Periodic Reporting for period 3 - FEAR (Fault Activation and Earthquake Rupture)
Reporting period: 2023-09-01 to 2025-02-28
Earthquakes are among the most significant hazards to human society and continue to remain the most elusive. Advancing our ability to understand the occurrence and severity of future earthquakes is of paramount importance to making society more resilient to earthquake risks. Progress in further understanding of earthquake physics is hindered by the lack of appropriate experimental facilities for observing the earthquake process at a close distance. The Bedretto Underground Laboratory for Geosciences and Geoenergies (BedrettoLab) is a deep underground experimental facility being constructed in the Bedretto tunnel, at 1’000m depth in the Swiss Alps and offers a unique opportunity to perform fault stimulation and earthquake nucleation experiments on a scale and depth not available until now. The core idea of the Fault Activation and Earthquake Rupture (FEAR) project is to gain understanding of how earthquakes start and stop by using hydraulic stimulation to modify stress and initiate small non-damaging earthquakes (magnitude 1 range) on a natural target fault in the vicinity of the BedrettoLab.
A suite of 4 stimulation experiments are planned along subsequent segments of the target fault. In Experiment I, an unperturbed fault section will be stimulated and will serve as a baseline experiment to characterize the fault response to stimulation and estimate pre- and post-stimulation stress conditions, injectivity, and permeability. In Experiment II, a larger fault segment will be stimulated, with up to three simultaneous injection boreholes, with the goal of triggering a non-damaging, larger rupture with a target moment magnitude of ~1.0. Experiment III will take place next to the segment ruptured in Experiment II and will test to what extent the rupture from Experiment II conditioned the adjacent Experiment III fault section for further rupture. In Experiment IV, we attempt to actively condition another segment of the same fault, e.g. by circulating cold and temperate water in the surrounding rock mass for several weeks, to either clamp or unclamp certain portions of the fault.
To facilitate the installation of the multidisciplinary FEAR Integrated Monitoring System, a 120m tunnel (the FEAR tunnel) is excavated parallel to the target fault. Numerous monitoring and stimulation boreholes will be drilled from this FEAR tunnel. The FEAR tunnel and the associated boreholes will facilitate the deployment of a dense network of multidisciplinary sensors to capture the rupture preparation phase, the earthquake rupture itself, and the post-rupture response of the target fault at unprecedentedly close distances.
Real-time data from this instrumentation network will flow as inputs into a real-time adaptive traffic light system for risk-mitigation for induced seismicity, serving as a unique testbed for state-of-the-art earthquake forecast models. In parallel to the in-situ activities in the Bedretto tunnel, rock samples from the target faults will be tested in rock deformation laboratories using state-of-the-art friction and fracture testing machines. Numerical models capturing the strongly coupled non-linear thermo-hydro-mechanical processes involved in fault rupture will be developed to address the question of bridging the gap between investigations at the laboratory scale and those on natural faults.
The science objectives of the FEAR project address key questions in 6 areas of earthquakes and faulting science, among which include:
- Earthquake physics: How do earthquakes nucleate, propagate, and arrest? What is the role of pre-stress conditions and geometrical/rheological complexities (i.e. barriers)?
- Role of fluids: What roles do fluids, pore-pressure changes, heterogeneity of frictional properties and dynamic parameters play in the initiation and evolution of individual earthquake ruptures, and in seismicity patterns?
- Earthquake precursors: Can we observe earthquake precursors, especially ones that have been observed at the laboratory scale but not yet in the field? Are there any specific transient process diagnostics of an impending rupture?
- What happens on and around the fault zone? What is the interrelation between seismic and aseismic deformation within the fault zone and in the surrounding volume? How does deformation localize inside the fault zone? Can we modify the mode of slip behavior?
- How do we best forecast earthquakes? What are the most successful earthquake forecast models? Can we use the unusually dense sensor network to inform a physics-based seismicity forecast model that significantly outperforms the purely statistical models? What is the current, and what is the inherent, limit on the predictability of earthquakes?
- Implications for induced seismicity in geo-energy applications: What stress and injection conditions produce larger magnitude events? Can we recognize activatable fault segments a priori, i.e. before injections? To what extent, and how, can induced earthquakes, and seismic and aseismic slip be controlled?
A suite of 4 stimulation experiments are planned along subsequent segments of the target fault. In Experiment I, an unperturbed fault section will be stimulated and will serve as a baseline experiment to characterize the fault response to stimulation and estimate pre- and post-stimulation stress conditions, injectivity, and permeability. In Experiment II, a larger fault segment will be stimulated, with up to three simultaneous injection boreholes, with the goal of triggering a non-damaging, larger rupture with a target moment magnitude of ~1.0. Experiment III will take place next to the segment ruptured in Experiment II and will test to what extent the rupture from Experiment II conditioned the adjacent Experiment III fault section for further rupture. In Experiment IV, we attempt to actively condition another segment of the same fault, e.g. by circulating cold and temperate water in the surrounding rock mass for several weeks, to either clamp or unclamp certain portions of the fault.
To facilitate the installation of the multidisciplinary FEAR Integrated Monitoring System, a 120m tunnel (the FEAR tunnel) is excavated parallel to the target fault. Numerous monitoring and stimulation boreholes will be drilled from this FEAR tunnel. The FEAR tunnel and the associated boreholes will facilitate the deployment of a dense network of multidisciplinary sensors to capture the rupture preparation phase, the earthquake rupture itself, and the post-rupture response of the target fault at unprecedentedly close distances.
Real-time data from this instrumentation network will flow as inputs into a real-time adaptive traffic light system for risk-mitigation for induced seismicity, serving as a unique testbed for state-of-the-art earthquake forecast models. In parallel to the in-situ activities in the Bedretto tunnel, rock samples from the target faults will be tested in rock deformation laboratories using state-of-the-art friction and fracture testing machines. Numerical models capturing the strongly coupled non-linear thermo-hydro-mechanical processes involved in fault rupture will be developed to address the question of bridging the gap between investigations at the laboratory scale and those on natural faults.
The science objectives of the FEAR project address key questions in 6 areas of earthquakes and faulting science, among which include:
- Earthquake physics: How do earthquakes nucleate, propagate, and arrest? What is the role of pre-stress conditions and geometrical/rheological complexities (i.e. barriers)?
- Role of fluids: What roles do fluids, pore-pressure changes, heterogeneity of frictional properties and dynamic parameters play in the initiation and evolution of individual earthquake ruptures, and in seismicity patterns?
- Earthquake precursors: Can we observe earthquake precursors, especially ones that have been observed at the laboratory scale but not yet in the field? Are there any specific transient process diagnostics of an impending rupture?
- What happens on and around the fault zone? What is the interrelation between seismic and aseismic deformation within the fault zone and in the surrounding volume? How does deformation localize inside the fault zone? Can we modify the mode of slip behavior?
- How do we best forecast earthquakes? What are the most successful earthquake forecast models? Can we use the unusually dense sensor network to inform a physics-based seismicity forecast model that significantly outperforms the purely statistical models? What is the current, and what is the inherent, limit on the predictability of earthquakes?
- Implications for induced seismicity in geo-energy applications: What stress and injection conditions produce larger magnitude events? Can we recognize activatable fault segments a priori, i.e. before injections? To what extent, and how, can induced earthquakes, and seismic and aseismic slip be controlled?
The FEAR project spans 6 years, with a timescale dictated by the experiments in the Bedretto tunnel and divided in three successive work phases. At 54 months into the project, we are currently in Phase II and are focusing on the data analysis of the first 2 FEAR experiments (Mzero and FEAR experiment 1) and preparation for FEAR experiments 3 and 4. On the administrative side, during the first year, the FEAR administration and scientific working group structures were defined. The members of the External Advisory Board were selected, and the FEAR website has been published (http://fear-earthquake-research.org(opens in new window)).
After selection of the fault zone which will be investigated during the FEAR experiments the Geology WG pursued their activities by creating a geological model. The history of deformation of the MC fault is a key component of the hydraulic workflow for the experiments to come. This general fracture network had been developed based on structural geological mapping, remote sensing, exploration drilling and borehole logging, ground-penetration radar, and laboratory investigations. The fault is now defined as a a ductile–brittle shear zone several meters wide with an intensely fractured volume spanning over 100 m in length. This multi-strand fault zone is made of quartz and biotite shear fractures, ~2m thick, in otherwise relatively intact Rotondo granite. The complex geological model highlights the possibilities of earthquake rupture and flow leakage in the rock volume.
The Monitoring WG finalized the design and technical specifications of the monitoring network required for FEAR experiment 1, which will be a major component of the FEAR Integrated Monitoring System. The design of the monitoring network is based on experience gained in experiments such as VALTER4FEAR and Mzero and a complex combination of acoustic sensors, accelerometers, geophones and SIMFIP probes.
With the densely distributed borehole monitoring network, we can characterize the finite fault system and understand how the mainshock ruptures initiate, propagate and terminate at sub-meter scales.
The activities of the Laboratory and Modeling WGs take place in parallel to the preparation of experimental activities in the BedrettoLab. Over the last year, a range of tests have been conducted to understand the behavior of dry and fluid saturated specimens. A range of metrics were collected such as permeability, strength, seismogenic potential, etc. These results are relevant for the modeling WG.
The Modeling WG focused on testing various numerical codes and models which are used to address the scientific questions of FEAR. During the latest reporting period the focus was aimed on the preseismic phase with multiple efforts on forecasting, driving dynamic effect on the rock characteristics and coseismic slip. Laboratory experiments have been conducted to model strain localization processes and acoustic emissions. The latter is essential to try upscaling laboratory results to the Bedretto experiments. Multiple models came out of different workflows to underline the fault rupture dynamics and the role of the permeability barriers in the seismicity activity.
Construction of the FEAR sidetunnel, providing access to the selected fault zone has started in Q3 2024. In January 2024 the first niches required for the tunneling were finalized. Construction was halted for a 6-month period to implement the FEAR Mzero experiments in the Geothermal testbed of the Bedretto tunnel. Experiment Mzero A and B took place respectively in April and August 2024. Tunneling activities resumed in September 2024 and have been interrupted briefly for the execution of FEAR experiment 1 in november and december 2024. Tunneling activites resumed in January 2025 and will be finalized by Q3 2025.
After selection of the fault zone which will be investigated during the FEAR experiments the Geology WG pursued their activities by creating a geological model. The history of deformation of the MC fault is a key component of the hydraulic workflow for the experiments to come. This general fracture network had been developed based on structural geological mapping, remote sensing, exploration drilling and borehole logging, ground-penetration radar, and laboratory investigations. The fault is now defined as a a ductile–brittle shear zone several meters wide with an intensely fractured volume spanning over 100 m in length. This multi-strand fault zone is made of quartz and biotite shear fractures, ~2m thick, in otherwise relatively intact Rotondo granite. The complex geological model highlights the possibilities of earthquake rupture and flow leakage in the rock volume.
The Monitoring WG finalized the design and technical specifications of the monitoring network required for FEAR experiment 1, which will be a major component of the FEAR Integrated Monitoring System. The design of the monitoring network is based on experience gained in experiments such as VALTER4FEAR and Mzero and a complex combination of acoustic sensors, accelerometers, geophones and SIMFIP probes.
With the densely distributed borehole monitoring network, we can characterize the finite fault system and understand how the mainshock ruptures initiate, propagate and terminate at sub-meter scales.
The activities of the Laboratory and Modeling WGs take place in parallel to the preparation of experimental activities in the BedrettoLab. Over the last year, a range of tests have been conducted to understand the behavior of dry and fluid saturated specimens. A range of metrics were collected such as permeability, strength, seismogenic potential, etc. These results are relevant for the modeling WG.
The Modeling WG focused on testing various numerical codes and models which are used to address the scientific questions of FEAR. During the latest reporting period the focus was aimed on the preseismic phase with multiple efforts on forecasting, driving dynamic effect on the rock characteristics and coseismic slip. Laboratory experiments have been conducted to model strain localization processes and acoustic emissions. The latter is essential to try upscaling laboratory results to the Bedretto experiments. Multiple models came out of different workflows to underline the fault rupture dynamics and the role of the permeability barriers in the seismicity activity.
Construction of the FEAR sidetunnel, providing access to the selected fault zone has started in Q3 2024. In January 2024 the first niches required for the tunneling were finalized. Construction was halted for a 6-month period to implement the FEAR Mzero experiments in the Geothermal testbed of the Bedretto tunnel. Experiment Mzero A and B took place respectively in April and August 2024. Tunneling activities resumed in September 2024 and have been interrupted briefly for the execution of FEAR experiment 1 in november and december 2024. Tunneling activites resumed in January 2025 and will be finalized by Q3 2025.
FEAR will be the first program to
i) perform 50-100m scale fault stimulation experiments in basement rock at over 1’000m depth,
(ii) pre-condition the stress distribution on the target fault to perform real-time tests of different physical sources and forecasting hypotheses,
(iii) deploy data-driven approaches and real-time modeling to conduct structured prospective earthquake forecasting experiments, and
(iv) integrate and validate results from deep underground experiments, experimental rock-deformation laboratories, numerical physics and dynamic modeling, and observations from natural earthquakes.
i) perform 50-100m scale fault stimulation experiments in basement rock at over 1’000m depth,
(ii) pre-condition the stress distribution on the target fault to perform real-time tests of different physical sources and forecasting hypotheses,
(iii) deploy data-driven approaches and real-time modeling to conduct structured prospective earthquake forecasting experiments, and
(iv) integrate and validate results from deep underground experiments, experimental rock-deformation laboratories, numerical physics and dynamic modeling, and observations from natural earthquakes.