Investigation of induced seismicity and aseismic deformation in response to CO2 injection and related pressure changes

In recent years of global warming, the reduction of CO2 emissions e.g. by employing carbon capture and storage (CCS) became increasingly important. To reduce the current speed of global warming, a tremendous amount of CO2 would need to be injected within geological subsurface storages. The main threat to the integrity of CO2 storages is that even small to moderate-sized earthquakes triggered by injection of these large volumes may break its seal. Since 2007, CCS research has been carried out by the CO2 Lab of the University Centre in Svalbard (UNIS). Whereas the first fluid injection experiment in 2010 was associated with small magnitude earthquakes, later injection tests did not generate any detectable events. Nevertheless, pressure and flow rate showed a pattern characteristic for fracture opening, potentially indicating aseismic fracture propagation or slow slip. Especially, we focussed on the following problems: - What is the value of microseismic monitoring? - Is there aseismic deformation during fluid injection at the CO2 Lab? - Can we make use of the continuity of seismic recordings by analysing the noise in addition to the signal? - Is it feasible to turn Svalbard into a CO2 neutral community by injecting a sufficient amount of CO2 in subsurface reservoir(s) without breaching the caprock? Within the SafeCO2 II project, we could gather important information to answer these questions. General project results are transferrable to other potential CO2 injection sites: 1) The results obtained in this project support the effort to transform Svalbard into a CO2 neutral community by injecting a sufficient amount of CO2 in subsurface reservoirs. The volumes of injected water have been small; however, the amount of CO2 injected at a potential storage site will also be small. From our experience in the project, a combination of downhole and shallow borehole sensors is efficient for detection and location of injection-induced seismicity. In addition, for a long-term monitoring of a potential storage site, data processing may be automated. Unexpected difficulties resulting from inherent timing problems between data recorders as well as electronic noise affecting the data recorded at the CO2 lab could be solved successfully and may be avoided in future projects. Both data analyses on the SEISVAL network and the SPITS array suggest that no local natural seismicity seems to occur, meaning that most probably, no active faults are located within Adventdalen endangering the success of a potential CO2 storage. Injection-induced seismicity, probably rather consists of small magnitude events. We succeeded in testing and comparing several methods to detect and locate such weak events. For this work, a good knowledge of the velocity model is essential and important groundwork has been performed on the improvement of a NORSAR3D model. This model could be validated by comparing Green's functions from finite difference wave propagation and ambient seismic noise measurements. 2) We successfully analysed ambient seismic noise in addition to microseismic events. During the course of the project, we concentrated on cross-correlation of ambient seismic noise that mainly allows for extraction of Green's functions describing the subsurface structure. When applying this technique to the permanent network at the CO2 Lab, the changes in daily cross-correlation function were challenging to interpret, maybe caused by the difficulties we experienced with the sensors of the CO2 Lab (being affected by electronic noise and, partially, by tube waves caused by gas ascending from organic layers). Nevertheless, the cross-correlation functions computed from ambient seismic noise recorded on the temporary SEISVAL network show clear signals of Rayleigh waves propagating through the valley. 3) The injection pressure/flow rate behaviour during injections, geomechanical modelling as well as the laboratory experiment on a shale sample from the Rurikfjellet formation, suggest that aseismic deformation occurs very probably during injection at the CO2 lab and may explain the low seismicity observed during injection experiments. Although we were not able to observe such events directly, we developed and tested the methodology to detect and locate them. 4) We were able to install and maintain a broadband seismometer network throughout Adventdalen despite the challenges imposed by the Arctic surroundings and climate conditions. 5) Due to the low number of microseismic events detected during injections at the CO2 lab site, we could not test methods for more sophisticated data analysis with this data set. Instead, we included the analysis of a data set from an Australian geothermal experiment at Paralana to prove that both stress drop and b-values vary laterally and temporally. Employing waveform cross-correlation, fracture systems found from event locations could be assessed critically.

In recent years of global warming, the reduction of CO2 emissions e.g. by employing carbon capture and storage (CCS) became increasingly important. To reduce the current speed of global warming, a tremendous amount of CO2 would need to be injected within geological subsurface storages. The main threat to the integrity of CO2 storages is that even small- to moderate-sized earthquakes triggered by injection of these large volumes may break its seal. Since 2007, CCS research has been carried out by the CO2 Lab of the University Centre in Svalbard (UNIS). So far, several injection tests have been conducted and a high-frequency geophone network surrounding the injection well has been established to detect microseismic events. Whereas the first fluid injection in 2010 was associated with high-frequency earthquakes, later injection tests did not generate any detectable events. Nevertheless, pressure and flow rate showed a pattern characteristic for fracture opening potentially indicating aseismic fracture propa gation or slow slip. We will focus on the following problems: - What is the value of microseismic monitoring: do event locations form a good representation of the main flow directions? Does microseismic activity allow for conclusions on rock properties or only helps to identify regions of high pore pressure? - Is there aseismic deformation during fluid injection at the CO2 Lab? How can we quantify its amount? What can these observations contribute to the understanding of the injective/hydraulic behaviou r of the rock? - Can we make use of the continuity of seismic recordings by analysing the noise in addition to the signal? Can we employ ambient seismic noise tomography to enhance our understanding of fluid flow and changes in the subsurface qualitativel y as well as quantitatively? - Is it feasible to turn Svalbard into a CO2 neutral community by injecting a sufficient amount of CO2 in subsurface reservoir(s) without breaching the caprock?