ERE5.1
Induced/triggered seismicity in geo-energy applications: monitoring, modeling, mitigation, and forecasting

ERE5.1

EDI
Induced/triggered seismicity in geo-energy applications: monitoring, modeling, mitigation, and forecasting
Co-organized by EMRP1/SM6
Convener: Antonio Pio Rinaldi | Co-conveners: Rebecca M. Harrington, Nadine Igonin, Marco Maria Scuderi, Victor Vilarrasa
Presentations
| Mon, 23 May, 13:20–14:40 (CEST), 15:10–18:27 (CEST)
 
Room 0.96/97

Presentations: Mon, 23 May | Room 0.96/97

Chairpersons: Antonio Pio Rinaldi, Federico Ciardo, Silvia De Simone
Hydraulic fracturing and wastewater disposal
13:20–13:27
|
EGU22-8688
|
ECS
|
On-site presentation
Riccardo Minetto, Agnès Helmstetter, and Ben Edwards

In August 2019 an hydraulic fracturing operation was carried out at the PNR-2 well in Preston New Road, UK. Hydraulic fracturing caused abundant seismic activity that culminated with a ML 2.9 event. This event prompted the operator (Cuadrilla Resources Ltd.) to halt any further stimulation of the well. The seismic activity was recorded by a downhole array of 12 sensors located in a monitoring well (PNR-1z). The operator released a seismic catalog created in real time during the fracturing operation. The catalog consists of 55555 events detected and located with a coalescence microseismic mapping method. The catalog also reports moment magnitudes, but no precise information on the method and on the parameters used to estimate them is available. In our study, we attempt to improve the number of detections and the location accuracy of the events by applying template matching and a double-difference relocation method, respectively. We also recalculate moment magnitudes using spectral fitting to look for any inconsistencies in the real-time catalog. Finally, we use the new information to better understand the spatio-temporal evolution of the seismicity and the dynamics that led to the ML 2.9 event.

How to cite: Minetto, R., Helmstetter, A., and Edwards, B.: Analysis of the spatio-temporal evolution of the seismicity induced by hydraulic fracturing operations in Preston New Road, UK, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-8688, https://doi.org/10.5194/egusphere-egu22-8688, 2022.

13:27–13:34
|
EGU22-3271
|
ECS
|
Virtual presentation
Xinyu Tan and Semechah K. Y. Lui

Occurred in the Delaware Basin, western Texas, near the town of Mentone, the Mw 5.0 Mentone earthquake is one of the largest induced earthquakes in the central US. Within 25 km of the epicenter, there are a few deep injection wells to the northwest injecting in the high permeable limestone layer at about 22 km averagely, as well as a lot of shallow wells injecting in the upper high permeable sandstone layer at around 18 km averagely. Between the shallow sandstone and deep limestone layers is a thick shale layer with low permeability, which excludes the possibility of downward percolation of the injected fluid in the shallow injection layer. However, the cumulative injection volume of shallow injection wells is about five times as much as that of deep injection wells. Motivated by this, we investigate whether the shallow injection wells may play a role in triggering the Mw 5.0 Mentone earthquake through the injection-induced coupled poroelastic stress perturbations. We first perform focal mechanism inversion and earthquake relocation with the Cut and Paste (CAP) and hypoDD methods, respectively, to constrain the fault plane on which the Mw 5.0 event occurred. A south-facing fault plane with strike/dip of 81o/52o is successfully fitted. We then calculate the change of the Coulomb failure stress (ΔCFS) caused by the shallow injection wells at the mainshock location based on the linear fully coupled poroelastic stress model. The calculated ΔCFS of shallow injection wells is approximately 20 kPa and it is mostly contributed by the change in coupled poroelastic stress. Based on findings from other studies, this value of ΔCFS is sufficient in reactivating faults that are well aligned with the local stress field. Since we only account for about half of the total injection volume from the shallow wells in the calculation, we also hypothesize that the actual perturbations caused by shallow injection wells via the coupled poroelastic stress change would be more prominent. Our result reveals the vital role of injection-induced coupled poroelastic stress in triggering seismicity, especially in low permeable geologic settings.

How to cite: Tan, X. and Lui, S. K. Y.: The non-negligible contribution of shallow injection wells on the triggering of the Mw 5.0 Mentone earthquake via coupled poroelastic stress perturbations, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-3271, https://doi.org/10.5194/egusphere-egu22-3271, 2022.

13:34–13:41
|
EGU22-11183
|
Presentation form not yet defined
|
Rebecca M. Harrington, Kilian B. Kemna, Marco P. Roth, Ricarda M. Wache, and Yajing Liu

The extensive development over the last decade of low-permeability tight shale formations in the Western Canada Sedimentary Basin (WCSB) using hydraulic fracturing (HF) techniques for oil and gas exploration has been associated with  an increasing number of M3+ earthquakes (e.g., ML 4.5 on 30 November 2018 near Dawson Creek, and a Mw 4.6 on 17 August 2016 near Fort St. John). Avoiding economic losses due to operational shutdowns and mitigating damage caused by ground shaking requires developing quantitative relationships between operational parameters and the rate of fault activation in areas of low historical seismicity rates such as the WCSB.

Here we present the first results of a detailed study of the relationship between earthquake occurrence and operational parameters using dense seismic array and the British Columbia Oil and Gas Commission operational database to quantitatively assess the relative influence of operational parameters and geological conditions on earthquake generation. We first enhance a local, automatically generated seismic catalog of > 8000 events in the Kiskatinaw (Montney Formation) in the time period between July 2017 -  December 2020 area using a multi-station matched-filter approach.  We then use a machine learning picker as an independent detection algorithm for the same time period and retain events with the best initial locations detected by both the matched-filter and machine-learning approaches. The combined approach leads to  > 30,000 additional earthquakes, which we relocate using a double-difference technique, lowering the magnitude of completeness Mc from ~1.3 to ~0.2.

As shown by several previous studies, while most earthquakes show a clear spatio-temporal correlation with HF operations, the majority of HF operations are not associated with felt earthquakes (e.g., M3+). To investigate the correlation between individual HF stage stimulation and earthquake occurrence, we correlate operational and geological characteristics with > 13000 HF stages. Geological data consists of the target formation for injection, which consists of either the Lower or Upper Montney Formations for the majority of stages. We then use a gradient-boosted decision tree machine learning algorithm combined with an approach to explain the model predictions to assess whether a specific stage is seismogenic. The decision-tree-algorithm allows us to estimate the importance of each injection parameter for the generation of seismicity. First results show that the target formation is the most influential parameter, where the Lower Montney Formation is more prone to higher rates of seismicity. In addition, the total pumped fluid volume and the maximum treating pressure are the important injection parameters that are positively correlated with seismicity. In contrast, the average injection rate and breakdown pressure may be relatively less influencial. We will present the results for specific stages and discuss the importance of their injection parameters in relation to seismicity. Our results could help to determine why only some HF wells are seismogenic.

How to cite: Harrington, R. M., Kemna, K. B., Roth, M. P., Wache, R. M., and Liu, Y.: Detailing the relationship between hydraulic fracturing parameters and induced seismicity using small-magnitude earthquakes, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-11183, https://doi.org/10.5194/egusphere-egu22-11183, 2022.

13:41–13:48
|
EGU22-9373
|
ECS
|
Presentation form not yet defined
Marco Pascal Roth, Kilian B Kemna, Alessandro Verdecchia, Ricarda M Wache, Andres F Pena Castro, Rebecca M Harrington, and Yajing Liu

The Western Canada Sedimentary Basin (WCSB) has experienced an increase in hydraulic fracturing (HF) operations in the last decade, accompanied by an increase in the number of felt earthquakes, including a Mw 4.6 on 17 August 2015 near Fort St. John and a ML 4.5 (Mw 4.2) on 30 November 2018 near Dawson Creek. While only a small percentage of HF operations induce seismicity, the majority of moderate-sized earthquakes occur in close spatial proximity to HF wells and temporal proximity to individual HF injection stages within the tight shale play of the Montney Formation. Whereas statistical analysis of an enhanced seismicity catalog suggests that the majority of seismicity occurs following HF operations in the relatively older and deeper compartments of the Montney Formation (Lower Montney; LM) and a low number of events are associated with the relatively younger and shallower layers (Upper Montney; UM), the detailed association and triggering mechanism(s) remains unclear.

In this study, we investigate induced earthquake source parameter variations resulting from spatial and/or temporal alteration of injection parameters, including injection time, depth, and volume, at three well pads operating between 2018 and 2020 in the Kiskatinaw area. We use dense local station coverage to create an enhanced seismicity catalog with double-difference relative hypocenter relocations to highlight potential fault orientations, confirmed by focal mechanism solutions. We estimate static stress drop values at the individual well pads and their variation over time as well as variation with the choice of empirical Greens function. We also investigate the temporal changes of the VP/VS-ratio in localized areas following HF operations as a proxy for increased fracture density and/or compliance.

The case study at three specific sites targeting both the UM and LM layers investigates the relative influence of a number of factors on the spatial and temporal distribution of source properties. Factors include the scale of HF injection parameters, the target formation layer, and site-specific factors, such as localized fluid accumulation. Preliminary results show that injection in the UM generally leads to significantly fewer earthquakes than injection in the LM, and that lateral variations in compartment properties may significantly influence the seismic response. Moreover, we investigate if repetitive injection at the same wellhead may repeatedly (re)activate sets of faults/fractures and lead to increased hydraulic connectivity between the target sedimentary layers and deeper, pre-existing basement faults. An increase in connectivity would imply an increased potential for triggering large mainshocks.

How to cite: Roth, M. P., Kemna, K. B., Verdecchia, A., Wache, R. M., Pena Castro, A. F., Harrington, R. M., and Liu, Y.: Variation in induced seismicity productivity by alteration of injection parameters: a comparative case study at three hydraulic fracturing wells in the Kiskatinaw area, British Columbia, Canada, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-9373, https://doi.org/10.5194/egusphere-egu22-9373, 2022.

Laboratory and In-Situ experiments
13:48–13:55
|
EGU22-4177
|
ECS
|
On-site presentation
Milad Naderloo, Aukje Veltmeijer, and Auke Barnhoorn

Over the last few decades, it has become apparent that different human activities in the subsurface, such as water waste injection, hydraulic fracturing, and geothermal energy production can lead to induced seismicity. Understanding the effects of fluid injection-related parameters on seismic response or evolution of it is essential for finding a method to manage and minimize the induced seismicity risk. Experimental and numerical studies indicate that varying injection patterns and rates can be used to effect and/or mitigate seismicity. However, most of the studies are for intact rock medium, and the mechanism of injection-induced seismicity of faulted rock medium is not clear yet. In this study, we performed fault reactivation experiments on faulted (saw-cut) Red Pfaelzer sandstones to provide new insight into the effect of stress/pressure cycling and rate on fault slip behavior and seismicity evolution. The saw-cut samples were subjected to two different reactivation mechanisms: 1) stress-driven and 2) injection-driven fault reactivation. Three different reactivation scenarios were performed during the stress-driven fault reactivation experiments: continuous sliding, cyclic sliding, and under-threshold cycling sliding. Ten AE transducers were used to detect microseismicity during the fault reactivation experiments, and consequently, different microseismic parameters, such as frequency-magnitude distribution (b-value), AE energy, and AE rate were estimated. Stress-driven fault reactivation experiments showed that (i) a below-threshold cycling scenario prevents seismicity and pure shear slip; however if the shear stress exceeds the previous maximum shear stress, seismicity risk increases drastically in terms of b-value, maximum AE energy, and magnitude. (ii) Compared to continuous sliding, cyclic sliding triggers less seismicity in terms of total b-value and large AE events due to the uniform reduction in roughness and asperity on the fault plane. (iii) By increasing the number of cycles, in general, the number of generated events and AE energy per cycle is reduced. Nevertheless, there is a risk of generating large AE events during the first cycles. In addition, results from the injection-driven fault reactivation experiments demonstrated that high injection rate results in higher peak slip velocity. Compared to the stepwise injection pattern, the cyclic recursive injection scenario showed higher peak slip velocity, due to the high hydraulic energy budget and fault compaction. A proper injection strategy needs to consider various factors, such as fault drainage, critical shear stress, injection rate, and injection pattern (frequency and amplitude). Our results demonstrate that selecting proper stress/pressure amplitude, and pressurization rate for the injection design strategy can help to reduce seismicity risk.  

How to cite: Naderloo, M., Veltmeijer, A., and Barnhoorn, A.: Fault reactivation process in the laboratory: The role of stress cycling and pressurization rate, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-4177, https://doi.org/10.5194/egusphere-egu22-4177, 2022.

13:55–14:02
|
EGU22-4117
|
ECS
|
On-site presentation
Aukje Veltmeijer, Milad Naderloo, and Auke Barnhoorn

With human activities in the subsurface increasing, so does the risk of induced seismicity. For mitigation of the seismic hazard and limiting the risk, monitoring and forecasting are essential. A laboratory study was performed to find precursors to fault failure. In this study, Red Pfaelzer sandstones samples were used, which are analog to the Groningen gas reservoir sandstones. A saw-cut fault was cut at 35 degrees, and the samples were saturated. Fault slip was induced by loading the sample at a constant strain rate, and simultaneously active acoustic transmission measurements were performed. Every 3 seconds 512 S-waves were sent, recorded, stacked to reduce the signal-to-noise ratio, and analyzed. The direct seismic wave velocity, coda wave velocity, and transmissivity were monitored before and during the reactivation of the faulted samples. Different loading patterns and confining pressures were investigated in combination with active acoustic monitoring. Velocity and amplitude variations were observed before the induced fault slip and can be used as precursory signals. Two methods to determine changing velocities were used. Direct S-wave velocities are compared to velocity change obtained by coda wave interferometry. Both analyses gave similar precursory signals, showing a clear change in slope, from increase to decreasing velocities and amplitudes prior to fault reactivation. Fault reactivation is preceded by fault creep and the destroying of some of the asperities on the fault plane, causing the seismic wave amplitude and velocity to decrease. Combining all precursors, the onset of fault slip can be determined and therefore upcoming slip can be forecasted in a laboratory setting. Our results show precursory changes in seismic properties under different loading situations and show a clear variation to the onset of fault reactivation. These results show the potential of continuous acoustic monitoring for detection and forecasting seismicity and help the mitigation of earthquakes.

How to cite: Veltmeijer, A., Naderloo, M., and Barnhoorn, A.: Acoustic Precursors to Laboratory Induced Fault Slip and Failure, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-4117, https://doi.org/10.5194/egusphere-egu22-4117, 2022.

14:02–14:09
|
EGU22-11535
|
On-site presentation
Francois Passelegue and Pierre Dublanchet

Fluid injections are known to induce earthquakes in the upper crust. Recent studies have highlighted that fluid injections can contribute to the nucleation of instabilities close to or far from the injection site due to stress transfer induced by poroelastic processes. In addition, recent studies have suggested the maximum magnitude earthquake is expected to be a function of the volume injected. However, the development of the slip front related to the fluid pressure front, as well as its implications on the induced seismic sequence in time and space, remain poorly constrained in the laboratory and in natural fault systems.

Here, we investigated the influence of the initial normal stress (i.e., the permeability of the fault plane) and of the injection rate on the development of both the fluid pressure front and associated slip front during the nucleation stage of laboratory fluid-induced earthquakes. Experiments were conducted on saw cut samples of andesite, presenting a negligible bulk permeability compared to the fault plane one. Strain gauges were glued all around the fault surface to track, (i) the strain transfer associated with slip front propagation during injection and the rupture velocity during dynamic rupture propagation. The dynamics of the fluid pressure front was inverted from pore pressure measurements located at both edges of the fault. The evolution of the slip distribution due to the change in fluid pressure around the injection site was inverted from strain gauge measurements, assuming a 3D modelling of the sample specimen using the Finite Element Method. Our preliminary results show that the initial stress acting on the fault controls the development of the slip front during the nucleation of the instability. In addition, the larger the injection rate, and the faster the propagation of the slip front compared to the fluid pressure front. Finally, the scaling between the volume of fluid injected and the associated nucleation moment differs from the one relating the volume injected to the seismic moment.

 

How to cite: Passelegue, F. and Dublanchet, P.: Development of slow slip front during the nucleation of laboratory fluid-induced earthquakes, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-11535, https://doi.org/10.5194/egusphere-egu22-11535, 2022.

14:09–14:19
|
EGU22-4204
|
ECS
|
solicited
|
On-site presentation
Yinlin Ji, Lei Wang, Hannes Hofmann, Grzegorz Kwiatek, and Georg Dresen

In this study, we conducted injection-driven shear tests on a sawcut fault in granite samples using a triaxial deformation apparatus. The granite samples were drilled from Odenwald basement rocks in Germany. The sawcut fault, inclined 30° to the sample axis, was ground using sandpaper with a particle size of 201 µm. Two boreholes (nominal diameter 1.8 mm) were drilled near the short edge of each sample half to allow direct fluid access to the fault surface. Eight strain gauges, and eight pairs of acoustic emission (AE) sensors attached on the sample surface were used to monitor the deformation, local strain and AE events. 

During the experiments, we first measured the peak shear strength of the faulted sample by advancing the axial piston at a constant rate of 1 µm/s under 36 MPa confining pressure and 1 MPa pore pressure. We then adjusted the shear stress to be 90% of the peak shear strength. Subsequently, the piston was fixed, and the first injection-driven shear test was initiated by injecting distilled water from the bottom borehole at a rate of 0.2 mL/min. We observed three full cycles of fast slip events until the injection pressure was increased up to approximately 18 MPa. We then reduced the pore pressure to the initial 1 MPa and the axial force was removed, followed by the second injection-driven shear test conducted at a higher injection rate of 0.8 mL/min using the same procedure as in the first test. We also observed three episodes of fast slip events until the injection pressure was increased to about 20 MPa. Fluid pressures were monitored continuously at the top and bottom boreholes. We employed a COMSOL model to obtain the time-dependent fluid pressure distribution along the sawcut fault during fluid injection.

For slow fluid injection, we find that the fault surface near the center experiences slight normal dilation and gradual shear stress release prior to the fast slip event. In contrast, for high-rate fluid injection, the same fault patch exhibits normal compaction and shear stress increase preceding fast slip. In both cases, significant normal dilation and abrupt shear stress drops were observed near the fault center during fast slip events. The distinct evolution of local fault deformation and stress are likely attributed to the distribution of slow slipping patches, as signified by the fluid pressure distribution and Mohr-Coulomb failure envelope. At slow injection rate, slow precursory slip may have occurred on the entire fault, initiating a fast slip event. In contrast, at higher rates, slow slip may have been localized around the injection port, resulting in local stress concentration beyond the slow slipping patch. Our results demonstrate that the evolution of local fault deformation and stress can be diverse in different fault patches, depending on the relative location to the fluid pressurized zone and the resulting slow slipping patch. This suggests that the strongly heterogeneous fault deformation should be considered when analyzing the precursors to injection-induced fault reactivation.

How to cite: Ji, Y., Wang, L., Hofmann, H., Kwiatek, G., and Dresen, G.: Injection-rate control on deformation and stress of an experimental fault in granite, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-4204, https://doi.org/10.5194/egusphere-egu22-4204, 2022.

14:19–14:26
|
EGU22-9233
|
Virtual presentation
|
Virginie Durand, Martina Rosskopf, Katrin Plenkers, Anne Obermann, Miriam Schwarz, Linus Villiger, Men-Andrin Meier, Hansruedi Maurer, Domenico Giardini, and Stefan Wiemer and the Bedretto Team

The Bedretto Underground Laboratory for Geoenergies and Geosciences (BULGG) is a multidisciplinary laboratory on the hundred meter scale run by ETH Zurich. It is located in the Swiss Alps, in the middle of a 5.2km long horizontal tunnel, 1.0km below the surface. 
Seven 250-300m long boreholes have been equipped with different instruments: Acoustic Emission Sensors, Accelerometers, Fiber Optics (allowing simultaneous DTS, DSS and DAS measurements), Strainmeters and Pore Pressure Sensors. The variety of the instrumentation allows a multidisciplinary analysis of the response of the rock volume to fluid injections. The fluid injections are realized through a 400m injection borehole located in the center of the instrument network. It is divided into 14 intervals, allowing us to make injections at different depths.
We will first present the methods used to generate a pico-seismic catalog with precise locations and a magnitude of completeness as low as -5, and the associated challenges. Then, we show a preliminary analysis of the spatio-temporal evolution of the pico-seismicity generated by different injection protocols. We interpret the evolution of the seismicity in comparison with the injection parameters (i.e., injection pressure and rate) and the stimulated intervals.

How to cite: Durand, V., Rosskopf, M., Plenkers, K., Obermann, A., Schwarz, M., Villiger, L., Meier, M.-A., Maurer, H., Giardini, D., and Wiemer, S. and the Bedretto Team: Analysis of the pico-seismic response of a fractured rock volume to fluid injections in the Bedretto Underground Laboratory, Switzerland, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-9233, https://doi.org/10.5194/egusphere-egu22-9233, 2022.

Reservoir Induced Seismicity
14:26–14:33
|
EGU22-7900
|
ECS
|
Presentation form not yet defined
|
Izabela Nowaczyńska and Grzegorz Lizurek

The Song Tranh 2 hydropower construction is located in the Quang Nam province (central Vietnam), it has a reservoir volume of 740 million cubic meters of water and a dam height of 96 m. The reservoir was filled to capacity for the first time in February 2011. The seismicity in the vicinity of reservoir is example of reservoir triggered seismicity (RTS).  The natural seismic activity of the Song Tranh 2 reservoir is very low. After the reservoir was filled, the seismic activity increased, and the number and frequency of the tremors also changed as the water level changed. Water level changes are accelerating the tectonic process leading the critically stressed faults to slip. Data suggest that reservoir exploitation stress field changes as triggering origin of this seismicity. The stress inversion method was used to check if there were any seasonal trends. The inverted stress tensor and, in particular, the stress ratio, which is very sensitive to data quality and scope and difficult to accurately retrieve, can be influenced by porous pressure changes. Has been checked, how the average annual seismic activity is related to the change of the water level and if it implies the orientation of the principal stress during high and low water levels in the reservoir.  The pore pressure changes and the stress ratio changes were also estimated in relation to the high and low water level periods. Coulomb stress transfer is a seismic-related geological process of stress changes to surrounding material caused by local discrete deformation events.Importantly, Coulomb stress changes have been applied to earthquake-forecasting models that have been used to assess potential hazards related to earthquake activity. It is also often assumed that changes in pore fluid pressure induced by changes in stress are proportional to the normal stress change across the fault plane. Coulomb stress changes was also calculated for low and high water period.

How to cite: Nowaczyńska, I. and Lizurek, G.: Seasonal stress inversion trends and Coulomb stress changes of RTS in Song Tranh2 reservoir, Vietnam, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-7900, https://doi.org/10.5194/egusphere-egu22-7900, 2022.

14:33–14:40
|
EGU22-9575
|
On-site presentation
|
Grzegorz Lizurek, Konstantinos Leptokaropoulos, Monika Staszek, Izabela Nowaczyńska, and Anna Tymińska

Water reservoirs play important role in energy production in Vietnam. Numerous dams were designed and built for hydropower plants and water storage during the wet season and its release during dry season. They were built in a different tectonic settings. We present our experience of several years of monitoring and research on two sites: first, tectonic active area of Lai Chau (North Vietnam) and relatively stable area of Song Tranh in Central Vietnam. We observed different seismicity patterns in this areas. Area of active tectonics in Lai Chau was less active in terms of reservoir triggering, while almost aseismic area of Song Tranh was highly active after reservoir impoundment. We proved, that this activity was related with seasonal water level changes in reservoir. Moreover, low water period during service works was proved to be more active and with significantly higher seismic hazard than during initial production regime and after the refilling. It suggests that decrease of water level and following pore-pressure change destabilize minor faults being closer to failure, than main faults in the area. We also found multiplet events triggered on minor normal faults in shallow depth despite the strike-slip regime of regional tectonic stress field. On the other hand in the active area of Lai Chau we observed triggering both on existing active strike-slip faults and minor normal fault discontinuities. However, the difference between seismic activity parameters before and after impoundment except spatial distribution directly after first filling didn’t differ substantially. We can conclude, that in stable tectonic setting triggering effect is clear and related with pore-pressure changes caused by reservoir water level fluctuations, which is main seismogenic factor. On the other hand in active seismic area reservoir water level fluctuation seems to be too small to significantly influence seismic activity in the long term.

This work was partially supported by the research project no. 2017/27/B/ST10/01267, funded by the National Science Centre, Poland under the agreement no. UMO-2017/27/B/ST10/01267 (GL and IN) and partially supported by the research project no. 2021/41/B/ST10/02618, funded by the National Science Centre, Poland under the agreement no. UMO-2021/41/B/ST10/02618 (GL and AT) and partially by National Statutory Activity of the Ministry of Education and Science of Poland No 3841/E-41/S/2022 (MS)

How to cite: Lizurek, G., Leptokaropoulos, K., Staszek, M., Nowaczyńska, I., and Tymińska, A.: Reservoir triggered seismicity in tectonically stable and seismically active areas of Vietnam, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-9575, https://doi.org/10.5194/egusphere-egu22-9575, 2022.

Coffee break
Chairpersons: Antonio Pio Rinaldi, Federico Ciardo, Silvia De Simone
Geothermal
15:10–15:17
|
EGU22-9826
|
ECS
|
On-site presentation
|
Dilshan Bandara, Thomas Heinze, Jeroen Smit, and Stefan Wohnlich

In the context of switching power generation towards renewable energy sources, the geothermal exploration of low enthalpy systems has gained interest also in regions with little to no recent tectonic or magmatic activity such as Sri Lanka. Sri Lanka has 9 low enthalpy systems with yet unknown heat generating mechanisms besides several existing hypotheses. Recent studies of such kind of low enthalpy geothermal systems hypothesize that fault network and recharge elevations are the main factors controlling the origin of the hot springs.

We studied the fault network, shear zones, and regional fracture networks to understand the heat flow causing the Sri Lankan hot springs. Remote sensing and geophysical methods were used to identify and analyze lineaments. We find that (1) The peak circulating temperatures of deeply circulating meteoric water depend on the elevation of the recharge zone for the corresponding hot spring. (2) Hot springs are formed in a terrain with a long fault / shear zone (starting from the highlands) when cross cuts with a regional fracture network occur in or near to the hot spring fields. (3) Highest number of hot springs in the country relates with the fault network that crosses the Mahaweli shear zone at the boundary of the two geological complexes Highland and Vijayan.

We conclude that the fault network that crosses both the central highlands and the Vijayan Complex plays a major role in the heating of deep percolating water, as it transports the water over more than hundred kilometers distance from the recharge zones to the hot springs. 

How to cite: Bandara, D., Heinze, T., Smit, J., and Wohnlich, S.: Structurally controlled regional groundwater circulation: Origin of geothermal springs in Sri Lanka, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-9826, https://doi.org/10.5194/egusphere-egu22-9826, 2022.

15:17–15:24
|
EGU22-5332
|
ECS
|
Virtual presentation
|
Iman Rahimzadeh Kivi, Estanislao Pujades, Jonny Rutqvist, and Victor Vilarrasa

The widespread development of geothermal energy is deemed to accelerate the transition to a low-carbon future. Hot Sedimentary Aquifers (HSA) provide cost-effective and non-intermittent geothermal resources. However, HSA development has reportedly been associated with seismic events, harming the public perception of exploiting these resources. This work digs deeper into thermo-hydro-mechanical (THM) mechanisms raised by water circulation in a HSA and their control on fault reactivation. We numerically simulate the problem by a 2D plane-strain model. The model consists of a porous and permeable hot aquifer sandwiched between the tight seal and base rocks and laterally bounded by two normal faults, representative of extensional tectonic environments. The horizontal injection-production well pairs are spaced 500 m apart at the middle of the aquifer, and the faults are located on each side of the doublet at a distance of 1 km. We consider two scenarios: low-permeability faults, mimicking a compartmentalized reservoir, and high-permeability faults, across which fluid flow takes place with further ease. We show that the fault permeability governs the hydraulic response of the reservoir. While the pore pressure slightly increases around the injector and decreases around the producer for the case of high-permeability faults, the compartmentalized reservoir experiences a global pore pressure decline. The latter is supported by the fact that the injected cold water is denser than the extracted hot water and occupies less space in the pore system. As soon as the thermal breakthrough occurs, which is after 12 years in the current setting, a more uniform temperature distribution across the doublet is established and the pressure begins to increase in the vicinity of the injector. Provided the high permeability of the reservoir rock, pore pressure and poroelastic stress perturbations impose rapid but minor effects on the fault stability. On the contrary, the cooling front formed around the injector lags much behind the pore pressure front toward the fault. The reservoir cooling contracts the rock and triggers stress reductions. Thermal stresses are transmitted much ahead of the cooled region and destabilize the fault located on the injection side. The fault begins to slip after 18 and 21 years of circulation for the high- and low-permeability scenarios, respectively. The reservoir pressure decrease in the latter case, attenuating the fault slip tendency, feeds into the observed difference in reactivation timings. Although thermal stresses initiate the slip, the static stress transfer jointly contributes to rupture nucleation along the fault. Interestingly, the slip-induced shear stress release, tied to a slip weakening frictional behavior, slows down elastic energy build-up on the other fault closer to the production well and impedes its reactivation. Our findings on the prolonged but dominant role of thermal stresses on the reactivation of distant faults have direct implications for safe and long-term production from geothermal systems.

How to cite: Rahimzadeh Kivi, I., Pujades, E., Rutqvist, J., and Vilarrasa, V.: Thermal stressing is likely to reactivate distant faults in hot sedimentary aquifers, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-5332, https://doi.org/10.5194/egusphere-egu22-5332, 2022.

15:24–15:31
|
EGU22-7670
|
On-site presentation
Multi-hazard risk analysis based on Bayesian networks for geothermal resources exploitation
(withdrawn)
Konstantinos Trevlopoulos and Fatemeh Jalayer
15:31–15:38
|
EGU22-2160
|
On-site presentation
Auregan Boyet, Silvia De Simone, and Vìctor Vilarrasa

Felt induced seismicity compromises the public perception on the deployment of geothermal power-plants in urban areas. Large induced earthquakes have led to the shutdown of Enhanced Geothermal Systems (EGS), such as Basel (Switzerland) and Pohang (Republic of Korea). In the majority of induced seismicity cases in EGS, the largest events occur after shut-in. Different mechanisms can trigger induced seismicity. Pore pressure diffusion is established as the most common triggering mechanism. It reduces the effective normal stress acting across pre-existing fault surfaces, weakening the shear resistance and allowing slip of faults. However, this is not the only triggering mechanisms and it cannot explain the large magnitude of post-injection induces seismicity. Additional influencing processes are poromechanical elastic stressing, shear stress transfer and local tectonic settings. Considering theses mechanisms simultaneously can provide a better understanding of the causes of post-injection seismicity and could allow to develop strategies to mitigate the occurrence of earthquakes with high magnitude. To explain these processes, we investigate the induced seismicity that led to the closure of the Basel EGS project. We set-up a hydro-mechanical finite element numerical model which contains faults corresponding with the clusters of induced events at Basel. We study the reactivation of these pre-existing fractures using a viscoplastic model. We are able to identify the process combinations bringing faults to failure. During injection, faults fail due to pore pressure diffusion in the vicinity of the well, and due to poroelastic stressing further in the reservoir. After the injection shut in, poroelastic stressing and shear stress transfer trigger seismicity, being the most relevant triggering mechanisms of post-injection induced seismicity.

How to cite: Boyet, A., De Simone, S., and Vilarrasa, V.: Identification of The Processes Triggering Induced Seismicity at the Enhanced Geothermal System of Basel (Switzerland), EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-2160, https://doi.org/10.5194/egusphere-egu22-2160, 2022.

15:38–15:45
|
EGU22-7969
|
ECS
|
On-site presentation
Annukka Rintamäki, Sebastian Heimann, Torsten Dahm, and Gregor Hillers

An experimental ~6 km deep enhanced geothermal system in Otaniemi, in the Helsinki capital region, southern Finland, was stimulated in 2018 and 2020. During the two stimulations that lasted seven and three weeks, respectively, signals of the induced earthquakes with a maximum local magnitude of 1.8 were recorded with dense and diverse seismic networks. The intraplate southern Finland setting of the experiment yields an intriguing opportunity to study earthquake and rock failure processes in the precambrian Fennoscandian Shield where the level of natural seismicity is comparatively low. The high confining pressure of 180 MPa at 6 km depth defines the key characteristics of the stress field, together with the previously estimated North-110-degrees-East direction of the maximum horizontal stress. The competent crystalline bedrock has very low attenuation, and yields high signal-to-noise ratio seismograms even at relatively high frequencies. We study the source mechanisms of ~250 induced earthquakes with Mw > 0.5. We perform probabilistic full moment tensor analysis with the Grond package of the software suite Pyrocko. We use data sets from the 2018 and 2020 stimulation experiments. Both experiments were monitored with more than 100 three-component surface stations operated by the Institute of Seismology, University of Helsinki, and 12 three-component borehole stations maintained by the St1 developer company installed at around 300 m depth. The diverse network elements help to evaluate the consistency of the results. We first present results of a detailed analysis of a small event subset characterized by the best data quality and solutions to assess the robustness of the different tensor components to different processing choices. This includes a comparison of surface and borehole sensor data. This allows us to conclude that the majority of the analysed earthquakes have a dominant reverse faulting mechanism and a small subset of events has strike slip mechanisms, which is compatible with solutions reported by the developer group. The predominant fault plane orientations are in agreement with the ambient stress conditions that also seem to control the thrust mechanism. Based on the best quality solutions we discuss the significance of the obtained non-double couple moment tensor components to assess if significant opening or closing elements in the induced earthquake source reflect genuine physical processes or spurious effects associated with imperfect resolution.

How to cite: Rintamäki, A., Heimann, S., Dahm, T., and Hillers, G.: Source mechanisms of earthquakes induced by the 2018 and 2020 geothermal stimulations in Espoo/Helsinki, southern Finland, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-7969, https://doi.org/10.5194/egusphere-egu22-7969, 2022.

15:45–15:52
|
EGU22-7486
|
ECS
|
On-site presentation
Amir Sadeghi-Bagherabadi, Tom Eulenfeld, Tommi A.T. Vuorinen, Annukka E. Rintamäki, and Gregor Hillers

In 2018 and 2020, two weeks-long geothermal reservoir stimulations were performed some 6 km below the Helsinki capital area, Finland. The seismic activity was recorded by a set of surface broadband sensors and 100 geophones installed by the Institute of Seismology, University of Helsinki, as well as Finnish National Seismic Network stations. The local magnitudes (ML) of the recorded earthquakes are estimated using a Finnish local magnitude scale and the local magnitude of the largest induced event was 1.8. We apply three different approaches for estimation of moment magnitudes (MW) to a data base of ~400 induced seismic events from the 2018 stimulation to explore the variability and sensitivity of the magnitude estimates. This is important for real-time monitoring and decision making when the induced event magnitudes approach the pre-defined magnitude limit, and to assess which trends can be robustly associated to earthquake source physics. (1) We employ a time-domain calculation of source parameters based on the application of Parseval's theorem to the integrals of the squared spectral displacement and velocity for the horizontal S-wave trains. The time window between the S-wave arrival time and twice the length of the S-wave travel time is considered for the S-wave train isolation. (2) We obtain moment magnitude estimates from an inversion of 50 s long three-component envelopes based on radiative transfer. (3) We apply a moment tensor inversion to 0.71 s long P and 0.81 s long S-wave signals. We fit a linear ML-MW conversion model to the values obtained from the different approaches. Considering the available local magnitude range between –0.5 and 1.8, a comparison of the linear conversion models shows that the moment magnitudes form the envelope inversion are systematically larger by ~0.2 units compared to those obtained from the moment tensor inversion. While the moment magnitudes determined by the time-domain calculation consistently exceed those of the envelope inversion for small local magnitudes (by ~0.2 units), they tend to yield similar estimates towards the larger local magnitudes. Other source parameter systematics include that the smallest seismic moment is obtained with the moment tensor inversion, and the largest with the time-domain equivalent of the spectral integrals. An initial extension of the analysis to 2020 data yields ML-MW as well as corner frequency-MW scaling relations that are, interestingly, different compared to the 2018 results; we will present updated results that inform about the reliability of these trends.

How to cite: Sadeghi-Bagherabadi, A., Eulenfeld, T., Vuorinen, T. A. T., Rintamäki, A. E., and Hillers, G.: Magnitude estimates of earthquakes induced by the geothermal stimulations in Espoo/Helsinki, southern Finland: a comparison of different approaches, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-7486, https://doi.org/10.5194/egusphere-egu22-7486, 2022.

15:52–15:59
|
EGU22-10183
|
ECS
|
On-site presentation
|
Lukas Krenz, Sebastian Wolf, Alice-Agnes Gabriel, Gregor Hillers, and Michael Bader

With this contribution, we expand the discussion of effects that earthquakes induced by geo-energy projects can have on local communities, and that should probably be considered in future legislation or permitting processes. Inspired by consistent reports of felt and heard disturbances associated with the weeks-long stimulation of a 6-km-deep geothermal system in 2018 below the Otaniemi district of Espoo, Helsinki, we conduct numerical simulations of wave propagation in the solid earth and the atmosphere to assess the sensitivity of the ground shaking and audible noise patterns to various parameters. We explore the effects of three different local velocity models, realistic topography, variations of the source mechanism, and earthquake size on the loudness of the synthetic waves at frequencies up to 20 Hz, therefore reaching the lower limit of human sound sensitivity. We discuss the results of 18 elastic-acoustic coupled scenario simulations conducted on the Mahti high-performance computing infrastructure of the Finnish IT Center for Science CSC using the SeisSol wavefield solver. The computationally challenging simulations target the Otaniemi case study, i.e., we discretize a 12 km x 12 km x 15 km domain with a 2 km thick air layer over the solid earth domain. The earthquake point source is located at the 6.5 km deep location of the largest M1.8 event induced by the stimulation. In the target central area, we use a mesh with element lengths of about 14 m in the air and 97 m in the solid earth. Inside each element, we approximate the solution by a fifth-degree polynomial, by which we achieve a resolution of roughly 2.3 m in the air and 16 m in the earth. We develop an interactive visualization to facilitate instant access to the results governed by the different parameter combinations, where the synthetics are shown on top of a map of the Helsinki metropolitan region. This tool facilitates “what-if” analyses by quickly comparing the effects of fault orientation, source mechanism, and the velocity model. This supports effective communication of physics-based nuisance analysis to decision-makers and stakeholders, not only in environments such as the case study where there is little experience with natural earthquake phenomena. Together, these results resolve for the first time synthetic nuisance sound patterns at the 50 – 100 m scale in a densely populated capital region. The study highlights the mostly disregarded spatially variable audible effects that can negatively impact the public attitude towards geothermal stimulations, even if the ground shaking limits are safe, and it provides first estimates of the resources needed for comprehensive scenarios for future stimulation projects.

How to cite: Krenz, L., Wolf, S., Gabriel, A.-A., Hillers, G., and Bader, M.: The variability of seismo-acoustic nuisance patterns: a case study from the Helsinki geothermal stimulation, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-10183, https://doi.org/10.5194/egusphere-egu22-10183, 2022.

15:59–16:09
|
EGU22-4703
|
ECS
|
solicited
|
On-site presentation
Federica Lanza and Stefan Wiemer and the DEEP team

The Swiss Energy Strategy 2050 anticipates that by 2050 up to ~7% of the future energy production will come from deep geothermal energy. Likewise, many other countries worldwide are investigating the potential of harnessing deep geothermal energy as a renewable solution. However, seismic risk reduction and reservoir efficiency is the current major coupled problem faced by Enhanced Geothermal System (EGS) reservoirs. Balancing risk and economic output is a key requirement in all EGS projects. The DEEP (Innovation for De-risking Enhanced geothermal Energy Projects) project is an international collaboration whose research-goal is to establish a full-scale protocol for real-time monitoring and risk analysis of potential seismicity triggered by EGS operations. To this end, the project will employ innovative seismic sensors, improved event-cataloguing techniques, fully probabilistic data-driven seismicity forecasts, and loss assessment strategies. In DEEP we plan to apply the so-called Adaptive Traffic Light System (ATLS) where forecasts are continuously updated with real-time data-feeds, providing an integrated and dynamic assessment of the seismic risk to the operators. Field test sites include the Frontier Observatory for Research in Geothermal Energy (FORGE) in Utah (USA), as well as at EGS sites in Germany and France. Parallel to the technology development, the project aims also at defining the next-generation good-practice guidelines and risk assessment procedures in order to reduce commercial costs and enhance the safety of future projects. Here, we will present an overview of the DEEP project to provide a framework for other DEEP presentations. We will also showcase a selection of results from new event detection and location algorithms based on machine learning and using Distributed Acoustic Sensing (DAS), as well as the results from a pilot test of the ATLS workflow for seismicity forecast models for the upcoming FORGE stimulation strategy.

How to cite: Lanza, F. and Wiemer, S. and the DEEP team: The DEEP Project: Innovation for De-Risking Enhanced Geothermal Energy Projects, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-4703, https://doi.org/10.5194/egusphere-egu22-4703, 2022.

Gas Storage and Production
16:09–16:16
|
EGU22-11538
|
Virtual presentation
|
Victor Vilarrasa, Silvia De Simone, Jesus Carrera, and Antonio Villaseñor

Cushion gas injection at the Underground Gas Storage (UGS) project of Castor, Spain, induced hundreds of events, including thirteen with magnitude higher than 3.5 that were felt by the local population and led to project cancellation. The sequence of felt events comprises the three largest earthquakes (M4.08, M4.01 and M3.97) ever induced by any of the more than 640 UGS facilities around the world. The largest earthquakes occurred 20 days after shut-in, when pore pressure buildup had already dissipated. The induced earthquakes nucleated at depths ranging from 4 to 10 km, significantly deeper than the storage formation, which is located at 1.7 km depth. These features of the induced seismicity disregard pore pressure buildup as the triggering mechanism. Our analyses show that seismicity was induced by gas injection, which reactivated the critically stressed Amposta fault. The Amposta fault, which bounds the storage formation, is a mature fault with very low permeability as a result of clay accumulation into its core resulting from its 1,000-m offset. Pore pressure buildup, but specially buoyancy of the gas, which continued to act after shut-in, destabilized the Amposta fault aseismically. The accumulation of aseismic slip caused stress transfer, destabilizing a deep critically stressed fault. Subsequently, shear slip stress transfer combined with slip-driven pore pressure changes, induced the sequence of felt earthquakes. We conclude that the induced earthquakes at Castor could have been avoided because fault stability analysis reveals the high risk of inducing seismicity.

 

Reference

Vilarrasa, V., De Simone, S., Carrera, J. and Villaseñor, A., 2021. Unravelling the causes of the seismicity induced by underground gas storage at Castor, Spain. Geophysical Research Letters, 48, e2020GL092038

How to cite: Vilarrasa, V., De Simone, S., Carrera, J., and Villaseñor, A.: Triggering mechanisms of the induced seismicity at the Underground Gas Storage of Castor, Spain, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-11538, https://doi.org/10.5194/egusphere-egu22-11538, 2022.

16:16–16:23
|
EGU22-8224
|
On-site presentation
Will Reis, Marcella Cilia, Neil Watkiss, Sally Mohr, Rui Barbara, and Phil Hill

Carbon Capture and Storage (CCS) sites require microseismic monitoring before, during and after operations to ensure safety of operational personnel and the wider public.

The high dynamic range and low self-noise of broadband seismometers allows for the detection of low magnitude microseismic events which fall below the threshold of less sensitive geophones. Higher long-period sensitivity also allows the full source spectra of earthquakes to be accurately measured, resulting in more accurate magnitude estimations which improve the integrity of any microseismic monitoring system.

Borehole instruments such as the Güralp Radian are a natural fit for detecting low magnitude microseismic events. Optional high gain at the higher frequencies makes the Radian extremely suitable for monitoring low-magnitude induced events while retaining long-period sensitivity for larger ruptures. The slim form factor and omni-angle operation allows the instrument to easily be lowered into decommissioned wells with little information about the orientation at depth.

The Radian is currently being utilised by the British Geological Survey as part of the UK GeoEnergy Test Bed (GTB) to monitor and improve understanding of fluid flow through natural subsurface pathways. A string of 6 interconnected Radians provides vertical profiling around the injection site with a maximum of 8 units able to join in a single string. The Radian will detect and monitor small changes in the subsurface at the GTB as part of the suite of monitoring technologies deployed onsite. 

In addition to onshore networks, offshore depleted gas fields are becoming increasingly scrutinised for potential to store CO2. The advent of Güralp omnidirectional sensor technology combined with acoustic near-real-time data transmission means the Aquarius OBS provides a cost-effective solution for monitoring offshore CCS sites, with infrequent and rapid battery recharging and acoustic data extraction while the unit is still on the seafloor.

How to cite: Reis, W., Cilia, M., Watkiss, N., Mohr, S., Barbara, R., and Hill, P.: Broadband seismic instrumentation for monitoring CCS sites, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-8224, https://doi.org/10.5194/egusphere-egu22-8224, 2022.

16:23–16:30
|
EGU22-5576
|
ECS
|
On-site presentation
|
Meng Li, Andre R Niemeijer, Femke C Vossepoel, and Ylona van Dinther

Induced seismicity triggered by fluid injection or extraction has been studied extensively in recent years. However, models relying on a Mohr-Coulomb yield criterion for interseismic loading or using a linear slip-weakening friction law for dynamic earthquake rupture cannot quantify well how much aseismic slip accumulates prior to nucleation or how to explain nucleation. Instead, a rate-and-state friction law is extensively utilized in earthquake cycle models to resolve and understand earthquake nucleation. Moreover, laboratory experiments indicate that the relevant lithologies in the Groningen subsurface are velocity-strengthening under in-situ temperature, pressure and fluid chemistry conditions [1]. This property should in theory lead to a lower chance of earthquake nucleation, which makes it difficult to explain the occurrence of earthquakes in Groningen. We study how to explain earthquake nucleation under velocity-strengthening friction and how much aseismic slip can be expected. In this study, we model the normal-fault setup of the Groningen field under reservoir depletion with rate-and-state friction. Initial conditions are chosen to mimic healing over millions of years prior to gas production. We implement fault loading due to fluid pressure reduction and validate our loading stresses with analytical predictions in Jansen et al. [2]. We provide constraints on how much aseismic slip to expect during nucleation and evaluate its relevance to induced seismicity in Groningen. We systematically investigate scenarios with various fluid extraction rates and different rate-and-state friction properties (including rate-strengthening, rate-weakening and a mixture of both) of surrounding lithologies using constraints from laboratory observations. In this way we explore the rate of stress change needed for nucleation under rate-strengthening friction. Currently, we produced an event with slip rate below seismic rate. If seismic rates cannot be reached, we will add a second state variable describing cohesion weakening with time to assess how it affects earthquake nucleation. The impact of frictional property and stress rate to aseismic slip build-up and earthquake nucleation is compared to what is caused by varied reservoir off-set distance and fault dipping angle. Sequences of earthquakes and aseismic slips are studied to understand the long-term effect of fluid extraction, with the influence of the planned gas production termination taken into account. We find that during continuous fluid depletion earthquakes reoccur at increasing recurrence interval. Large dipping angle and relatively low Poisson ratio are necessary to achieve this if reservoir offset is zero. Slip or strain nucleation and distribution patterns produced by our models provide hints that can guide seismologists to identify aseismic slip from natural observations, which can in turn, support this study and constrain simulated fault properties. Ultimately, this will help to better understand the nucleation of induced seismicity with similar lithologies that are present across northwestern Europe and lead to a better understanding of the relevance of aseismic slip.

 

References

[1] Hunfeld, L. B., Niemeijer, A. R., & Spiers, C. J. (2017).  Journal of Geophysical Research: Solid Earth, 122(11), 8969-8989.

[2] Jansen, J. D., Singhal, P., & Vossepoel, F. C. (2019).  Journal of Geophysical Research: Solid Earth, 124, 7193– 7212.

 

How to cite: Li, M., Niemeijer, A. R., Vossepoel, F. C., and van Dinther, Y.: Deciphering fluid extraction-induced earthquake nucleation in Groningen under rate-strengthening friction, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-5576, https://doi.org/10.5194/egusphere-egu22-5576, 2022.

16:30–16:37
|
EGU22-5398
|
ECS
|
On-site presentation
|
La Ode Marzujriban Masfara, Cornelis Weemstra, and Thomas Cullison

In May 2019, an earthquake with a magnitude of 3.4 (local magnitude) hit the area of the Westerwijtwerd village in the province of Groningen, the Netherlands. The event is the result of the gas extraction in the Groningen gas field and is one of the largest events to date. To better understand the source characteristics of the event, we apply a probabilistic full-waveform inversion technique that we recently developed to the event's recordings. Specifically, we use a variant of the Hamiltonian Monte Carlo (HMC) algorithm. When sampling high-dimensional model spaces, HMC is proven to be more efficient than the generic Metropolis-Hasting algorithm. Compared to probabilistic inversions of tectonic events, two main challenges arise while applying the algorithm. First, the prior information of the event is usually incomplete and inaccurate. That is, the only available information is (an estimate of) the hypocenter and origin time. Second, the frequency content of the induced event's seismograms is higher than that of typical tectonic events. This implies a higher non-linearity, which in turn complicates the ability of a probabilistic inversion algorithm to sample the model spaces, particularly when considering the first challenge. Consequently, to address both challenges, first, we develop a procedure to estimate the necessary prior information and use it as input to the HMC variant. Second, we run our HMC algorithm iteratively to mitigate the non-linearity. Using the relatively detailed velocity model of the Groningen gas field, we eventually estimate ten posteriors of the source parameters. The latter being the hypocenter (three parameters), the moment tensor (six independent parameters), and the origin time.  

How to cite: Masfara, L. O. M., Weemstra, C., and Cullison, T.: Characterizing induced seismic events in the Groningen gas field using an efficient Hamiltonian Monte Carlo sampler: a case study, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-5398, https://doi.org/10.5194/egusphere-egu22-5398, 2022.

Coffee break
Chairpersons: Antonio Pio Rinaldi, Federico Ciardo, Silvia De Simone
Modeling
17:00–17:07
|
EGU22-5536
|
ECS
|
On-site presentation
Federico Ciardo and Antonio Pio Rinaldi

Injection of fluid in fractured reservoirs triggers seismicity that migrates away from injection point. The enlarging cloud of (micro-)seismicity can be driven by pore-fluid diffusion within fractured rock mass, thus propagating in space proportional to square root of time for an effective isotropic and homogenous medium, or by elastic-stress interactions between over-stressed pre-existing fractures.

In this contribution we adopt an hybrid approach to model seismicity evolution driven by pore-fluid propagation into a Discrete Fracture Network and apply it to a large-scale injection experiment at FORGE Test Site in Utah (USA). We couple a statistical seed model for seismicity with a physic-based solver for non-linear pore-fluid diffusion into a three-dimensional DFN (using TOUGH3). Local inelastic permeability changes mimick irreversible deformations and affect pore-fluid evolution and hence seismicity cloud.

Several synthetic catalogs are generated and compared with one generated with a pure physic-based numerical solver.

 

How to cite: Ciardo, F. and Rinaldi, A. P.: Modeling of injection-induced seismicity in fractured rock masses with TOUGH3-seed hybrid solver, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-5536, https://doi.org/10.5194/egusphere-egu22-5536, 2022.

17:07–17:14
|
EGU22-6361
|
Virtual presentation
|
Hossein Ebrahimian, Fatemeh Jalayer, and Vincenzo Convertito

Methodology:

Induced earthquakes have peculiar characteristics such as, relatively shallow depths, small magnitude, correlation with field operations, non-GR recurrence law, and eventually non-homogenous Poisson recurrence time. Indeed, induced seismicity tends to cluster in limited volumes near the wells where field operations (e.g., fluids injection, extraction, fracking, etc.) are performed. A novel and fully-probabilistic simulation-based procedure is presented for providing temporal and volumetric predictions of induced events’ occurrence in a prescribed forecasting time interval (in the order of hours or days). The procedure aims at exploiting the information provided by the ongoing sequence in quasi-real time (even in the presence of very limited registered data) to adaptively update the seismicity forecasts based on the incoming information as it becomes available. The clustering of seismic events in volume (3D seismicity) and time is modelled based on an Epidemic Type Aftershock Sequence (ETAS) model. The proposed 3D ETAS model encompasses a decoupled depth-area volumetric probabilistic kernel that incorporates kernel density functions for areal extent as well as the focal depth. The ETAS parameters are going to be re-calibrated in order to take into account non-GR long-term temporal boundary conditions in case of induced seismicity. Moreover, exact spatial integrals will be used to consider the 3D boundary conditions. The proposed procedure considers the uncertainties in the earthquake occurrence model parameters in a Bayesian updating framework. Pairing up the Bayesian inference and the suitable efficient simulation schemes (using Markov Chain Monte Carlo Simulation) provides the possibility of performing the forecasting procedure with minimum (or no) need of human interference.

Application:

The procedure is demonstrated through retrospective forecasting of induced seismicity recorded at the Geysers geothermal field in northern California in the time period of 2011-2015. Injection of cold water and heavier liquids in the hot reservoir caused induced earthquakes with moment magnitudes in the range of [0.0, 4.0] and depth ranging up to 5 km. The proposed procedure is examined for both Bayesian updating of the proposed 3D ETAS model parameters and forecasting of the number of events of interest expected to occur in various time intervals before and after a number of main events within the seismic sequence. The seismicity is predicted within a confidence interval from the mean estimate. Adding a kernel density for the focal depth and moving towards the 3D seismicity forecasting leads to the forecasted number of events that better match the events that actually took place in the forecasting interval, as compared to the 2D ETAS model. Therefore, it is concluded that the proposed 3D ETAS model is quite effective in case of induced seismicity.

How to cite: Ebrahimian, H., Jalayer, F., and Convertito, V.: A 3D ETAS model for forecasting spatio-temporal distribution of induced seismic events, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-6361, https://doi.org/10.5194/egusphere-egu22-6361, 2022.

17:14–17:21
|
EGU22-12568
|
ECS
|
Virtual presentation
|
Victor Clasen Repolles, Antonio Pio Rinaldi, Federico Ciardo, and Luigi Passarelli

Within the workflow of Adaptive Traffic Light System, it is important to evaluate the performance of different induced seismicity forecasting models in order to properly weight the forecasts during seismic hazard calculation. In this respect, we propose a standardize test bench approach capable of comparing outputs’ models (in terms of seismicity rate) and their uncertainties in real time. We test this approach using different models that are trained using existing datasets from geothermal exploration campaigns. Notably, we use two statistical models that link injection volumetric rate to seismicity rate with the difference that a Bayesian approach (EM1_BH) additionally adds epistemic uncertainty to the aleatoric uncertainty introduced in a purely frequentist approach (EM1_MLE), one pressure-based seismicity model (HM0_CAPS) based on 1D analytical solution for linear pore-fluid diffusion and finally one hybrid 1D model that includes a physic-based module for linear and non-linear pore-fluid diffusion linked to a stochastic model for seismicity generation using a seed approach (HM0_SEED and HM1_SEED). By using these different models and their uncertainties in our numerical investigations, we show the robustness of the proposed testbench approach.

How to cite: Clasen Repolles, V., Rinaldi, A. P., Ciardo, F., and Passarelli, L.: Performance comparison of induced seismicity forecasting models with existing datasets, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-12568, https://doi.org/10.5194/egusphere-egu22-12568, 2022.

17:21–17:28
|
EGU22-12474
|
ECS
|
On-site presentation
Mohammad Javad Afshari Moein and Qinghua lei

Induced seismicity is a major challenge for fluid injection operations performed by geo-energy industry to exploit the underground resources. Despite recent developments in the understanding of induced earthquakes, many high-pressure fluid injection operations can still trigger unexpectedly large-magnitude events. A physical understanding of geological parameters controlling the induced seismicity is of central importance for improving our ability to forecast and mitigate the risk of inducing large earthquakes. Current physics-based numerical models are typically based on some simplifications that disregard the multiphysical interactions among fractures and faults. Therefore, the physical linkage between geometrical attributes of the fracture system and the statistics of induced seismicity is poorly understood. The final objective of this research is to determine the impact of fracture network properties on the spatiotemporal evolution of injection-induced seismicity and the emergence of large earthquake events.  

Here, we numerically capture the occurrence of seismic and aseismic slips in fracture systems, represented as discrete fracture networks (DFNs), spanning over two orders of magnitude over the length scale (1-100 m). Then, a 2D finite element model is used to simulate the coupled hydraulic and mechanical processes during fluid injection and analyze the occurrence of earthquakes. We present some preliminary results of our numerical simulations based on synthetic fracture network realizations. We particularly focus on power-law exponent of fracture length distribution and analyze the potential controls on the magnitude frequency of induced seismic events. The results of the analysis could have significant implications injection-related activities such as enhanced geothermal systems.

How to cite: Afshari Moein, M. J. and lei, Q.: Impact of fracture length distribution on the injection-induced seismicity in fractured rocks, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-12474, https://doi.org/10.5194/egusphere-egu22-12474, 2022.

17:28–17:35
|
EGU22-9880
|
ECS
|
On-site presentation
Jinlin Jiang, Pierre Dublanchet, Franҫois Passelègue, Dominique Bruel, and Frederic Pellet

Geothermal energy is one of the most promising techniques to exploit renewable energy resources from the Earth and to limit emissions of greenhouse gas. Deep geothermal exploitations are associated with long term fluid circulation and pressure perturbations at great depth, in fractured and faulted zones and are likely associated with a risk of triggering earthquakes. Such earthquakes are usually interpreted as the reactivation of rapid (m/s) shear slip on critically stressed faults caused by fluid flow and poroelastic stress changes. In some cases however, slow aseismic slip (m/d) can take place on faults in response to fluid flow. How fluid pressure perturbations reactivate aseimic or rapid slip still remains poorly understood. A better understanding of the hydromechanical processes controlling fault slip is therefore crucial to mitigate seismic hazards associated with geothermal exploitation.

In this framework, our study aims at constraining the influence of stress state, fluid injection rate, diffusivity and frictional failure criterion on the reactivation of slip on pre-existing faults through mechanical modelling of a set of laboratory experiments. The experiments consist of a fluid injection into a saw-cut rock sample loaded in a triaxial cell. Fault reactivation is triggered by injecting fluids through a borehole directly connected to the fault. This experimental setup is modelled by a 3D Finite Element Method (FEM) coupled with a solver of the fluid diffusion. The sample fault is modelled as a contact surface obeying slip-weakening Mohr-Coulomb friction law. This approach allows to compute slip and stress evolutions, as observed during the laboratory experiment. The FEM model is calibrated and is able to reproduce the experimental results. We show that fluid injection triggers a shear crack that propagates varying from 1 to 300 m/d along the fault. This approach can be used to investigate the relationship between fluid front and slip front during reactivation, which is an important issue to control the effects of fluid injections at depth.

How to cite: Jiang, J., Dublanchet, P., Passelègue, F., Bruel, D., and Pellet, F.: Numerical modelling of fluid-induced fault slip reactivation,application to Geo-Energy systems, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-9880, https://doi.org/10.5194/egusphere-egu22-9880, 2022.

17:35–17:42
|
EGU22-10392
|
On-site presentation
Antonio Pio Rinaldi, Vanille Ritz, Shyam Nandan, Raymi Castilla, Dimitrios Karvounis, and Stefan Wiemer

The Hellisheiði Geothermal Field is situated in Southwest Iceland and composes the Southern part of the Hengill Volcanic System. This area is characterized by a complex triple junction between three tectonic features: the Reykanes Peninsula rifting, South Iceland Volcanic Zone and West Volcanic  Zone. Reinjection of spent geothermal fluids is distributed mostly in two areas (Gráuhnúkar and Húsmúli), comprising respectively 6 and 5 active injection wells. The Húsmúli reinjection area, commissioned in September 2011 and has seen significant seismicity associated with drilling and injection operations.
In the framework of the Geothermica project COSEISMIQ (http://www.coseismiq.ethz.ch/en/home/), a dense temporary network was installed to monitor the seismicity in the Hengill region between December 2018 and August 2021. With this enhanced network, novel analysis and relocation techniques, a high resolution relocated catalogue was curated and comprises over 3600 events in the Húsmúli area.
We use numerical models, some purely statistical (ETAS and Seismogenic index) and a hybrid model (TOUGH2-Seed) to reproduce observed seismicity in the Húsmúli reinjection area during the COSEISMIQ project. We employ a pseudo-forecasting approach and compare models performances
and fit to the recorded data.

How to cite: Rinaldi, A. P., Ritz, V., Nandan, S., Castilla, R., Karvounis, D., and Wiemer, S.: Modelling injection induced seismicity in the Hengill geothermal field, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-10392, https://doi.org/10.5194/egusphere-egu22-10392, 2022.

17:42–17:49
|
EGU22-8716
|
ECS
|
Presentation form not yet defined
Sri Kalyan Tangirala, Francesco Parisio, and Victor Vilarrasa

Enabling a widespread exploitation of Enhanced Geothermal Systems (EGS) around the world by tapping into the heat trapped by the radioactive granites demands a better understanding of the fluid-induced seismicity associated with their stimulation. Induced seismicity occurs not only during hydraulic stimulation, but also after shut-in. The induced earthquakes of Mw > 3 at Basel and Mw = 5.5 at Pohang are two well-known examples that have caused a negative public perception on EGS. Here, we numerically compare the effect of bleed-off on the mitigation of post-injection seismicity for three stimulation schemes: constant rate, step rate and cyclic injection. We find that applying bleed-off in the post-injection phase significantly reduces the post-injection induced seismicity when compared to not applying bleed-off in all the injection schemes.

How to cite: Tangirala, S. K., Parisio, F., and Vilarrasa, V.: Numerical investigation of hydraulic stimulation strategies to mitigate post-injection seismicity in Enhanced Geothermal Systems, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-8716, https://doi.org/10.5194/egusphere-egu22-8716, 2022.

17:49–17:56
|
EGU22-10601
|
ECS
|
Virtual presentation
|
Riddhi Mandal and Semechah Lui

Heightened seismic activity due to human activities, such as wastewater injection, carbon storage and geothermal energy production, has been a rising problem in recent years. Various injection parameters and geological conditions have been shown to affect fault behaviour differently when fluid is injected on the faults, although existing observational studies about their effects often show contradictory results. Aseismic slip is also known to affect seismicity, but its exact contribution remains elusive.

To address these, we perform numerical modelling to study the effects of various injection parameters on fault slip behaviour. Our fully dynamic fault model is governed by the rate-and-state friction laws and spontaneously resolves all stages of an earthquake cycle and long-term fault slip. Our results show several interesting observations on the role of injection volume and rate: First, the injected volume can advance or delay the next earthquake if no earthquakes are directly triggered during perturbation. Second, if earthquakes are triggered, the number of triggered earthquakes is controlled by the rate at which fluid is injected, while the timings of the triggered earthquakes are controlled by the injected volume. Large triggered earthquakes are usually preceded by smaller precursors. Third, the pore-pressure threshold at which earthquakes are triggered changes depending on the injection parameters. In most cases, it increases with the volume of injected fluid, but in some cases when the injection is slow, it can also depend on the rate of injection. The change with respect to injection rate is not a smooth positive trend, however, as increasing the rate causes aseismic transients to grow stronger and transition into seismic events, thus advancing the triggering time and causing decrease in the threshold pore pressure in the process. Overall, the effects of perturbation do not end as soon as injection stops. Instead, heightened aseismic activities, as well as oscillating earthquake timings and magnitudes occur for multiple seismic cycles after the end of pore-pressure perturbation. We also see large variations in aseismic moment release under different perturbation scenarios and its intricate relationship with the resulted seismicity pattern, which confirms the vital role of aseismic slip in earthquake triggering. Similar to previous studies, we find that energy on the fault is primarily released aseismically.

Our results thus far are based on spatially uniform pore-pressure evolution, and we are currently developing models that resemble environments with temporally and spatially heterogeneous pore pressure by coupling the temporal evolution of pore pressure with spatial diffusion. We are also incorporating geologic information of the crustal medium, which will be more fitting for modelling realistic scenarios such as the injection-induced earthquakes in Oklahoma.  

How to cite: Mandal, R. and Lui, S.: Quantifying the effects of injection parameters on fault response under spatially homogenous and heterogenous pore-fluid conditions, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-10601, https://doi.org/10.5194/egusphere-egu22-10601, 2022.

General
17:56–18:06
|
EGU22-16
|
ECS
|
solicited
|
Presentation form not yet defined
Ryan Schultz, William Ellsworth, and Gregory Beroza

Earthquakes caused by human activities receive scrutiny due to the risks and hazards they pose.  Seismicity that occurs after the causative anthropogenic operation stops has been particularly problematic – both because of high-profile cases of damage caused by this trailing seismicity and due to the loss of control for risk management.  With this motivation, we undertake a statistical examination of how induced seismicity stops.  We borrow the concept of Båth’s law from tectonic aftershock sequences.  Båth’s law anticipates the difference between magnitudes in two subsets of seismicity as dependent on their population count ratio.  We test this concept for its applicability to induced seismicity, including ~80 cases of earthquakes caused by hydraulic fracturing, enhanced geothermal systems, and other fluid-injections with clear operational end points.  We find that induced seismicity obeys Båth’s law: both in terms of the magnitude-count-ratio relationship and the power law distribution of residuals.  Furthermore, the distribution of count ratios is skewed and heavy-tailed, with most earthquakes occur during stimulation/injection.  We discuss potential models to improve the characterization of these count ratios and propose a Seismogenic Fault Injection Test to measure their parameters in situ.  We conclude that Båth’s law quantifies the occurrence of earthquake magnitudes trailing anthropogenic operations.

How to cite: Schultz, R., Ellsworth, W., and Beroza, G.: Statistical bounds on how induced seismicity stops, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-16, https://doi.org/10.5194/egusphere-egu22-16, 2022.

18:06–18:13
|
EGU22-5699
|
On-site presentation
Luigi Passarelli, Louis De Barros, Antonio Pio Rinaldi, and Stefan Wiemer

A growing number of direct and indirect measurements and observation indicates that aseismic slip transients are often induced during fluid injection operation alongside with swarm-like seismicity. The detection of fluid-induced aseismic slip has made a paradigm shift on our understanding of the spatio-temporal evolution of earthquake activity during injection operation, classically interpreted as triggered by a diffusive front of high pore pressure. Instead, unclamping of the fault by pressurization induced by fluid injection creates the condition to nucleate synchronous aseismic and seismic slip transients.  In this scenario, the spatio-temporal evolution of the induced seismicity is driven by the stressing rate imparted at the leading edges of the aseismic rupture front. However, the relationship between the magnitude of aseismic slip and the hydraulic energy input in the system remains still elusive. A similar mechanism has been proposed for natural earthquake swarms triggered by shallow (5-10 km depth) slow slip events (SSEs), for which a robust power-law scaling has been demonstrated between seismic and aseismic slip. Notably, the power-law moment scaling of shallow SEEs and associated earthquake swarms has been interpreted with a mechanism of fault pressurization enhanced by intense fracturing in a seismogenic volume with abundance of crustal fluids. Similar fault conditions are at play for fluid-induced seismicity. Here, we collected several case studies of recorded induced aseismic deformation during injection experiments together with the accompanying seismic activity. We investigated the spatial distribution and temporal evolution of the seismicity with respect to the ongoing transient aseismic slip. We focused in particular on the seismic and aseismic slip budget of induced seismicity and compared it with previous scaling of SSEs. The aseismic and seismic moments of induced events are compatible with the power-law scaling of shallow natural earthquakes swarms triggered by SSEs, although a data gap exits for SSEs in 0-4 magnitude range, where no SSEs have never been recorded due to the low resolution of surface geodetic instrumentation. We performed also numerical simulation using a 3D hydro-mechanical model using realistic fault and hydraulic parameters in order to fill in the data gap. Our results serve as a basis to build up empirical models that incorporate aseismic slip together with injected volume and pressure to forecast seismicity during fluid-injection experiments.

How to cite: Passarelli, L., De Barros, L., Rinaldi, A. P., and Wiemer, S.: Seismic to aseismic slip scaling during fluid injection experiments, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-5699, https://doi.org/10.5194/egusphere-egu22-5699, 2022.

18:13–18:20
|
EGU22-10043
|
Virtual presentation
Fuzhen Li, TianXiang Ren, ShunLiang Chi, Huai Zhang, and YaoLin Shi

Sufficient shreds of evidence have proved the existence of the remote triggering effect of large earthquakes. To understand its mechanism, it is necessary to conduct detailed investigations on the influence of the far-field dynamic stress changes on the stress state of faults. As an important tool of ultra-broadband crustal stress monitoring, a four-component borehole strainmeter can directly record the dynamic changes of horizontal strain and stress caused by seismic waves. These data are of great importance to study the dynamic Coulomb stress changes and related triggering effects, but have not been paid sufficient attention to so far. This paper analyzes the data of the four-component borehole strainmeter at Gaotai and Tonghua stations, which recorded the far-field strain changes of four major earthquake events in the Pacific region in 2018. We successfully identify the seismic phases of P, S, and surface waves, and analyze the characteristics of different phases through the stress petal method. The dynamic stress changes are calculated, demonstrating the feasibility of using borehole strainmeter data to quantitatively study the triggering effect of teleseismic waves of earthquakes with different magnitudes at different epicentral distances. We find that the direction of the principal stress axis of the dynamic stress changes is generally consistent with the azimuth of the earthquake epicenter. We further discuss the Coulomb stress changes on the major faults near the stations. According to the results, the peak values of the dynamic Coulomb stress changes produced by four earthquakes on the fault planes near Gaotai and Tonghua stations are at the magnitude of hundreds of Pa, which are lower than the threshold value of dynamic triggering. This is also consistent with the observation that no dynamically triggered earthquakes are found on the faults. However, the idea and method of this paper provide useful insight into the detection of possible dynamic triggering of large earthquakes.

How to cite: Li, F., Ren, T., Chi, S., Zhang, H., and Shi, Y.: The dynamic Coulomb stress changes caused by remoteearthquakes based on the borehole strainmeter data, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-10043, https://doi.org/10.5194/egusphere-egu22-10043, 2022.

18:20–18:27
|
EGU22-10911
|
ECS
|
Virtual presentation
|
Lavinia-Cristina Varzaru and Felix Borleanu

Seismicity in western Romania is the result of tectonic processes that continuously shaped the landscape generating a fractured crust, which showed significant movements as a result of overall tectonic stress in the area as well as secondary effects such as erosion or lateral density fluctuations. At the same time, this region has an important natural resource, being identified here various deposits that have been intensively explored lately. The exploitation of these resources, as well as the development of the infrastructure in the region, led to the generation of anthropogenic seismic events. Due to the recent improvement of the Romanian Seismic Network, the coverage with seismic stations increased and these events were detected and located as natural tectonic events, contaminating the Romanian earthquakes (ROMPLUS) catalog.

To eliminate anthropogenic event contamination in the ROMPLUS catalog, we ran a complex statistical approach on the catalog data. In addition, to build a robust discriminant, we further applied cross-correlation and spectral analysis algorithms on the seismic waveforms recorded between 2014 and 2021 by the Surduc (SURR) and Gura Zlata (GZR) stations, which are located in the proximity of the major clusters of seismic events.

Our results showed a good distinction between tectonic and anthropogenic events and revealed that most of the clustered events are located near the explorations sites. We also noted that most of the events occurred during working hours. At the same time, the high similarity among these events indicates the existence of repetitive seismic sources.

How to cite: Varzaru, L.-C. and Borleanu, F.: Identifying anthropogenic seismic events generated in western Romania using statistical approaches and novel waveform processing techniques, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-10911, https://doi.org/10.5194/egusphere-egu22-10911, 2022.