Logo Bedretto Lab and logo ETH Zurich

2021-05-07

Detective work on earthquakes - with updates

Detective work on earthquakes - with updates

As part of the ongoing Bedretto Reservoir Project, high-resolution seismic networks have been installed along the Bedretto main tunnel and in dedicated boreholes to monitor the behaviour of the rock volume and the possible occurrence of seismicity.

The water injections performed in the past two weeks by Geo Energie Suisse (GES) induced numerous very small earthquakes in the vicinity of the borehole – an expected and desired effect of the stimulations. The last stimulation activities took place three days ago, and the induced seismicity near the injection borehole has since decreased and ceased.

The seismic network allows to detect events in the whole tunnel area, and on the night of 6 May 2021, a couple of events have been registered at about 500 meters away from the injection area, in the direction of the entrance of the Bedretto tunnel. The largest event had a magnitude of +0.1, according to the Swiss Seismological Service at ETH Zurich. Therewith, this event is considerably lower than the magnitude threshold above which people can usually notice an earthquake. Only the very dense seismic monitoring system made it possible to record and locate such small events. A full analysis of the data recorded over the past weeks enabled us to detect so far nine events with magnitudes down to ML= -1.8, some of which occurring already several months before the start of the water injections.

Although these events are too small to raise any immediate safety concerns, they provide a unique possibility to understand how pressure perturbations connected to activities in the reservoir and to water flow propagate through the rock volume. The BedrettoLab team has initiated a range of investigations to gain a better understanding of the causes of these unusual events and will use this new knowledge to steer future stimulation activities. 7 May 2021

Update 12 May 2021: Detective work continues

The Bedretto team continues to investigate the micro-earthquakes observed first on May 6th. In the past five days, three events with a maximum magnitude of -0.5 were observed, plus one event at more than 1 km from the tunnel. Using additional reprocessing steps, the team was able to reconstruct the timeline of the earthquakes of the past two years in more detail. Seismologists use an approach called “template matching”: Once a signal is known, they can scan past data for similar patterns that could not be detected before. This analysis revealed that another seismicity episode occurred unknown to the Bedretto team in October 2020 a few hundred meters further into the tunnel as compared to the May 6th events, with a maximum magnitude of about 0.2. However, as early as March 2020, another episode of very small events of magnitude -1.5 and smaller was detected.

So far, the events in March and October 2020 seem to have no obvious connection to drilling or stimulation activities in the Lab. The events in October followed within a few days of extreme rainfall at Bedretto, so it is possible that the seismicity is related to changes in pore pressures induced by rainfall, but other sequences do not so clearly correlate. Rain induced earthquakes have been observed in other parts of Switzerland (see for example here), mainly in Karst areas where water from the surface can quickly reach seismogenic depths of 1 kilometre or more.

The detective work therefore continues: The Bedretto team is installing additional seismic stations, gathering fluid samples and is re-analysing data, and they are considering hydro-mechanic models to build plausible models for the micro-earthquakes. At the same time, activities in the tunnel continue with microdrilling, notching and stress measurements in one of the long boreholes; activities that do not result in any micro-seismicity. Injections are not planned for the next days. 

Update 27 May 2021: Clues found, investigations continue

The detective work on the cause of the observed micro-earthquakes has produced some clues, but further analysis is necessary to better understand the seismicity around the BedrettoLab.

One of the clues is that much of the observed seismicity occurred during a period of heavy rainfalls in the Bedretto region. Rainfall-driven microearthquake triggering is a plausible but somewhat debated scientific hypothesis, and our observations form an interesting test case for the theory. We are currently collecting regional precipitation and meltwater data, to do an in-depth study, with recently developed microearthquake detection algorithms.

Another clue is that microearthquake sequences are not at all uncommon in underground operations such as the ones performed in the BedrettoLab. Activities like drilling and tunnelling can change crustal stresses and pore pressures in the host rock, even at some distance from the tunnel itself. Such perturbations are also likely candidates to drive the observed microquake activity.

The Bedretto team is currently expanding the seismic monitoring network in order to be able to better characterize the microseismic events, not just in the immediate vicinity of the laboratory, but also across a larger region around it.

The reservoir stimulation activities are completed for the moment, which will allow the reservoir system to equilibrate. With the expanded monitoring network, we will be able to even better study the microseismic activity and hunt for further clues. With the additional knowledge about the background seismicity in Bedretto, the team is preparing next stimulations starting in autumn.