismogenic index could potentially help to prevent or to delay the occurrence of such an  earthquake.

This paper provides a simplified consideration based on analytical formulations for an effective homogeneous porous medium and monotonic injection operations. Numerical simulations of more realistic injection configurations,  an analysis of modeling results along the indicated here directions, further enhanced processing and analysis of seismologic records are required for more detailed understanding of processes led to the Pohang event. 

How to cite: Shapiro, S. and Ree, J.-H.: Modeling of the Pohang Earthquake Probability Using the Seismogenic Index , EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-18960, https://doi.org/10.5194/egusphere-egu2020-18960, 2020

D932 |
Ryan Haagenson and Harihar Rajaram

It has been long understood that the injection of fluids into the subsurface, a common practice in several industries, often leads to seismic activity by altering the fluid pressures and stress states acting along fault structures. In some cases, this puts the people and infrastructure located nearby at considerable risk. The effective mitigation of this potential hazard relies heavily on understanding the physical mechanisms controlling the behavior of injection-induced seismicity. Here, we aim to better understand the spatiotemporal patterns of the seismicity through the concept of triggering fronts (i.e. the propagating front where the onset of seismicity occurs). Previously, triggering fronts have been studied mostly in the context of homogenous porous media. Here, field scale simulations of fluid injection into fractured rock are modeled as linear, uncoupled fluid flow. While injection-induced seismicity is certainly affected by poroelastic stressing and nonlinear hydraulic parameters of the rock, the focus of this study is to understand the impact of a discrete fracture network on patterns of seismicity. Therefore, poroelastic and nonlinear effects are ignored. Results indicate that the pathways of high permeability within the fracture network greatly influence the migration of the triggering front. While the triggering front clearly follows a diffusive process as expected, the corresponding diffusivity is found to be distinct from the effective hydraulic diffusivity of the domain. We, therefore, call this diffusivity the seismic diffusivity of the fractured rock. Understanding seismic diffusivity may help us better interpret datasets of injection-induced seismicity and potentially forecast the patterns of injection-induced seismicity in well-characterized formations.

How to cite: Haagenson, R. and Rajaram, H.: Seismic diffusivity: The influence of fracture networks on the patterns of induced seismicity., EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-20308, https://doi.org/10.5194/egusphere-egu2020-20308, 2020

D933 |
Omid Khajehdehi, Kamran Karimi, and Jörn Davidsen

Seismic hazard due to fluid invasion in hydraulic fracturing, wastewater disposal, and enhanced geothermal systems has become a concern for industry and nearby residents. One of the challenges associated with this seismic hazard is the estimation of the spatial effects of these industry operations. Based on a large set of real-world fluid-induced seismicity catalogs, it was recently found that the spatial decay of seismic activity with distance from injection wells exhibits two typical behaviors: short-range decay and long-range decay. The distinction between the two groups can be captured by the exponent in the seismicity density but the underlying origin remains unknown. Here, we introduce a novel conceptual model that not only can capture the observed frequency magnitude distribution of fluid-induced seismic events but also explains different spatial decay exponents observed. In particular, previous models of fluid-induced seismicity have assumed that the permeability and porosity field is either uniform or random and spatially uncorrelated. However, power-law scaling in the spatial frequency power spectrum of well-logs, S(k)∝1/k^β, has been observed for many different physical properties of rocks such as sonic velocity, porosity, and log(permeability). Our model takes advantage of this by introducing a spatially correlated field for porosity and permeability. Our analysis shows that increasing β can decrease the spatial decay exponent, leading to more seismic activity at larger distances from the injection site. In particular, our model explains the two different types of behavior in the spatial distribution of fluid-induced microseismic events as a consequence of different correlations in permeability.

How to cite: Khajehdehi, O., Karimi, K., and Davidsen, J.: The effect of correlated permeability on spatiotemporal distribution of microseismic events in a conceptual model of fluid-induced seismicity, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-5826, https://doi.org/10.5194/egusphere-egu2020-5826, 2020

D934 |
Hideo Aochi and Jonny Rutqvist

We consider seismogenic asperities loaded by aseismic slip on a fault, which is induced by fluid circulation, as a simple example of fault reactivation. For this purpose, we combine two methods. The TOUGH2 (Transport Of Unsaturated Ground water and Heat) code is used for modeling the pore pressure evolution within a fault and then a Boundary Integral Equation Method (BIEM) is applied for simulating fault slip, including aseismic slip on the entire fault plane and fast slip on seismogenic asperities. The fault permeability is assumed stress-dependent and therefore is not constant but varies during a simulation. We adopt the Coulomb friction and a cyclic slip-strengthening-then-weakening friction model governing the fault slip, which allows for repeated asperity slip. We were able to demonstrate the entire process from the fluid injection, pore pressure increase, aseismic slip to seismogenic asperity slip. We tested a step-like increase of injection rate with time, which is common for hydraulic fracturing and reservoir stimulation at deep geothermal sites. Under this configuration, the pore pressure increase is not proportional to the injection rate, as the permeability depends on the stress.  Fault slip on seismogenic asperities is triggered repeatedly by surrounding aseismic slip. We find, in a given example, that the reccurence of the fast slip on asperity is approximatively proportional to the injected fluid volume, inferring that the aseismic slip amount increases proporitionally to the fluid volume as well.

How to cite: Aochi, H. and Rutqvist, J.: Hydro-mechanical modeling of seismogenic asperity loaded by aseismic slip through TOUGH-BIEM simulation, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-6643, https://doi.org/10.5194/egusphere-egu2020-6643, 2020

D935 |
Brice Lecampion, Federico Ciardo, Alexis Saèz Uribe, and Andreas Möri

We investigate via numerical modeling the growth of an aseismic rupture and the possible nucleation of a dynamic rupture driven by fluid injection into a fractured rock mass. We restrict to the case of highly transmissive fractures compared to the rock matrix at the scale of the injection duration and thus assume an impermeable matrix. We present a new 2D hydro-mechanical solver allowing to treat a large number of pre-existing frictional discontinuities. The quasi-static (or quasi-dynamic) balance of momentum is discretized using boundary elements while fluid flow inside the fracture is discretized via finite volume. A fully implicit scheme is used for time integration. Combining a hierarchical matrix approximation of the original boundary element matrix with a specifically developed block pre-conditioner enable a robust and efficient solution of large problems (with up to 106 unknowns). In order to treat accurately fractures intersections, we use piece-wise linear displacement discontinuities element for elasticity and a vertex centered finite volume method for flow.

We first consider the case of a randomly oriented discrete fracture network (DFN) having friction neutral properties. We discuss the very different behavior associated with marginally pressurized versus critically stressed conditions. As an extension of the case of a planar fault (Bhattacharya and Viesca, Science, 2019), the injection into a DFN problem is governed by the distribution (directly associated with fracture orientation) of a dimensionless parameter combining the local stress criticality (function of the in-situ principal effective stress, friction coefficient and local fracture orientation) and the normalized injection over-pressure. The percolation threshold of the DFN which characterizes the hydraulic connectivity of the network plays an additional role in fluid driven shear cracks growth. Our numerical simulations show that a critically stressed DFN exhibits fast aseismic slip growth (much faster than the fluid pore-pressure disturbance front propagation) regardless of the DFN percolation threshold. This is because the slipping patch growth is driven by the cascades of shear activation due to stress interactions as fractures get activated. On the other hand, the scenario is different for marginally pressurized / weakly critically stressed DFN. The aseismic slip propagation is then tracking pore pressure diffusion inside the DFN. As a result, the DFN percolation threshold plays an important role with low percolation leading to fluid localization and thus restricted aseismic rupture growth.

We then discuss the case of fluid injection into a fault damage zone. Using a linear frictional weakening model for the fault, we investigate the scenario of the nucleation of a dynamic rupture occurring after the end of the injection (as observed in several instances in the field). We delimit the injection and in-situ conditions supporting such a possibility.

How to cite: Lecampion, B., Ciardo, F., Saèz Uribe, A., and Möri, A.: Induced seismicity associated with fluid injection into a fractured rock mass, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-13568, https://doi.org/10.5194/egusphere-egu2020-13568, 2020

D936 |
Mohammad Hadi Mehranpour, Suzanne J. T. Hangx, and Chris J. Spiers

Predicting reservoir compaction resulting from fluid depletion is important to assess potential hazards and risks associated with fluid production, such as surface subsidence and induced seismicity. Globally, many producing oil and gas fields are experiencing these phenomena. The giant Dutch Groningen gas field, the Netherlands, is currently measuring up to 35 cm of surface subsidence and experiencing widespread induced seismicity. To accurately predict reservoir compaction, reservoir-scale models incorporating realistic grain-scale microphysical processes are needed. As a first step towards that aim, Discrete Element Method (DEM) modeling can be used to predict the compaction behavior of granular materials at the cm/dm-scale, under a wide range of conditions representing realistic in-situ stress and pressure conditions.

Laboratory experiments on the reservoir of the Groningen gas field, the Slochteren sandstone, have shown elastic deformation, inelastic deformation due to clay film consolidation, and inelastic deformation due to grain sliding and grain failure. Since the available contact models for DEM modeling do not yet incorporate all of these grain-scale processes, a new contact model, the Slochteren sandstone contact model (SSCM), was developed to explicitly take these mechanisms into account and integrate them into Particle Flow Code (PFC), which is a powerful DEM approach.

In SSCM the blunt conical contact with an apex angle close to 180˚ is assumed to properly model the elastic behavior, as well as the grain failure mechanism. Compacting an assembly of particles with this type of contact model, results in a range of contact shapes, from point to long contacts, which is compatible with microstructural observations of Slochteren sandstone.  The deformation of thin intergranular clay coatings is implemented following the microphysical model proposed by Pijnenburg et al. (2019a).

The model allows for the systematic investigation of porosity, grain size distribution and intergranular clay film content on compaction behavior. The model was calibrated against a limited number of hydrostatic and deviatoric stress experiments (Pijnenburg et al. 2019b) and verified against an independent set of uniaxial compressive experiments (Hol et al. 2018) with a range of porosities, grain size distributions and clay content. The calibrated model was also used to make predictions of the compaction behavior of Slochteren sandstone. These predictions were compared to field measurements of in-situ compaction and showed an acceptable match if the uncertainties of field measurements are considered in calculations.


Pijnenburg, R.P.J., Verberne, B.A., Hangx, S.J.T. and Spiers, C.J., 2019. Intergranular clay films control inelastic deformation in the Groningen gas reservoir: Evidence from split‐cylinder deformation tests. Journal of Geophysical Research: Solid Earth.

Pijnenburg, R.P.J., Verberne, B.A., Hangx, S.J.T. and Spiers, C.J., 2019. Inelastic deformation of the Slochteren sandstone: Stress‐strain relations and implications for induced seismicity in the Groningen gas field. Journal of Geophysical Research: Solid Earth.

Hol, S., van der Linden, A., Bierman, S., Marcelis, F. and Makurat, A., 2018. Rock physical controls on production-induced compaction in the Groningen Field. Scientific reports, 8(1), p.7156.

How to cite: Mehranpour, M. H., Hangx, S. J. T., and Spiers, C. J.: Discrete Element Method modelling of Groningen reservoir compaction using a new contact model describing elastic and inelastic grain-scale interactions, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-13941, https://doi.org/10.5194/egusphere-egu2020-13941, 2020

D937 |
Alessandro Verdecchia, Bei Wang, Yajing Liu, Rebecca Harrington, Marco Roth, Andres Peña Castro, and John Onwuemeka

The Dawson-Septimus area near the towns of Dawson Creek and Fort St. John, British Columbia, Canada has experienced a drastic increase in seismicity in the last ~ 6 years, from no earthquakes reported by Natural Resources Canada (NRCan) prior to 2013 to a total of ~ 200 cataloged events in 2013 – 2019. The increase follows the extensive horizontal drilling and multistage hydraulic fracturing activity that started to extract shale gas from the unconventional siltstone resource of the Montney Formation. In addition to hydraulic fracturing, ongoing wastewater disposal in the permeable sandstones and carbonates located stratigraphically above and below the Montney formation may also be contributing to elevated seismicity in the region. Earthquakes occur in close spatial and temporal proximity to hydraulic fracturing wells, at distances up to ~ 10 km. The expected diffusion time scales in the low-diffusivity siltstone rock units and the temporal and spatial scale of seismic activity beg questions about the possible processes controlling the location and timing of earthquakes.


Here, we investigate the causative mechanisms for two of the largest events in the Montney Basin, British Columbia: the August 2015 M4.6 earthquake near Fort St. John, and the November 2018 M4.5 earthquake near Dawson Creek. Both events are thought to have occurred within the crystalline basement, ~2 km below the injected shale units (Montney formation).  We use a finite-element 3D poroelastic model to calculate the coupled evolution of elastic stress and pore pressure due to injection at several hydraulic fracturing stages. Initially, we consider a simple layered model with differing hydraulic parameters based on lithology. Subsequently, also considering the seismicity distribution for each sequence, we introduce hypothetic hydraulic conduits connecting the injection intervals with the crystalline basement, where the respective mainshock occurred. We test a range of permeability values (10-15 m2– 10-12 m2) commonly implemented for fault zones.


Our results show that, for both cases, the poroelastic stress perturbation may be not sufficient to trigger events in the basement. Instead, a scenario with a high-permeability (10-13 m2– 10-12 m2) conduits connecting the Montney formation to the fault responsible for the mainshock could better explain the relationship between the hydraulic stimulation and the timing of the two M > 4 earthquakes. For the 2018 M4.5 event, aftershock distribution can be mainly attributed to earthquake-earthquake interaction via static Coulomb stress transfer from the mainshock slip. In addition to the modeling of single well/event sequences, future work will include the long-term poroelastic effect due to multiple disposal wells located in the region.

How to cite: Verdecchia, A., Wang, B., Liu, Y., Harrington, R., Roth, M., Peña Castro, A., and Onwuemeka, J.: Exploring possible causative mechanisms for earthquakes triggered by hydraulic fracturing: examples from the Montney Basin, BC, Canada., EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-10680, https://doi.org/10.5194/egusphere-egu2020-10680, 2020

D938 |
Yajing Liu, Alessandro Verdecchia, Kai Deng, and Rebecca Harrington

Fluid injection in unconventional hydrocarbon resource exploration can introduce poroelastic stress and pore pressure changes, which in some cases may lead to aseismic slip on pre-existing fractures or faults. All three processes have been proposed as candidates for inducing earthquakes up to 10s of kilometers from injection wells. In this study, we examine their relative roles in triggering fault slip under both wastewater disposal and hydraulic fracturing scenarios. We first present modeling results of poroelastic stress changes on a previously unmapped fault near Cushing, Oklahoma, due to injection at multiple wastewater disposal wells within ~ 10 km of distance, where over 100 small to moderate earthquakes were reported between 2015/09 to 2016/11 including a Mw5.0 event at the end of the sequence. Despite the much larger amplitude of pore pressure change, we find that earthquake hypocenters are well correlated with positive shear stress change, which dominates the regimes of positive Coulomb stress change encouraging failure. Depending on the relative location of the disposal well to the recipient fault and its sense of motion, fluid injection can introduce either positive or negative Coulomb stress changes, therefore promoting or inhibiting seismicity. Our results suggest that interaction between multiple injection wells needs to be considered in induced seismicity hazard assessment, particularly for areas of dense well distributions. Next, we plan to apply the model to simulate poroelastic stress changes due to multi-stage hydraulic fracturing wells near Dawson Creek, British Columbia, where a dense local broadband seismic array has been in operation since 2016. We will investigate the relative amplitudes, time scales, and spatial ranges of pore pressure versus solid matrix stress changes in influencing local seismicity.

Finally, we have developed a rate-state friction framework for calculating slip on a pre-existing fault under stress perturbations for both the disposal and hydraulic fracturing cases. Preliminary fault slip simulation results suggest that fault response (aseismic versus seismic) highly depends on 1) the relative timing in the intrinsic earthquake cycle (under tectonic loading) when the stress perturbation is introduced, 2) the amplitude of the perturbation relative to the background fault stress state, and 3) the duration of the perturbation relative to the “memory” timescale governed by the rate-state properties of the fault. Our modeling results suggest the design of injection parameters could be critical for preventing the onset of seismic slip.

How to cite: Liu, Y., Verdecchia, A., Deng, K., and Harrington, R.: Coupling poroelastic stress change and rate-state fault slip models to simulate fluid injection induced seismic and aseismic slip, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-6369, https://doi.org/10.5194/egusphere-egu2020-6369, 2020

D939 |
Linus Villiger, Dominik Zbinden, Antonio Pio Rinaldi, Paul Antony Selvadurai, Hannes Krietsch, Valentin Gischig, Joseph Doetsch, Mohammadreza Jalali, Florian Amann, and Stefan Wiemer

Several decameter-scale in-situ stimulation experiments were conducted in crystalline rock at the Grimsel Test Site, Switzerland, with the aim to advance our understanding of the seismo-hydro-mechanical processes associated with deep geothermal reservoir stimulation. To allow comparability between the experiments, a standardized injection protocol was applied for all experiments. Induced seismicity was recorded using acoustic emission sensors and accelerometers, which were distributed along tunnel walls and within four boreholes. Hydro-mechanical responses of the fault zones were measured using grouted longitudinal fiberoptic strain sensors and open pressure monitoring borehole intervals. A total of four ductile shear zones (with brittle overprint) and two brittle-ductile shear zones have been stimulated during these experiments.

Here we present an analysis of heterogeneous permeability evolution within a target shear zone during ongoing stimulation. The shear zone in question is an originally ductile shear zone which contains a single fracture in the injection interval. The observed planar seismicity cloud indicates that most of the stimulation process was confined within the target shear zone. Hydraulic characterization of the injection interval before and after stimulation revealed an enhancement in interval transmissivity from 8.3-10-11 m2/s to 1.5-7 m2/s. Within the reservoir, the seismo-hydro-mechanical data (i.e. seismicity cloud, pressure peaks and local deformation) spatiotemporally coincide, suggesting that permeability enhancement along the shear zone is highly localized and heterogeneous. Thus, we argue that the permeability evolution is linked to asperity distribution and breakdown within the shear zone.

The conceptual model developed from the experimental analysis is implemented in a three-dimensional numerical model, with which we attempt to simulate the directional permeability creation observed in the experiment. The model accounts for a discrete planar fault zone of finite thickness with distributed low-permeability, brittle asperities embedded in a more permeable damage zone mimicking the ductile shear zone at Grimsel. The hydro-mechanical processes are modeled with the TOUGH-FLAC simulator, which sequentially couples fluid flow and poroelastic deformation within the fault and the surrounding medium. A Mohr-Coulomb failure criterion is used to simulate asperity reactivation, which can lead to permeability enhancement of the reactivated area.

How to cite: Villiger, L., Zbinden, D., Rinaldi, A. P., Selvadurai, P. A., Krietsch, H., Gischig, V., Doetsch, J., Jalali, M., Amann, F., and Wiemer, S.: The influence of the heterogeneous asperity distribution on induced seismicity and permeability evolution during hydraulic fault zone stimulation, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-18774, https://doi.org/10.5194/egusphere-egu2020-18774, 2020

D940 |
Dominik Zbinden, Antonio Pio Rinaldi, Tobias Diehl, and Stefan Wiemer

Industrial projects that involve fluid injection into the deep underground (e.g., geothermal energy, wastewater disposal) can induce seismicity, which may jeopardize the acceptance of such geo-energy projects and, in the case of larger induced earthquakes, damage infrastructure and pose a threat to the population. Such earthquakes can occur because fluid injection yields pressure and stress changes in the subsurface, which can reactivate pre-existing faults. Many studies have so far focused on injection into undisturbed reservoir conditions (i.e., hydrostatic pressure and single-phase flow), while only very few studies consider disturbed in-situ conditions including multi-phase fluid flow (i.e., gas and water). Gas flow has been suggested as a trigger mechanism of aftershocks in natural seismic sequences and can play an important role at volcanic sites. In addition, the deep geothermal project in St. Gallen, Switzerland, is a unique case study where an induced seismic sequence occurred almost simultaneously with a gas kick, suggesting that the gas may have affected the induced seismicity.

Here, we focus on the hydro-mechanical modeling of fluid injection into disturbed reservoir conditions considering multi-phase fluid flow. We couple the fluid flow simulator TOUGH2 with different geomechanical codes to study the effect of gas on induced seismicity in general and in the case of St. Gallen. The results show that overpressurized gas can affect the size and timing of induced earthquakes and that it may have contributed to enhance the induced seismicity in St. Gallen. Our findings can lead to a more detailed understanding of the influence of a gas phase on the induced seismicity.

How to cite: Zbinden, D., Rinaldi, A. P., Diehl, T., and Wiemer, S.: Multi-phase hydromechanical modeling of induced seismicity: general insights and the case study of the deep geothermal project in St. Gallen, Switzerland, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-21428, https://doi.org/10.5194/egusphere-egu2020-21428, 2020

D941 |
Vanille A. Ritz, Antonio P. Rinaldi, Elisa Colas, Raymi Castilla, Peter M. Meier, and Stefan Wiemer

Monitoring micro-seismicity during operations of a geothermal field is critical to the understanding of seismic hazard and changes in the reservoir. In the context of a geothermal project, induced earthquakes are an important tool to enhance the permeability and thus productivity of reservoirs and to image structure and processes. However, felt and/or damaging earthquakes are a major threat to societal acceptance and regulatory license to operate. With the adaptive data-driven tool ATLS (Adaptive Traffic Light System), we aim at managing and mitigating the risk posed by induced earthquakes during stimulation and operations, while at the same time ensuring and optimising the productivity.

The demonstration site for the application of ATLS lies in the Hengill volcanic region located in the South-West of Iceland, host to two power plants (Hellisheiði and Nesjavellir) with a total production capacity of 423 MWe and 433MWth. The production of energy and heat is accompanied by reinjection of the spent geothermal water in dedicated areas, both to maintain production and to comply with legal requirements. These reinjection areas have been showing different seismic responses to drilling and injection operations. We investigate these different behaviours by performing numerical modelling for two of the reinjection regions.

Two models are compared: TOUGH2-Seed, a full 3-dimensional stochastic simulator and an analytical model based on a cumulative density function linking maximum pressure in the reservoir and reactivation. Those two models fulfil two different aspects of the development of an ATLS, with the full 3D allowing an in-depth dive in the driving mechanisms of induced seismicity; and the analytical solution providing a robust and fast approximation of the forecast for real-time application. We show that both models can reproduce observed seismicity patterns in the Hengill geothermal field.

How to cite: Ritz, V. A., Rinaldi, A. P., Colas, E., Castilla, R., Meier, P. M., and Wiemer, S.: Geomechanical modelling of spent fluid reinjection in the Hengill geothermal field, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-10907, https://doi.org/10.5194/egusphere-egu2020-10907, 2020

D942 |
Lisa Johann and Serge A. Shapiro

It is understood that the recent acceleration of seismic event occurrences in Kansas and Oklahoma, U.S., can be connected to the large-volume disposal of wastewater. These highly saline fluids are co-produced with oil and gas and are re-injected under gravity into the highly porous Arbuckle aquifer. Since 2015, injection rates have been decreasing. However, the seismic hazard in that region remains elevated. Furthermore, it has been noticed that some events in Kansas occur far from disposal wells.

To analyse spatio-temporal patterns between the fluid injection and earthquake locations, we applied a time-dependent 2D cross-correlation technique. This reveals a vectorial migration pattern of the seismic events. Whereas early events occur towards the east-sourtheast, later events are located preferably in northeastern direction of large volume injectors. With time, event locations migrate further in that direction. We explain this observation as well as measured Arbuckle pore pressures by a directional pore-fluid pressure diffusion and poroelastic stress propagation. This also follows from our principal two-dimension poroelastic finite element model which is of predictive power and identifies controlling parameters of the observations. These are mainly the permeability of the target injection formation and the seismogenic basement as well as the anisotropic permeability and the critical fault strength distribution. Our results lead to the conclusion that remote locations are destabilised also when injection rates are declining.

Thus, volume reductions may only provide a direct effect to lower earthquake rates locally. However, a state-wide decrease of the seismicity may require longer times such that the seismic hazard due to wastewater disposal induced seismicity may remain for decades. 

How to cite: Johann, L. and Shapiro, S. A.: Large-Scale Migration Patterns of Wastewater-Induced Earthquakes in the Central U.S., EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-5245, https://doi.org/10.5194/egusphere-egu2020-5245, 2020

D943 |
Francesco Parisio, Victor Vilarrasa, Wenqing Wang, Olaf Kolditz, and Thomas Nagel

Geothermal energy is a fundamental piece of the puzzle in the carbon-free energy transition necessary to mitigate adverse effects of climate change. In this context, so-called supercritical geothermal systems (where resident brine is above its critical point) can reach a power output of up to 50 MW per-well and are becoming a reality thanks to the research effort of the international community. Supercritical systems are inherently complex because of fluid and solid rheology and because of the non-linear couplings involved between pore pressure, temperature and deformation in the porous and fractured rock. In this contribution, we have performed finite element analyses of coupled thermo-hydro-mechanical (THM) conditions in a supercritical geothermal system. The formulation employed includes a porosity-dependent permeability relationship that is derived through mass balance equations of the solid skeleton. The equations of state of water are based on IAPWS standards and span the whole range of temperature and pressure of interest in supercritical geothermal systems. We have analyzed an injection-production doublet scenario at 5.5 km depth, where the two wells are spaced 500 m apart and a major fault lies between them. When injecting in isothermal conditions, because of water mobility of the supercritical phase is higher than the liquid phase, minor perturbations are observed in the major fault. Seismicity increases rapidly following a pressure-diffusion response and only micro-seismicity is expected as the stress in the major fault shows minor changes. On the contrary, cold fluid injection generates large thermal-induced stress changes that diffuse following advective heat transport laws. Micro-seismicity is quickly triggered and, within 10 years in the current setting, the perturbation reaches the 250 m distant fault, where larger magnitude events are possible. Our findings bear important consequences in terms of safety of supercritical geothermal systems and proved that seismic response in amplified by re-injection-induced cooling.

How to cite: Parisio, F., Vilarrasa, V., Wang, W., Kolditz, O., and Nagel, T.: Cooling effects on induced seismicity in supercritical geothermal systems, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-16283, https://doi.org/10.5194/egusphere-egu2020-16283, 2020

Chat time: Thursday, 7 May 2020, 16:15–18:00

Chairperson: Antonio Pio Rinaldi, Marco Scuderi, Rebecca Harrington
D944 |
Victor Vilarrasa, Francesco Parisio, Roman Makhnenko, Haiqing Wu, and Iman Rahimzadeh Kivi

Geological media is envisioned as a strategic resource to store large volumes of CO2 and mitigate climate change. Geo-energy applications, such as geologic carbon storage, geothermal energy, and subsurface energy storage, involve injection and extraction of fluids that cause pressure diffusion. Pore pressure changes may induce seismicity, especially in faults that intersect the injection formation or are hydraulically connected with it. We numerically study with finite element analysis of coupled hydro-mechanical conditions how fault stability is affected by fluid injection into a porous aquifer that is overlaid and underlain by low permeability clay-rich formations. We model a layered sedimentary basin with alternating soft and low permeability with stiff and high permeability formations and include the crystalline basement at the bottom. Additionally, a low permeability steep fault, whose offset ranges from zero to the reservoir thickness, crosses the system. We consider a normal faulting stress regime typical of extensional environments. Simulation results show that the reservoir pressurization as a result of fluid injection causes significant stress changes around the fault that affect its stability. The stress changes depend on the stiffness of the rock juxtaposed to the pressurized reservoir. If there is no offset, the rock is stiff on both sides of the fault, inducing a homogeneous horizontal total stress increase along the thickness of the reservoir. As a result, the deviatoric stress becomes smaller and the induced seismicity potential is low. As the fault offset increases, some part of the base rock gets juxtaposed to the pressurized reservoir. The soft base rock deforms more than the reservoir rock in response to the reservoir expansion, inducing a lower horizontal total stress. Thus, fault stability reduces when the pressurized reservoir rock is juxtaposed with the softer base rock. This finding shows that the induced seismicity potential may increase with the fault offset.

How to cite: Vilarrasa, V., Parisio, F., Makhnenko, R., Wu, H., and Rahimzadeh Kivi, I.: The role of fault offset in induced seismicity potential, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-12863, https://doi.org/10.5194/egusphere-egu2020-12863, 2020

D945 |
Oliver Heidbach, Moritz Ziegler, and Sophia Morawietz

The Bavarian Molasse Basin is one of the largest European areas for low-enthalpy hydrothermal applications. In the last decades, more than 20 geothermal applications for district heating and even power generation have been established, and more are planned or under construction. At about one third of the projects, seismicity of up to Ml 2.4 has been observed after the onset of production and reinjection of fluids while no seismicity has been observed at the other project sites. In order to assess the potential for induced seismicity at a specific site, three parts of information are required: (1) The initial stress state, (2) the changes in the stress state due to production and reinjection, (3) a fault geometry (strike, dip), and (4) a fault properties in terms of a failure criterion. While a modelling of the production and/or injection induced stress changes is performed commonly and basic information on the failure behaviour of rocks is available, information on stress magnitudes is rare and unevenly distributed. Thus, 3D geomechanical-numerical modelling is used to estimate the stress state in a target area based on the few data records available.

We present a 3D geomechanical-numerical model of the initial stress state in the Bavarian Molasse Basin in order to assess the individual potential for induced seismicity at different geothermal sites. Our model area contains several lithological units with different rock properties. Additionally, in our approach, we quantify the uncertainties introduced by the variability of the 13 stress magnitude data records used for the calibration of the model. We further reduce the large uncertainties by introduction of additional observables that limit the range of acceptable stress states. From the model results, we extract the stress state and its uncertainties at the two geothermal sites Aschheim/Feldkirchen/Kirchheim and Poing that are in a distance of 4 km. While seismicity of up to Ml 2.1 has been observed in Poing, the other site remained seismically quiet. Our modelled stress state at the two sites in combination with according failure criteria on optimally oriented faults is in agreement with the seismological observations. Even considering uncertainties of 2σ, the modelled stress state in Aschheim/Feldkirchen/Kirchheim is stable, while in Poing even the average stress state is already critical. These results indicate the relevance of a geomechanical assessment of sites of subsurface applications in order to minimize the potential for induced seismicity.

How to cite: Heidbach, O., Ziegler, M., and Morawietz, S.: Geomechanical Assessment of Potential for Induced Seismicity, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-10514, https://doi.org/10.5194/egusphere-egu2020-10514, 2020

D946 |
Camilla Rossi, Francesco Grigoli, Simone Cesca, Sebastian Heimann, Paolo Gasperini, Vala Hjörleifsdóttir, Torsten Dahm, Christopher J. Bean, and Stefan Wiemer

Geothermal systems in the vicinity of the Hengill volcano, SW Iceland, started to be exploited for electrical power and heat production since the late 1960s, and today the two largest operating geothermal power plants are located at the Nesjavellir and the Hellisheidi. This area is a complex tectonic and geothermal site, being located at the triple junction between the Reykjanes Peninsula (RP), the Western Volcanic Zone (WVZ), and the South Iceland Seismic Zone (SISZ). The region is seismically highly active with several thousand earthquakes located yearly. The origin of such earthquakes may be either natural or anthropogenic. The analysis of microseismicity can provide useful information on natural active processes in tectonic, geothermal and volcanic environments as well as on physical mechanisms governing induced events. Here, we investigate the microseismicity occurring in Hengill area to understand physical source mechanisms and the origin of these microseismic events. We use a very dense broadband monitoring network deployed since November 2018 with support of the GEOTHERMICA project COSEISMIQ and apply robust and full-waveform based methods for earthquake location, clustering analysis and source mechanism determination. Our dataset consists of about 637 events with ML ranging between 0.8 and 4.7 from December 2018 to January 2019. We use this rich and large dataset for testing a workflow for automated processing. Earthquake location and clustering analysis show that seismicity is spatially clustered, with shallower events at the center of geothermal site in proximity to geothermal plants, and deeper earthquakes in the southern part of the study area. Most of our moment tensors can suggest the influence of geothermal activity and geothermal energy exploitation operations on the subsurface. This work is supported by the COSEISMIQ project of the EU GEOTHERMICA program .

How to cite: Rossi, C., Grigoli, F., Cesca, S., Heimann, S., Gasperini, P., Hjörleifsdóttir, V., Dahm, T., Bean, C. J., and Wiemer, S.: Analysis of microseismicity in the Hengill Geothermal Area, SW Iceland, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-8704, https://doi.org/10.5194/egusphere-egu2020-8704, 2020

D947 |
Francesco Grigoli, Sebastian Heimann, Claus Milkereit, Stefan Mikulla, Nima Nooshiri, Malte Metz, Gesa Petersen, Simone Cesca, Vala Hjörleifsdóttir, Rögnvaldur Magnússon, Ragnheidur St. Ásgeirsdóttir, Hannes Hofmann, Marco Broccardo, Dimitrios Karvounis, Arnaud Mignan, Kristjan Augustsson, Stefan Audunn Stefansson, Gunter Zimmermann, Torsten Dahm, and Stefan Wiemer

At Geldinganes Island, Reykjavik, Iceland a hydraulic stimulation was recently conducted to enhance the productivity of an existing hydrothermal well. An experimental cyclic soft stimulation concept was applied. Seismic risk was assessed with an appropriate monitoring network which was set up and operated before, during, and for some time after the stimulation activities. An advanced traffic light system was developed and operated for the first time in this setup.

A crucial element in such traffic light systems is the real-time monitoring of background and induced seismicity. During the experiment, real-time seismograms from the monitoring network were streamed over the internet to three different institutions (ISOR, ETHZ and GFZ), where they were analysed independently, with different combinations and setups of automatic, semi-automatic and manual methods. Both, classic pick based approaches and modern full-waveform methods were applied. Locations, magnitudes, and centroid moment tensor solutions were determined.

Many things can go wrong in real-time or near-real-time processing of seismic data. Sensor failures, transmission failures, timing issues, processing hardware failures, computational limitations, software bugs and human error, just to name a few. In a temporary network the challenges are additionally salted by the need to validate sensor responses, orientations, gain factors and site conditions in a short time frame between station setup and beginning of the experiment. Furthermore, tuning of advanced analysis methods can be difficult without example events at hand.

In this contribution, we would like to share our lessons learned in near-real-time processing of data from a heterogeneous temporary seismic network. 

How to cite: Grigoli, F., Heimann, S., Milkereit, C., Mikulla, S., Nooshiri, N., Metz, M., Petersen, G., Cesca, S., Hjörleifsdóttir, V., Magnússon, R., Ásgeirsdóttir, R. St., Hofmann, H., Broccardo, M., Karvounis, D., Mignan, A., Augustsson, K., Stefansson, S. A., Zimmermann, G., Dahm, T., and Wiemer, S.: Contribution of continuous waveform processing to induced seismicity realtime monitoring during geothermal stimulation at Geldinganes, Iceland. , EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-16199, https://doi.org/10.5194/egusphere-egu2020-16199, 2020

D948 |
Ortensia Amoroso, Ferdinando Napolitano, Vincenzo Convertito, Raffaella De Matteis, and Paolo Capuano

Nesjavellir Geothermal Field is located in the Northern part of the Hengill central volcano in South West Iceland. The Hengill volcanic complex consists of three smaller volcanic systems feeding several geothermal fields with surface manifestations.

Geothermal energy is currently produced at two power plants, in Nesjavellir and in Hellisheidi. After an exploitation period started in 1947, the construction of Nesjaveillir power plant was completed in 1990. Nowadays it produces geothermal energy of up to 300 MW, which is 1,640 l/sec of hot water and up to 120 MW of electricity.

Part of the surplus geothermal water from the plant goes into the injection wells and in analogy with the nearby Hellisheidi power plant the re-injection of geothermal gases into basaltic formations is planned. To this aim several tests of fluids deep injection are being conducted to prepare the experimental re-injection of carbon dioxide and hydrogen sulphide.

In the framework of the H2020-Science4CleanEnergy project, S4CE, a multi-disciplinary project aimed at understanding the underlying physical mechanisms underpinning sub-surface geo-energy operations and to measure, control and mitigate their environmental risks, we investigate the seismicity evolution through the b-value and study the elastic properties of the propagation medium through the 3D/4D seismic tomography.

The seismicity recorded in the study area is due to different mechanisms. Indeed, while in Hengill the seismicity is originated by volcano-tectonic processes, small earthquake swarms between Hengill and Grensdalur volcano are due to the geothermal activity. Finally, the seismicity in proximity of Hellishedi and Nesjaveiilir power plant appears to be induced by re-injection of waste water from the geothermal production.

Seismic data are recorded by the Icelandic Meteorological Office (IMO) but also from Iceland GeoSurvey (ÍSOR) and by the COSEISMIQ project. The production data are from the OR energy company.

We used an iterative linearized delay-time inversion to estimate both the 3D P and S velocity models and earthquake locations. The velocity model is parametrized by trilinear interpolation on a 3D grid. The inversion starts from the 1D velocity model, optimized for the area. Time variations of the medium seismic properties are observed in term of Vp, Vs and Vp/Vs ratio obtained by 4D tomography. The technique consists in applying the 3D tomography at consecutive epochs. Spatial and temporal characteristics of the re-located earthquakes are then analysed by using the ZMAP code to image the b-value in the investigate volume.

The images obtained for each epoch in terms of b-value, Vp and Vs velocities are then correlated with operational data.


This work has been supported by S4CE ("Science for Clean Energy") project, funded from the European Union’s Horizon 2020 - R&I Framework Programme, under grant agreement No 764810 and by PRIN-2017 MATISSE project funded by Italian Ministry of Education and Research.

How to cite: Amoroso, O., Napolitano, F., Convertito, V., De Matteis, R., and Capuano, P.: Seismicity time evolution and 3D/4D seismic tomography of Nesjavellir geothermal field (Iceland), EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-8462, https://doi.org/10.5194/egusphere-egu2020-8462, 2020

D949 |
Olivier Lengliné, Léna Cauchie, and Jean Schmittbuhl

The injection of fluid in the upper crust, notably for the development or exploitation of geothermal reservoirs, is often associated with the onset of induced seismicity. Although this process has been largely studied, it is not clear how the injected fluid influences the rupture size of the induced events. Here we re-investigate the induced earthquakes that occurred during an injection at Soultz-sous-Forêts, France in 1993 and studied the link between the injected fluid and the source properties of the numerous induced earthquakes. We take advantage that deep borehole accelerometers were running in the vicinity of the injection site. We estimate the moment and radius of all recorded events based on a spectral analysis and classify them into 663 repeating sequences.  We show that the events globally obey the typical scaling law between radius and moment. However, at the scale of the asperity, fluctuations of the moment are important while the radii remain similar suggesting a variable stress drop or a mechanism that prevents the growth of the rupture. This is confirmed by linking the event source size to the geomechanical history of the reservoir. In areas where aseismic slip on pre-existing faults has been evidenced, we observed only small rupture sizes whereas in part of the reservoir where seismicity is related to the creation of new fractures, a wider distribution and larger rupture sizes are promoted. Implications for detecting the transition between events related to pre-existing faults and the onset of fresh fractures are discussed.

How to cite: Lengliné, O., Cauchie, L., and Schmittbuhl, J.: Seismic asperity size evolution during fluid injection: case study of the 1993 Soultz-sous-Forêts injection, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-13421, https://doi.org/10.5194/egusphere-egu2020-13421, 2020

D950 |
Michele D'Ambrosio, Carlo Giunchi, Davide Piccinini, Rebecca Bruni, and Gilberto Saccorotti

Located in the inner sector of the Northern Apennines (southern Tuscany), Mt. Amiata is a quaternary volcano that was active in between 0.3 My and 0.19 My b.p.. It now hosts a water-dominated geothermal field which is being exploited for the production of electric power since the early 1960s. Historical records report at least ten moderate (M < 5.3), yet damaging earthquakes, occurred well before the geothermal exploitation started. Nonetheless, public concern is rapidly raising due to the possibility that the geothermal production processes (i.e., vapor extraction and re-injection of condensates back into the subsurface) provoke stress perturbations that may trigger earthquakes. A critical issue thus consists in discerning whether seismicity is related to the exploitation of the geothermal resource, rather than to the natural tectonic stresses at the site. Here, we report data from a temporary seismic network operated at Mt Amiata during the 2016-2019 time span. We obtained precise, absolute locations using a non-linear probabilistic location procedure and a minimum-misfit, 1-D velocity model specifically derived for the area. Complementary information on the velocity structure was also derived from the inversion of the surface-wave group- and phase-velocity dispersion curves, as obtained from frequency-time analysis (FTAN) of regional earthquakes and noise correlation functions, respectively. Our catalog amounts to more than 1000 earthquakes, with a completeness magnitude of about 0.4. The improvement in earthquake detection with respect to the catalog from the national monitoring program is of the order of 1 magnitude unit. Hypocenters are clustered within a few, distinct focal volumes, two of which closely correspond to the productive fields. Most hypocenters are shallower than 5km, getting deeper as the distance from the geothermal areas increases. Overall, the lower bound of hypocentral depths follows the K-horizon, a regional-scale seismic reflector inferred to mark the upper limit of the brittle-to-ductile transition and, possibly, the top of the Pleistocene granitic intrusions. Throughout the observation period, the largest magnitude observed within the geothermal area is ML=2.9 (ML=2.1 in the national catalog); 95% of the earthquakes have magnitudes lower than 1. Earthquakes occur at a rather constant rate of less than 1 event/day, occasionally interspersed by short-duration (1-3 days), swarm-like bursts accounting for tens of earthquakes that do not exhibit any clear mainshock-aftershock sequence. The scaling relationships of the catalog are examined by computing clustering in the magnitude-distance-time space domains. The background and stationary components have similar relevance (55% and 45%, respectively), thus not resulting diagnostic about the principal driving mechanism of the observed seismicity. Although the proximity between the main focal volumes and production areas would suggest that geothermal exploitation plays a major role in the earthquake-generation process, the lack of any industrial data prevents from inferring any causative relationship between the two phenomena.

How to cite: D'Ambrosio, M., Giunchi, C., Piccinini, D., Bruni, R., and Saccorotti, G.: Seismicity in the geothermal area of Mt. Amiata Volcano (Italy), EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-9942, https://doi.org/10.5194/egusphere-egu2020-9942, 2020

D951 |
José Ángel López-Comino, Martin Galis, P. Martin Mai, Xiaowei Chen, and Daniel Stich

Exploring the connections between injection wells and seismic migration patterns is key to understanding processes controlling growth of fluid-injection induced seismicity. Numerous seismic clusters in Oklahoma have been associated with wastewater disposal operations, providing a unique opportunity to investigate migration directions of each cluster with respect to the injection-well locations. We introduce new directivity migration parameters to identify and quantify lateral migration toward or away from the injection wells. We take into account cumulative volume and injection rate from multiple injection wells. Our results suggest a weak relationship between migration direction and the cluster-well distances. Migration away from injection wells is found for distances shorter than 5-13 km, while an opposite migration towards the wells is observed for larger distances, suggesting an increasing influence of poroelastic stress changes. This finding is more stable when considering cumulative injected volume instead of injection rate. We do not observe any relationship between migration direction and injected volume or equivalent magnitudes.

How to cite: López-Comino, J. Á., Galis, M., Mai, P. M., Chen, X., and Stich, D.: Lateral migration patterns toward or away from injection wells for earthquake clusters in Oklahoma , EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-2649, https://doi.org/10.5194/egusphere-egu2020-2649, 2020

D952 |
Kilian B. Kemna, Alexander Wickham-Piotrowski, Andrés F. Peña-Castro, Elizabeth S. Cochran, and Rebecca M. Harrington

The LArge-n Seismic Survey in Oklahoma (LASSO) array recorded local seismicity in 2016 in a region of active saltwater disposal. The month-long deployment of 1,833 vertical-component nodes had a nominal station spacing of 400 m and covered a 25 by 32 km2 area. We estimate local event magnitudes and focal mechanisms of the induced seismicity using the vertical component waveforms from a catalog of 1375 earthquakes. Here we use the developed catalog to investigate the spatio-temporal evolution of seismicity and the source properties of the induced events.

The catalog is complete to a local magnitude of ~0.9, with a b-value of ~1.1. Focal mechanisms, which we determined using the HASH method, show a mix of strike-slip and normal faulting. The majority of the events are located at 1.5 – 5.0 km depth, where injection depths range from 0.1 – 2.0 km, and the basement contact is located at 1.5 – 2.5 km. Analysis of the coefficient of variation of interevent times suggests that the time evolution of seismicity is close to Poissonian, with minimal temporal clustering. We observe spatial clustering, with larger (M > 2) events occurring within dense clusters near the footprint of the array.

The dense station coverage of the array permits the exploration of variations in corner frequency and resulting stress drop estimates as a function of azimuth, i.e. radiation pattern. We calculate stress drops for the local catalog within 5 km of the array footprint from individual spectral and spectral ratio corner frequency values. Single spectra corner frequency estimates for events within the array footprint on individual nodes show evidence of variation related to radiation pattern, and vary as much as 100% from the mean for an individual event. Stress drop estimates from spectral ratio corner frequency estimations range between 10 – 100 MPa, show self-similar scaling, and fall within the typical range observed for intraplate (tectonic) earthquakes. Both single spectra and spectral ratio corner frequency estimates show a significant sample bias in the corner frequency estimation by using less than ~10 stations, and highlight the importance of azimuthal coverage for the stability of spectral estimates.

How to cite: Kemna, K. B., Wickham-Piotrowski, A., Peña-Castro, A. F., Cochran, E. S., and Harrington, R. M.: Exploring the Evolution and Source Properties of Injection-Induced Seismicity in Northern Oklahoma Using a Large-N Seismic Array, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-7553, https://doi.org/10.5194/egusphere-egu2020-7553, 2020

D953 |
Marco Pascal Roth, Alessandro Verdecchia, Kilian B. Kemna, John Onwuemeka, Rebecca M. Harrington, and Yajing Liu

An increasing number of M3+ earthquakes have been associated with Hydraulic Fracturing (HF) injection activity in low-permeable tight shale formations in the Western Canada Sedimentary Basin (WCSB) in the last decade. These include a Mw 4.6 on 08/17/2015 near Ft. St. John, a ML 4.5 on 11/30/2018, and two ML 3.2 on 10/05/2019, 10/08/2019 near Dawson Creek, British Columbia. Increased seismic activity in the Dawson-Septimus area prompted a temporary deployment of seismic stations in a joint effort between McGill University and the Ruhr University Bochum in order to perform higher-resolution monitoring relative to the regional seismic station coverage. Here, we use waveform data from that deployment of 22 (dominantly broadband) stations in close proximity to numerous HF wells in an area of roughly 60 x 70 km2, between July 2017 and August 2019, as well as records from 6 additional seismic stations northwest of the study area. In total, we detect 6222 local earthquakes, of which 5325 surpass a quality control criterion of having a horizontal location error ≤ 3 km. An investigation of the spatial and temporal correlation between injection and earthquake initiation using a cross-correlation based event similarity analysis during seismically active time periods reveals a high degree of event similarity within various clusters and a strong correlation with individual injection episodes at specific HF wells. In addition, event clusters also exhibit similar patterns in daily cumulative seismic moment, independent of differences in waveform characteristics.

As individual clusters may represent the activation of specific geological structures, we perform double-difference relative relocation of seismicity to identify fault orientations. In addition, we invert for focal mechanism solutions per event cluster to check consistency with structures inferred with relocated hypocenters, and perform spectral fitting for source parameter analysis. Event relocations are performed on individual families, where the total catalog is divided into subsets corresponding to 24 seismic active time periods where 43 event families are active. Relocating each earthquake family separately allows us to successfully relocate 4571 out of the total 5325 events. The relative relocations align in two dominant orientations, with one roughly perpendicular to the maximum horizontal regional stress orientation, and the other at low angles to the maximum regional stress orientation on a regional scale around individual HF wells. Focal mechanism estimates for events with M > 2.0 result in two primary groups of faulting mechanisms: strike-slip deformation on faults implied by lineations striking at low angles to SH, and thrust-faulting deformation on faults implied by lineations perpendicular to SH. Seismic moment and corner frequency estimates from single spectrum and spectral ratio fitting as well as scaling relations will be presented.

How to cite: Roth, M. P., Verdecchia, A., Kemna, K. B., Onwuemeka, J., Harrington, R. M., and Liu, Y.: Earthquake source parameter analysis shows hydraulic fracturing induced events are consistent with fault reactivation under regional stress in northeastern British Columbia, Canada, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-8240, https://doi.org/10.5194/egusphere-egu2020-8240, 2020

D954 |
Guiyun Gao, Chandong Chang, Chenghu Wang, and Jin Jia

We conduct geomechanical study for a seismogenic fault in Hutubi underground gas storage site, northwestern China. The Hutubi reservoir has undergone active production from 1990s to 2012, leading to a complete depletion, and then sequential gas injection and extraction from 2013 for the gas storage project. First, we constrain the orientation and magnitudes of the stress state at the reservoir depths (~3.6 km depth) at the time of a complete depletion in 2012, using image-logged wellbore breakouts in a borehole. Then we estimate the variation of the stress state with time as a result of pore pressure change based on a simple assumption of coupling between horizontal stresses and pore pressure. Our results show that the stress state was initially in a reverse faulting regime before production and switched to a strike-slip faulting regime during production. Gas injection from 2013 turned the stress regime again in favor of reverse faulting. We use the estimated variation of the reservoir stress state with time to calculate temporal changes of slip tendency of the major earthquake fault (Hutubi fault) in the reservoir. Slip tendency of the fault decreased continuously with production, and then increased with injection. The first earthquake swarm associated with gas injection occurred ~2 months after the commencement of injection, possibly due to slow pore pressure diffusion. Thereafter, earthquakes were induced whenever gas was injected, while few earthquakes were detected during gas extraction phases. Our preliminary assessment of slip tendency suggests that earthquake swarms are induced during increasing phases of pore pressure when slip tendency reaches a value between 0.4 and 0.5, which can provide information on friction coefficient of the fault.

Funding information: This work is supported by the National Natural Science Foundation of China (41574088,41704096) 

How to cite: Gao, G., Chang, C., Wang, C., and Jia, J.: Variation of slip tendency of a seismogenic fault associated with fluid extraction and injection in the Hutubi underground gas storage, China, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-7742, https://doi.org/10.5194/egusphere-egu2020-7742, 2020

D955 |
Sonja Gaviano, Davide Piccinini, Luisa Valoroso, Luigi Improta, and Carlo Giunchi

The southern Apennines range hosts a well documented case of protracted Reservoir Induced Seismicity (RIS) associated to the Pertusillo artificial lake. Since the deployment of a local monitoring network in 2001, M3+ swarms were recorded to the south of this medium-sized water reservoir. Interpretation in terms of RIS relies on the positive correlation found between seasonal water level changes and earthquake rate that increases during the winter-spring refill. We present a new high-resolution catalogue of RIS obtained by running a matched-filter (MF) detection technique on data recorded during a dense passive survey between 2005-2006. We aim at producing a very-high quality catalogue in terms of completeness magnitude (Mc) and hypocenter location accuracy to precisely track the spatio-temporal distribution of seismicity, pinpoint the activated faults, investigate the rupture mechanisms and the role played by crustal fluids in triggering RIS. All these issues are critical to improve understanding of the physical mechanism behind the RIS.

Our initial catalogue includes 406 handpicked templates recorded by 3C 24-stations temporary network run by INGV. Local magnitudes range between 0.06 and 2.63, with a MC of 0.4. Templates are correlated to the 13-month-long data streams by the MF algorithm. A matched event is declared when the average value of cross-correlation function (CC) computed over all stations exceeds 0.65. The procedure furnishes 10056 matched events with associated P- and S-phase automatic picks, weighted according to the uncertainties of template event picks and the CC values of each trace. Matched events are preliminary located in a 1-D model using the NonLinLoc software and then selected based on quality criteria. The final catalog has MC=0.1 and includes 6012 high-quality events with ML > -0.9 that are then relocated through the high-precision double-difference relative technique. We recognize four main clusters confined at 2-6 km depth within a fractured, liquid-bearing carbonate antiform characterized by high-Vp (>6.0 km/s) and very-high Vp/Vs ratio (>2.0) that indicates high-pressure pore fluids. Hypocentral alignments delineate NW-trending high-angle faults dipping to the NE or SW that measure up to 2 km along strike and dip. Prevailing extensional focal mechanisms are coherent with the fault geometry and local stress field. These results suggest re-activation of inherited thrust-faults with associated back-thrusts optimally oriented in the present extensional stress field.  

The spatiotemporal seismicity distribution indicates a positive correlation between the seasonal oscillation of the lake level and the progressive activation of the 4 clusters of seismicity. Distant clusters from the PWR are delayed with respect to the closer ones, suggesting that seismicity migrates away from the reservoir following a pore fluid pressure triggering process. The b-value is high and it also varies with time between 1.2 and 1.8 with a trend anti-correlated to the lake level. Therefore, the proportion of large earthquakes to small ones increases during the re-fill stage characterized by intense earthquake production and vice-versa. The two southern clusters, more distant from the lake, with events that delineate clear fault-zones, share the lower b-values (1.4).

How to cite: Gaviano, S., Piccinini, D., Valoroso, L., Improta, L., and Giunchi, C.: New insights into the Pertusillo Lake reservoir induced seismicity (Italy) from a high-resolution matched-filter earthquake catalogue, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-9168, https://doi.org/10.5194/egusphere-egu2020-9168, 2020

D956 |
Francesca Silverii, Djamil Al-Halbouni, Magdalena Stefanova Vassileva, Gudrun Richter, Rongjiang Wang, Borja Gonzalez Cansado, Sebastian Hainzl, Torsten Dahm, and Francesco Maccaferri

Within the framework of the SECURE project, we test modeling techniques used for natural geothermal and volcanic reservoirs and apply them to anthropic underground gas storage facilities. These systems indeed share similar mechanics and physical properties, however gas reservoirs are often extensively monitored, and better imaged. In order to manage fluctuations between gas supply and demand, natural gas can be temporarily stored in different underground storage facilities, such as depleted gas/oil fields, natural aquifers, and salt cavern formations. When properly monitored during storage and withdrawal (production) of gas, these systems provide a unique opportunity to investigate how reservoirs evolve at different time scales, modify the surrounding stress state, produce deformation coupled with diffusion processes, and possibly induce/trigger earthquakes on nearby faults.

In the first case study we addressed within the framework of SECURE project, we take advantage of well constrained reservoir geometry and physical parameters, records of gas injection/production rates, pore pressure variations, and a local seismic catalog at a gas reservoir in Spain. We implement a poro-elastic model to simulate pressure temporal variations, estimate related stress-state variations, and study eventual relationship with small recorded seismic events. The model is based the software POEL by Wang et al., (2003), a semi-analytical physics-based numerical scheme which allows the computation of transient and steady-state solutions in response to pore-pressure variations. Being 2D axisymmetric, POEL drastically simplify the geometry of the reservoir, but it is particularly suitable to link observables such as pressure variations within the reservoir with the physical/mechanical processes occurring in the surroundings.

In the second case study we address the stability condition for salt caverns which has been excavated for salt mining purposes. We make use of 2D discrete-element geomechanical models to compare numerical simulation results with field observations in terms of surface subsidence. With this numerical model we consider different pressure conditions for the fluid (brine) filling the cavity, and return different scenarios for the stability of a salt cavern. Such modeling effort aims at improving our understanding of middle-to-long term stability conditions, for those cavities that have been dismissed after anthropic operations such as salt extraction, but also seasonal gas storage.

How to cite: Silverii, F., Al-Halbouni, D., Stefanova Vassileva, M., Richter, G., Wang, R., Gonzalez Cansado, B., Hainzl, S., Dahm, T., and Maccaferri, F.: Underground seasonal storage of gas: testing numerical modelling tools with application to i) a deep aquifer-layer, and ii) salt caverns., EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-10311, https://doi.org/10.5194/egusphere-egu2020-10311, 2020

D957 |
Bo Zhang, Baoshan Wang, Zhanbo Ji, Jinxin Hou, Lu Li, Bin Wei, Nier Wu, and Zhide Wu

Underground Gas Storages (UGSs) are important large-scale industrial facilities used to bridge the gap between the natural gas demand and supply. The UGS production can cause periodical changes of the subsurface stresses and probably change the seismicity pattern. The UGS operation related seismicity has important affects on the UGS, but is rarely reported. Hutubi UGS is the largest Underground Gas Storage in China and well equipped with seismic observations from the beginning of the UGS operation in Jun. 2013. The Hutubi UGS provides an unprecedented opportunity to study the seismicity related gas injection. The seismicity around the Hutubi UGS was detected and located by using matched filter and double difference relocation techniques. More than 7000 earthquakes were detected and located within 20 km of the UGS from Jan. 2011 to Dec. 2018 (i.e., 2 years before the operation and 6 injection-extraction cycles). The seismicity can be clustered into three groups South of, North of, and beneath the UGS. Two (South and North) of three groups occurred along two south-dipping planes with dipping angle ~40 degree, corresponding to local geological structures. While the underlaying group occurred along the direction conjugate to the other two groups, which was in accordance with the starting of the second injection stage and gradually migrated deeper. The northern cluster occurred mainly after the Murghob earthquake (M7.2) and the Hutubi earthquake (M6.2), which may be related to dynamic stress triggering from these two earthquakes. The seismicity in the southern cluster persists, but the rate shows seasonality, which is likely modulated by the gas operation and underground water exploitation. The seismicity in the Hutubi UGS area may have different origins. Understanding the mechanism of the impact of UGS operation on different clustering seismicity will help us to optimize the production parameters and reduce the risk of induced earthquake.

How to cite: Zhang, B., Wang, B., Ji, Z., Hou, J., Li, L., Wei, B., Wu, N., and Wu, Z.: Spatial and temporal evolution of micro-earthquakes during Multi-cycle operation of the Hutubi underground gas storage, Xinjiang, China, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-9984, https://doi.org/10.5194/egusphere-egu2020-9984, 2020

D958 |
Rebecca M. Harrington, Hongyu Yu, Honn Kao, Yajing Liu, and Bei Wang

Seismicity related to fluid injection during unconventional oil and gas exploration has increased dramatically in North America in the last decade. The Western Canadian Sedimentary Basin experienced a significant increase in the number of M3+ earthquakes, including several M4+ associated with high-pressure stimulation during Hydraulic Fracturing (HF) activity. The vigorous seismic response to injection activity and low historical seismicity rates pose critical questions as to the triggering mechanism(s) and seismic hazard assessment in the affected areas. To monitor seismicity linked to injection, a dense local network of eight broadband seismic stations was installed in 2015 at distances of ~2 km around an active well pad with the purpose of monitoring seismicity prior to, and following, a HF well stimulation in the Montney Play in British Columbia, Canada. Here we present an earthquake source process study using observations from the local station network, and provide evidence for a slow-rupture seismic signal which may bridge the spectrum of fault slip rates from aseismic near the well bore, to typical seismic velocities at distances beyond ~1 km.

Initial detection and relocation of seismicity between May 28 – October 15, 2015 yielded 350 well-constrained hypocenters of high-frequency events with a maximum magnitude of Mw 1.8 that resemble typical tectonically generated earthquakes. The detection procedure also yielded a total of 31 events with high-frequency (or broadband) onsets, that transition to protracted, low-frequency ringing relative to event magnitude, which we term hybrids. Both hybrid and high-frequency events occur at similar depths to the active well bore and at distances of ~1-2 km from injection stages, yet exhibit varying source characteristics in spite of their proximal source volumes. Hybrid waveforms are marked by broader P- and S-wave arrival pulse shapes, and spectral fitting suggests that the stress drop values are roughly an order of magnitude lower than high-frequency events, with average static stress drop values of 0.3 MPa and 4.9 MPa, respectively. We interpret wider phase arrival pulse widths and lower stress drop values as resulting from lower rupture velocities of hybrid events relative to high-frequency events. A dilatant strengthening effect would be expected in close proximity to the well bore, and near the hybrid sources, where material is weaker and pore pressures are elevated, which may result in slower rupture propagation when slip is initiated relative to further distances where material damage and pore pressure perturbation are both lower. Thus, hybrid earthquakes may mark regions where slip velocities transition from aseismic sliding directly next to the well bore, which has been observed in laboratory and meso-scale experiments, to typical seismic velocities at further distances. The size-duration scaling of the induced hybrids observed here also extends the scaling of slow earthquakes occurring in tectonic fault transition zones, and may provide the first observations to extend the scaling down to seismic moment values of ~1010.

How to cite: Harrington, R. M., Yu, H., Kao, H., Liu, Y., and Wang, B.: Hydraulic fracturing induced hybrid earthquakes in the Montney Basin, British Columbia, Canada may mark the transition from aseismic slip near the wellbore to seismic slip at greater distances, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-4544, https://doi.org/10.5194/egusphere-egu2020-4544, 2020

D959 |
Adam Klinger and Max Werner

Hydraulic fracturing underpins tight shale gas exploration but can induce seismicity. During stimulations, operators carefully monitor the spatio-temporal distribution and source parameters of seismic events to be able to respond to any changes and potentially reduce the chances of fault reactivation. Downhole arrays of geophones offer unique access to (sub) microseismic source parameters and can provide new insights into the processes that induce seismicity. For example, variations in stress drop might indicate changes in the seismic response to injection (e.g. pore pressure variations). However, borehole arrays of geophones and the high frequencies of small events also present new challenges for source characterization. Stress drop depends on the corner frequency, a parameter with great uncertainty that is sensitive to attenuation, especially for (sub-) microseismicity. Here, we explore the behavior of microseismic spectra measured along borehole arrays and the effect of attenuation on estimates of corner frequency. We examine a dataset of over 90,000 microseismic events recorded during hydraulic fracturing in the Horn River Basin, British Columbia. We only see clear phase arrivals for events Mw > -1 and restrict our initial analysis to a subsample of Mw> 0 events that vary in space and time.

Our first observation is that some stations in the borehole array show an unexpected increase in the displacement energy from the low frequency to the corner frequency in the P and SH phases as well as high-frequency energy spikes inconsistent with a smooth Brune source model. A shorter time window that only captures the direct arrival results in a flatter low frequency plateau and reduces the amplitude of the pulses but compromises the resolution. The spikes may be caused by high frequency coda energy. We also find that corner frequency estimates decrease with decreasing station depth along the array in both the P and SH phases, a likely result of high frequency attenuation along the downhole array. The findings suggest Brune corner frequencies of moment magnitudes < 0.5 may not be resolvable even with downhole arrays at close proximity. Our results will eventually contribute to a better characterization of microseismic source parameters measured in borehole arrays.


How to cite: Klinger, A. and Werner, M.: Insights into downhole array spectra of hydraulic-fracturing induced seismicity in the Horn River Basin, British Columbia, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-20394, https://doi.org/10.5194/egusphere-egu2020-20394, 2020

D960 |
Antoine Delvoye and Ben Edwards
Recent examples have shown that fluid injection during hydraulic fracturing can result in felt, or even damaging, seismic activity. In the vicinity of the Preston New Road site (Lancashire, UK), almost 200 earthquakes of ML -0.8 to 2.9 have been recorded by the British Geological Survey (BGS) over the period from October 15th 2018 to September 2019. This corresponds to the period during which hydraulic fracturing (fracking) was carried out by the operator, Cuadrilla Resources. Throughout the operation, fracking had to be suspended temporarily five times as the ML 0.5 ‘red light’ of the UK regulatory Traffic Light System (TLS) was exceeded. Since 2017, the University of Liverpool has operated a seismic monitoring network comprising nine broadband Nanometrics Trillium 120 across the Blackpool-Preston region in order to determine the baseline seismicity and monitor induced events associated with the fracking operations atthe site. In addition to this network, both Cuadrilla and the BGS deployedseismometers-including borehole geophone strings-over the region making it oneof the best places in Europe for monitoring induced seismicity. The superficial geology of the region is dominated by thick sand, till and clay deposits, poten-tially leading to significant amplification of seismic waves. This amplification may lead to over-estimation of earthquake magnitude, and therefore increased likelihood of triggering mitigation measures associated with the TLS. In order to understand amplification effects near the PNR site, surface-wave measurements (both MASW and seismic Ambient Vibration Arrays, AVAs) have been used to derive dispersion curves and obtain VS profiles through an inversion process for station-site characterization. By using small local arrays (hundreds of meters wide) to regional arrays (tens of km wide), we reconstruct a velocity model down to the bedrock depth. This velocity model can then be used to compute a parshly non-ergodic ground motion and subsequently seismic hazard assessment. This approach allow us to account for site to site variability and result in reduced uncertainty in the hazard assessment. We find Vs30 in the range 200 – 300 m/s at the sites investigated, leading to significant amplification effects that may bias event magnitudes determined on a surface array.

How to cite: Delvoye, A. and Edwards, B.: Induced Seismicity at the Preston New Road Shale Gas Site in Lancashire, UK – Site Characterisation and impact on the TLS, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-20620, https://doi.org/10.5194/egusphere-egu2020-20620, 2020

D961 |
Gudrun Richter, Sebastian Hainzl, Peter Niemz, Francesca Silverii, Torsten Dahm, Gert Zöller, Arno Zang, and Francesco Maccaferri

In the framework of the Geo:N project SECURE (Sustainable dEployment and Conservation of Underground Reservoirs and Environment) we developed a Python software toolbox to model the rate and distribution of seismicity induced by anthropogenic stress changes at various production sites (gas production, hydrofracturing, gas storage). This toolbox tests different frictional behavior of the underground (linear or rate-and-state stressing rate dependent, critically or subcritically prestressed faults) and takes into account the uncertainties of the production site parameters. The knowledge on the location and orientation of pre-existing faults can be considered as well. Model parameters are estimated by fitting the model to recorded historical seismicity using a maximum likelihood approach. We discuss applications at conventional gas fields, hydraulic fracturing experiments and an aquifer gas storage site, covering a wide range of spatial and temporal scales of induced seismicity in different settings and for different production schemes. This enables to investigate the underlying physical processes by the comparison of the different models. Additionally, the model parameters are linked to frictional material properties and the best performing model can be used to forecast the seismicity rates in space and time with their uncertainties according to the production plans.

Induced seismicity at gas fields in the Northern Netherlands and in Germany have similar tectonic settings but very different extents, depths and production histories. The data set of two sites are compared which both show a large delay of the first recorded seismicity after the start of production. Using our model we can reproduce the long delay for both sites. Thanks to the long and detailed data set we successfully reproduce the spatiotemporal pattern of the seismicity of one site, whereas the limited number of seismic events result in large uncertainties for the other site. In the comparative testing of the models the critically prestressed rate-and-state model performs best. This means that the complete stressing history influences the resulting seismicity. We also applied the model to a hydraulic fracturing experiment in granite comparing data sets for different fracturing methods and different phases of a stimulation experiment. Hundreds of microearthquakes are localized in a volume of roughly 15x15m with increasing number of events for later refraction stages indicating the growth of rock fracturing. A third application is run for a gas storage in an aquifer layer, which is loaded by injection and production operations. Here the proportion of the tectonic versus the anthropogenic induced seismicity is investigated analyzing the varying number of small local earthquakes in the region.

How to cite: Richter, G., Hainzl, S., Niemz, P., Silverii, F., Dahm, T., Zöller, G., Zang, A., and Maccaferri, F.: Spatiotemporal modelling of induced seismicity with a stress-based statistical approach applied to different production sites, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-10905, https://doi.org/10.5194/egusphere-egu2020-10905, 2020

D962 |
Elisa Venturini, Licia Faenza, Irene Munafò, Mario Anselmi, Lucia Zaccarelli, Alexander Garcia‐Aristizabal, and Andrea Morelli

Ground motion prediction equations (GMPEs) based on data collected on hydrocarbon extraction areas have to deal with lower magnitude and smaller distances compared to the natural seismicity. This study focuses on the Mirandola-Cavone oil field located in the Po Plain (Northern Italy), an area that has been struck by the Emilia earthquake sequence in 2012.
We start with the compilation of a new homogeneous seismic catalogue, in terms of locations and moment magnitudes. The data come from the local network run by the industrial operator, integrated by the closest stations of the Italian seismic network managed by Istituto Nazionale di Geofisica e Vulcanologia.
Subsequently, we calculate the intensity parameters of interest (e.g., Peak Ground Acceleration, Peak Ground Velocity and spectral values) for the available set of about 250 earthquakes.
Lastly, we develop a functional form for the GMPE using software tools available at  IS-EPOS Platform, derived for the geometrical mean of the horizontal components of seismograms. The resulting attenuation curve is calibrated for magnitudes higher than 0.1 and distances up to 50 km, appropriate for monitoring local seismicity. This work represents the first attempt to construct the GMPEs for an oil field in Italy, starting from the raw data, in support to the Italian Guidelines for monitoring seismicity, deformation and pore pressure in hydrocarbon extraction areas.

How to cite: Venturini, E., Faenza, L., Munafò, I., Anselmi, M., Zaccarelli, L., Garcia‐Aristizabal, A., and Morelli, A.: Ground Motion Prediction Equations for shallow, small-magnitude events: application to the Mirandola-Cavone oil field, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-11794, https://doi.org/10.5194/egusphere-egu2020-11794, 2020

How to cite: Venturini, E., Faenza, L., Munafò, I., Anselmi, M., Zaccarelli, L., Garcia‐Aristizabal, A., and Morelli, A.: Ground Motion Prediction Equations for shallow, small-magnitude events: application to the Mirandola-Cavone oil field, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-11794, https://doi.org/10.5194/egusphere-egu2020-11794, 2020

How to cite: Venturini, E., Faenza, L., Munafò, I., Anselmi, M., Zaccarelli, L., Garcia‐Aristizabal, A., and Morelli, A.: Ground Motion Prediction Equations for shallow, small-magnitude events: application to the Mirandola-Cavone oil field, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-11794, https://doi.org/10.5194/egusphere-egu2020-11794, 2020

D963 |
Vincenzo Convertito, Vincenzo De Novellis, Francesco Casu, Riccardo Lanari, Fernando Monterroso, Sotiris Valkaniotis, and Nicola Alessandro Pino

In the last decades, the triggered seismicity has represented one of the most debated issue. Fluid pressure changes with/without fluid flow in rock fractures/pores, thermal stress changes due to temperature gradients, and the volumetric changes or mass removal/accumulation can be included in the geoengineering processes that can induce or trigger seismic activity.

In this work we analyse the Le Teil earthquake (LTe), which occurred on November 11, 2019 (MW 4.9) in Ardèche region (south-eastern France), as a possible triggered event originated by rock mass removal in the Lafarge quarry operating since 1833. The structural area where LTe occurred is part of the St. Thomé‐La Rouvière fault system, located in the Cévennes fault bundle, which marks the south-eastern border of the Massif central over almost 150 km long; this fault presents a NE-SW trend and its geometry characterized by several uncertainties due to the absence of dip measurements.

As first step, to estimate the removed volume of rock in the Lafarge quarry, we use multi-temporal digital surface models and, in particular, the archive stereo aerial image pairs from IGN for 1946, 1979, 2007 and 2011; our analysis is restricted to the modern period where detailed topographic data are available. Subsequently, we generate the coseismic deformation maps by applying the Differential Synthetic Aperture Radar Interferometry (DInSAR) technique to SAR data collected along ascending and descending orbits by the Sentinel-1 (S1) constellation of the European Copernicus Programme. In order to retrieve the seismogenic fault parameters, we jointly invert the so-generated S1 DInSAR measurements by performing a consolidated two-step approach: it consists of a non-linear optimization to constrain the fault geometry with uniform slip, followed by a linear inversion to retrieve the slip distribution on the fault plane.

Finally, our Coulomb stress changes analysis on the fault along the slip direction suggests a clear positive triggering relation between the long-term activity in the Lafarge quarry and the Le Teil earthquake.

This work is supported by: the 2019-2021 IREA-CNR and Italian Civil Protection Department agreement; the EPOS-SP project (GA 871121); and the I-AMICA (PONa3_00363) project.

How to cite: Convertito, V., De Novellis, V., Casu, F., Lanari, R., Monterroso, F., Valkaniotis, S., and Pino, N. A.: The Mw 4.9 Le Teil seismic event (France): a possible case of triggered seismicity, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-21210, https://doi.org/10.5194/egusphere-egu2020-21210, 2020

D964 |
Nicolai Gestermann and Thomas Plenefisch

Induced and triggered seismicity in Germany is related to various mining operations such as hydrocarbon extraction, geothermal exploitation and classical mining techniques, i.e. coal and potash mining.

After some larger events small damages to buildings were observed that might have been caused by the ground shakings. This led to public discussions on compensation and to political discussions on improving legal regulations. The possibility of damages caused by mining induced seismic events and difficulties in financial compensation reduced the acceptance of mining projects in the past, e.g. geothermal projects are inhibited.

In case of verified damage due to an induced event, the causative mining company has to pay compensations. In 2016 new legal regulations entered into force. The Federal Mining Act was revised with an improved legal situation for the population by expanding the prima facie evidence on mining activities using boreholes. The new legal regulations define, that damages at buildings are assumed to be caused by the seismic event in the responsibility of the operator of the mining activities, if they occur within a certain area defined by the mining authority (impact area, German: “Einwirkungsbereich”).

From the seismological perspective, local measurements of PGV are often rare. Thus, it is difficult to assess the damage potential of the seismic events in detail, especially if intensities are around V (EMS-98). In many cases, a relation between individual damages at buildings and the seismic event is only hardly verifiable. Actually, detailed survey reports could neither prove nor disprove the relation between damages and seismic events in some cases. In conclusion, some of the widely discussed events might have led to small damages.

A brief introduction about the existing legal regulations will be presented. We used synthetic seismogram to model the wave propagation and amplitude effects for induced seismic events in the magnitude range between ML 2.9 and 3.6, for which it was necessary to define the impact area for legal regulations. Results from amplitude measurements at existing seismic stations were taken to calibrate the absolute amplitudes of the modeling. The synthetic seismograms could help to quantify the effects from the radiation pattern of the source and the impact of sediment coverage between source and receivers. They could improve the definition of the area of impact.

How to cite: Gestermann, N. and Plenefisch, T.: Synthetic seismograms for assessing areas of potential damage after induced events for legal regulations in Germany, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-9381, https://doi.org/10.5194/egusphere-egu2020-9381, 2020

D965 |
Gregor Mokelke and Manfred Joswig

The Nothern German basin was considered aseismic in most of the seismic hazard maps of last century. Then seismic events occurred in the last decades, and were located in the vicinity of gas production. Until now more than 70 earthquakes are documented with magnitudes ranging from ML 1.0 to 4.5. Especially the 2004 Rotenburg ML 4.5 event caused much concern, and first locations of different authors disagreed in depth. Dahm et al (2007) argued for 5 km depth close to the horizons of gas production, and suggested a depletion-induced event. Macroseismic studies and other authors, however, determined focal depths of 8-12 km, clearly below gas production.
Within the last 15 years new stations from BGR (Bundesanstalt für Geologie und Rohstoffe) and the BVEG (Bundesverband Erdöl, Erdgas und Geoenergie e.V.) were established in the region between Cloppenburg and Soltau. Our own work is based on a small-scaled, but dense network with arrays and single stations that were installed from 2014 to 2018 in the eastern central part of the gas fields near Rotenburg. Results resolve that seismic activity can occur in a great range of depths down to 30 km, and it is not exclusively focussed on the reservoir horizons. We found strong dependence of depth determinations from parameter settings – notably vP/vS – and station selection. Besides obvious mis-locations of weak, low SNR events based on few phase readings we also traced this dependency back to the 2004 Rotenburg event which at that time was recorded only by a sparse network of remote stations.
In summary, the Northern German basin offers a complex regime of weak seismicity, ranging from single low-crust earthquakes to frequent, induced events of gas production. The 2004 Rotenburg event does not fit either category, and geomechanical modelling will be needed to decide on its relation to gas production.

How to cite: Mokelke, G. and Joswig, M.: Seismicity in the Northern German basin - from simple model to complex regime, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-20078, https://doi.org/10.5194/egusphere-egu2020-20078, 2020

D966 |
Sepideh Karimi, Adam Baig, Aaron Booterbaugh, Yoones Vaezi, Mark Stacey, Dario Baturan, and Benjamin Witten

Seismicity potentially induced through wastewater disposal, hydraulic fracture completion, or other industrial operations, has been a cause for increasing public concern over the last decade. Monitoring for this activity has focussed on the problems of location and characterization, often to a relatively rough degree of precision. Regulations typically spell out responses for operators should an event exceed a magnitude threshold within a specified distance of their facilities. While this type of monitoring is critical for ensuring injections be conducted effectively while minimizing potential damage from shaking and public alarm, it often leaves many unanswered questions in terms of the underlying processes.

Understanding these questions entails that we demand more out of the seismic networks, essentially upgrading the data products to a “next generation” level.  The data from the network needs to be used to provide a detailed understanding of critical geological structures and geomechanics of the study area. This goal is facilitated through both a densification of hardware and a higher order of event processing. High-precision locations delivered through relative relocation methodologies delineate slipping fault structures, often resolving previously unknown features. Moment tensor inversion processing also helps reveal the orientations of faults and provides information on stress in the region. The resolution of these structures provides critical insight into understanding how a field is reacting.

We illustrate the application of this “next-generation” seismicity monitoring system to the Delaware Basin in West Texas, where we have deployed a network of 25 broadband seismometers complementing monitoring from TexNet and other networks. Despite being an exceptionally challenging recording environment, by aggregating all of these data we obtain a high-resolution catalog of earthquake hypocenters delineating a number of fault features. Inverting the stresses from the moment tensors of the highest-quality events shows a dominantly normal stress regime and tangibly resolves a rotation of axes transitioning across the basin. We illustrate both the logistical and processing requirements necessary for timely delivery of results highlighting the dynamics of seismicity in an active study area.

How to cite: Karimi, S., Baig, A., Booterbaugh, A., Vaezi, Y., Stacey, M., Baturan, D., and Witten, B.: Seismic Monitoring for High-Precision Delineation of Fault Geometry and Stress; Case Study for the West Texas Subscription Array, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-12786, https://doi.org/10.5194/egusphere-egu2020-12786, 2020

D967 |
Alina Besedina and Dmitry Pavlov

Previous studies of microseismic noise before earthquakes in seismically active regions showed the possibility of detecting the preparation processes of seismic events. This effect manifests itself as decrease of the frequency of natural oscillations before earthquakes. In this work, the method for detecting the natural frequencies of blocks in seismic noise is adapted to a lower hierarchical level. Using known empirical relations for faults with a characteristic length of less than 500-1000 m, the characteristic natural oscillation frequencies that can be used to diagnose a fault zone are estimated. For small seismic events with magnitudes Mw from -2 to -1, we calculated the expected frequencies of natural vibrations of 350-1100 Hz, and for events with Mw from 0 to 1,  - below 35 Hz. For analysis, we used the recording data of high-frequency accelerometers at a depth of 300 m from the free surface in the area of the city of Gubkin (Russia) within iron ore deposits. Before small events with an amplitude of more than 0.01 m/s2. The intervals of decrease in the spectral centroid in the range of 20-1200 Hz were identified. The minimum values of the spectral centroid obtained on the basis of experimental data are generally in a good agreement with theoretical estimates. This work was supported by RFBR (project # 18-05-00923).

How to cite: Besedina, A. and Pavlov, D.: Development of the method for detecting natural frequencies of blocks in seismic noise before small seismic events, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-18098, https://doi.org/10.5194/egusphere-egu2020-18098, 2020

D968 |
Marcelo Bianchi, Lucas Schirbel, and Alexandre Ausgusto

We put SeisComp3 to test by using it to analyze a very dense (9 squared kilometers) local network of 712, four components sensors (stations). Each station had a 3-component accelerometer and a pressure sensor deployed at the ocean bottom, close to the Brazilian platform near an oil exploration field. Noise levels were extreme. During the two months of the operation time, the network recorded an earthquake swarm sequence, and later analysis indicated more than 1000 earthquakes detected in a one-hour interval employing a coherency stacking method. While still not a common practice, real-time earthquake detection and location in this situation would be beneficial since this could support decisions while drilling or oil recovering is in place. Traditional tools as SeisComp3 are routinely used and allows for real-time detection and location along with the rapid revision of regional and teleseismic events, but are not widely adapted to work in a very local environment. Our experience so far showed that SeisComp3 efficiently handled the data volume (4 components at 500 samples per second times 712 stations) with a modern average workstation. Traditional SEG-Y data can be routinely converted and fed in real-time to SeedLink FIFO using ObsPy. Still, data must be correctly rotated since SeisComp3 needs at least a vertical component. Processing workflow included parallel picking using scautopick with STA/LTA, nucleation of origins using scautoloc, and location using Locsat and Hypo71 tools. In this harsh environment, the optimal window size for STA is about the size of the P-wave (0.05-0.1 s) and, LTA is about 30-60 times the S-P times (60-120 s). Using those parameters, SeisComp3 managed to generate from 400-1200 readings per data channel. We fed all picks into scautoloc that handled origin nucleation and location. Despite parameters supplied to scautoloc, the tool has many limits and relations hardcoded that inhibit it from respecting maximum requested residuals. In other words, its nucleation algorithm is adapted to work on the teleseismic and regional scale. Actual results indicate that we were able to nucleate and locate only 10-20% of known origins. Due to the flexibility of the tool, we also developed a pipeline using S-waves only. S-waves had a higher SNR for the events of interest and, due to lower velocities, presents a larger moveout on the small array easing the location. Manually picked and relocated detections returned an RMS lower as 0.04 s. Additional tests performed using the Scanloc module (GEMPA closed source nucleator) showed a higher performance during the nucleation of new origins. In this case, Hypo71 was the used locator. We did not observe any clear difference between LocSat and Hypo71 performance once the earthquake source is nucleated, and a proper velocity model is supplied.

How to cite: Bianchi, M., Schirbel, L., and Ausgusto, A.: Realtime detection and location of very local seismicity using SeisComp3, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-3912, https://doi.org/10.5194/egusphere-egu2020-3912, 2020

D969 |
Gulam Babayev and Fakhraddin Kadirov (Gadirov)

Absheron peninsula (Azerbaijan) area was hit by the strong Caspian earthquakes on November 25, 2000 with Mw6.1 and 6.2 magnitudes. The seismic networks successfully recorded the foreshock, main shock and many aftershocks at respective locations. By using probabilistic analysis, magnitude of design earthquake for the current study in the oilfield was taken as 6.3. From this concept design (scenario) earthquake, accelerations were estimated for the distance of 35 km. In the second phase of the study, soil amplification factors and site characteristics data from boreholes were determined and estimated. In the next phase, the study uses synthesized accelerograms formed on the basis of simulation of the seismic wave propagation processes through ground layer aiming to determine the quantitative characteristics of seismic effect on the oilfield region. Soil amplification values estimated by empirical relationships in terms of shear wave velocities are in the range of 0.7 and 1.9 values. Shear wave velocity (Vs, 30) values are 100 and 110 (m/s). The PGA values for the study area were evaluated by considering the local site effects. Peak ground acceleration varies between 100 – 380 gal. On the basis of the empirical relationship between MSK-64 and peak ground acceleration, the special distribution of intensity of the design earthquake with intensity of >8 is represented. Finally, the study presents possible relationship between seismic effect and daily oil recovery which states the direct proportional characteristics.

Keywords: ground classification, oilfield, scenario earthquake, Vs30, amplification factor, peak ground acceleration

How to cite: Babayev, G. and Kadirov (Gadirov), F.: Earthquake-induced site effect in the oil field deposit of Absheron peninsula (Azerbaijan), EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-3965, https://doi.org/10.5194/egusphere-egu2020-3965, 2020

D970 |
Hongfeng Yang, Pengcheng Zhou, Nan Fang, Gaohua Zhu, Wenbin Xu, Jinrong Su, Fanbao Meng, and Risheng Chu

Coinciding with the extensive hydraulic fracturing activities in the southern Sichuan basin, seismicity in the region has surged in the past a few years, including a number of earthquakes with magnitudes larger than 5. On 25 February 2019, an ML4.9 earthquake struck the Rongxian County, Sichuan, China and caused 2 fatalities and 12 injuries, the first deadly earthquake associated with shale gas production. The earthquake was preceded by two foreshocks with magnitudes of ML4.7 and ML4.3 within two days. We relocated the earthquake sequence using local and regional seismic network, and obtained the focal depths of the mainshock and two foreshocks at 1 and 3 km, respectively, much shallower than the report from catalogue. Most other smaller quakes were located at 2-6 km. The mainshock had also been well captured by InSAR images, which confirmed the shallow depth of ~1 km. Both seismic and geodetic data yielded thrust faulting mechanism for the mainshock, consistent with the mapped Molin fault in the region. The two foreshocks, however, occurred on an unmapped fault that has different orientation than the Molin fault. Injection wells are found in the vicinity of the two foreshocks and the fracking depth (~2.7 km) coincides with their focal depths, suggesting a possible causal relationship. The mainshock is located in the region with positive Coulomb failure stress caused the two foreshocks. The value of Coulomb failure stress change is 0.03 bar, smaller than the typical static triggering threshold. Therefore, the mainshock is likely caused by fracking by poroelastic stress transfer.

How to cite: Yang, H., Zhou, P., Fang, N., Zhu, G., Xu, W., Su, J., Meng, F., and Chu, R.: Induced or triggered? The deadly February 2019 Rongxian-Weiyuan ML 4.9 earthquake in the shale gas field in Sichuan, China, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-12096, https://doi.org/10.5194/egusphere-egu2020-12096, 2020