DFOS based fault creep monitoring - insights from an underground mine experiment

In tectonic active regions, fault creep induced ground motions represent a hazard that may put at risk infrastructures (railroad, highway, bridges etc.), buildings and industrial constructions. Proper monitoring in this context is not only a prerequisite for risk assessment but is also of value to provide insights into the understanding of fault loading process which helps to constrain seismic hazard. Today, monitoring of fault creep is widely done by means of local in-situ (GPS, extensometers) or broad scale remote sensing (INSAR ect.) measurement techniques which lack either in spatial continuity and range or temporal resolution. Given its quasi-continuity at kilometer scale in space and time, Distributed Fiber Optic Sensing (DFOS) monitoring techniques may represent a promising complementary tool in this respect. Here we provide insights on the monitoring potential using Distributed Strain Sensing (DSS) from an in-situ fault monitoring experiment in a deep underground mine in Sweden. At the so-called Garpenberg mine, seismicity is associated with long-term occurrences (several months to years) of seismic repeaters and multiplets documenting repetitive fault failure in specific zones. Comparison to in-situ strain measurement shows that this repetitive seismic signature is widely driven by aseismic creep of the rockmass following the excavation of stopes (volumes of ~ 27000 m³). Further investigations confirmed that rockmass readjustment and stress redistribution following excavation is dominated by aseismic creep which itself may (but not always) trigger seismicity. DSS monitoring has been applied together with Distributed Acoustic Sensing (DAS) and other fiber optic technologies in order to monitor the full seismic cycle of certain repeater/multiplet targets. DSS allowed detecting multiple active fault structures and associated creeping sequences either triggered from excavation progress and/or self-triggered from interactive loading processes. In addition, reliable first order approximation of fault slip (displacement) could be derived from the recorded strain using a simplified shear zone geometry model. Next to these promising results, currently, the potential of DSS is further explored in a real fault creep monitoring scenario at an outcropping actively creeping fault structure.