Earthquake catalog of induced seismicity associated with 2020 hydraulic stimulation campaign at OTN-2 well in Helsinki, Finland
Cite as:
Kwiatek, Grzegorz; Martínez-Garzón, Patricia; Karjalainen, Aino (2022): Earthquake catalog of induced seismicity associated with 2020 hydraulic stimulation campaign at OTN-2 well in Helsinki, Finland. GFZ Data Services. https://doi.org/10.5880/GFZ.4.2.2022.001
Status
I N R E V I E W : Kwiatek, Grzegorz; Martínez-Garzón, Patricia; Karjalainen, Aino (2022): Earthquake catalog of induced seismicity associated with 2020 hydraulic stimulation campaign at OTN-2 well in Helsinki, Finland. GFZ Data Services. https://doi.org/10.5880/GFZ.4.2.2022.001
Abstract
This data publication contains seismic catalog developed by the analysis of seismicity recorded during hydraulic stimulation campaign performed in May 2020 in the 5.8-km deep OTN-2 well near Helsinki, Finland as part of the St1 Deep Heat project (Kwiatek et al., 2022). The original seismic data to develop the seismic catalog were acquired with the high-resolution seismic network composed of 22 geophones surrounding the project site. The centerpiece of the network was a 10-level borehole array of Geospace OMNI-2400 geophones (3C/15 Hz) sampled at 2 kHz placed in the OTN-3 well adjacent to the OTN-2 injection well, and located at 1.93 - 2.55 km depth, approx. 3km from injection intervals. Additional 12 stations at distances <10 km from project site formed the satellite network that was equipped with short-period 3C 4.5 Hz Sunfull PSH geophones, completing the seismic network. Near-real-time processing of induced seismicity data started on Jan 26, 2020, i.e. about 3 months prior to the onset of the injection, covering entire period of the stimulation campaign in May 2020. The monitoring stopped end of June 2020, about one month after the stimulation finished. The monitoring campaign resulted in initial industrial seismicity catalog containing 6,243 events that was refined and further extended (cf. Kwiatek et al., 2022). The final catalog associated with this data publication contains 6,318 earthquakes, including 197, 5427 and 694 events recorded before, during, and after stimulation campaign. The core catalog data contains origin time, local magnitude, (re)location and focal mechanism data.
Methods
The initial induced seismicity catalog provided by the industrial operator contained 6,243 detections including earthquakes, electronic noises and signals originating from or near the surface. A coincidence trigger was first run on the database of P-wave arrivals not associated with any earthquake (see details of the procedure in Kwiatek et al., 2019; Leonhardt et al., 2021a,b). This enhanced the initial catalog by 3,720 newly detected events resulting in a total number of 9,963 detected events.
Automated inspection of observed hodographs was then performed, comparing the observed patterns of P- and S-wave arrivals on sensors forming the OTN-3 array with those predicted for earthquakes originating from injection site. This allowed to confirm that 6,318 detections out of the 9,963 are seismic events originating from the stimulated crustal volume, and these events constitute the catalog included in the data publication. The remaining 3,645 events, manually identified as electronic noises of various types or surface blasts were excluded from the catalog.
The vast majority of the 6,318 confirmed induced seismic events were visible only on seismograms from sensors forming the OTN-3 array, but not on the satellite sensors located within Helsinki municipal area. This means that the accurate hypocentral locations could not be achieved for most of events without additional P- and S- phase arrivals from sensors forming the satellite network. In fact, only the 72 largest events could be reliably located using additional P- and S-wave onsets available from the satellite network. These 72 earthquakes were further relocated using the Double-Difference method (Waldhauser and Ellsworth, 2000). For the location and relocation, the 1D P-wave velocity model based on a vertical seismic profiling campaign (Leonhardt et al., 2021a,b) assuming a V_P⁄V_S ratio of 1.71 for inverting S-wave arrival times was used. The relocation limited the final hypocentral location catalog to 45 events. The achieved relative location precision (95% confidence interval) for these events in horizontal and vertical direction was not exceeding ±85 m and ±42 m, respectively.
Although limited number of hypocenter locations could be estimated, the local “Helsinki” magnitude was calculated for all 6,318 events. This was possible because S-P times could be easily extracted from downhole recordings of OTN-3 array. To calculate the local magnitude, we followed the procedure presented by Uski and Tuppurainen (1996) and Uski et al. (2015), which was applied in our previous studies as well (Kwiatek et al., 2019; Leonhardt et al., 2021a,b).
Finally, 8 double-couple constrained focal mechanism were calculated using the hybridMT moment tensor inversion package (Kwiatek et al., 2016) and time integrals of the first P-wave ground displacement pulses including sign information (e.g. Amemoutou et al., 2021).
Authors
Kwiatek, Grzegorz;GFZ German Research Centre for Geosciences, Potsdam, Germany
CharacterString: This data publication contains seismic catalog developed by the analysis of seismicity recorded during hydraulic stimulation campaign performed in May 2020 in the 5.8-km deep OTN-2 well near Helsinki, Finland as part of the St1 Deep Heat project (Kwiatek et al., 2022). The original seismic data to develop the seismic catalog were acquired with the high-resolution seismic network composed of 22 geophones surrounding the project site. The centerpiece of the network was a 10-level borehole array of Geospace OMNI-2400 geophones (3C/15 Hz) sampled at 2 kHz placed in the OTN-3 well adjacent to the OTN-2 injection well, and located at 1.93 - 2.55 km depth, approx. 3km from injection intervals. Additional 12 stations at distances <10 km from project site formed the satellite network that was equipped with short-period 3C 4.5 Hz Sunfull PSH geophones, completing the seismic network. Near-real-time processing of induced seismicity data started on Jan 26, 2020, i.e. about 3 months prior to the onset of the injection, covering entire period of the stimulation campaign in May 2020. The monitoring stopped end of June 2020, about one month after the stimulation finished. The monitoring campaign resulted in initial industrial seismicity catalog containing 6,243 events that was refined and further extended (cf. Kwiatek et al., 2022). The final catalog associated with this data publication contains 6,318 earthquakes, including 197, 5427 and 694 events recorded before, during, and after stimulation campaign. The core catalog data contains origin time, local magnitude, (re)location and focal mechanism data.
pointOfContact
CI_ResponsibleParty
individualName
CharacterString: Kwiatek, Grzegorz
organisationName
CharacterString: GFZ German Research Centre for Geosciences, Potsdam, Germany
CI_OnLineFunctionCode (codeList=http://www.isotc211.org/2005/resources/Codelist/gmxCodelists.xml#CI_OnLineFunctionCode codeListValue=http://www.isotc211.org/2005/resources/Codelist/gmxCodelists.xml#CI_OnLineFunctionCode_information): information
CharacterString: The initial induced seismicity catalog provided by the industrial operator contained 6,243 detections including earthquakes, electronic noises and signals originating from or near the surface. A coincidence trigger was first run on the database of P-wave arrivals not associated with any earthquake (see details of the procedure in Kwiatek et al., 2019; Leonhardt et al., 2021a,b). This enhanced the initial catalog by 3,720 newly detected events resulting in a total number of 9,963 detected events.
Automated inspection of observed hodographs was then performed, comparing the observed patterns of P- and S-wave arrivals on sensors forming the OTN-3 array with those predicted for earthquakes originating from injection site. This allowed to confirm that 6,318 detections out of the 9,963 are seismic events originating from the stimulated crustal volume, and these events constitute the catalog included in the data publication. The remaining 3,645 events, manually identified as electronic noises of various types or surface blasts were excluded from the catalog.
The vast majority of the 6,318 confirmed induced seismic events were visible only on seismograms from sensors forming the OTN-3 array, but not on the satellite sensors located within Helsinki municipal area. This means that the accurate hypocentral locations could not be achieved for most of events without additional P- and S- phase arrivals from sensors forming the satellite network. In fact, only the 72 largest events could be reliably located using additional P- and S-wave onsets available from the satellite network. These 72 earthquakes were further relocated using the Double-Difference method (Waldhauser and Ellsworth, 2000). For the location and relocation, the 1D P-wave velocity model based on a vertical seismic profiling campaign (Leonhardt et al., 2021a,b) assuming a V_P⁄V_S ratio of 1.71 for inverting S-wave arrival times was used. The relocation limited the final hypocentral location catalog to 45 events. The achieved relative location precision (95% confidence interval) for these events in horizontal and vertical direction was not exceeding ±85 m and ±42 m, respectively.
Although limited number of hypocenter locations could be estimated, the local “Helsinki” magnitude was calculated for all 6,318 events. This was possible because S-P times could be easily extracted from downhole recordings of OTN-3 array. To calculate the local magnitude, we followed the procedure presented by Uski and Tuppurainen (1996) and Uski et al. (2015), which was applied in our previous studies as well (Kwiatek et al., 2019; Leonhardt et al., 2021a,b).
Finally, 8 double-couple constrained focal mechanism were calculated using the hybridMT moment tensor inversion package (Kwiatek et al., 2016) and time integrals of the first P-wave ground displacement pulses including sign information (e.g. Amemoutou et al., 2021).
Location
Please, click on markers, line or bounding boxes to see related details in popup.
Released
Dataset
Abstract
;GFZ German Research Centre for Geosciences, Potsdam, Germany
Files
