The DSeis Project: Drilling into Seismogenic zones of M2.0 to M5.5 earthquakes in South African gold mines

The DSeis project under ICDP consists of drilling in three mines; MK, TT and C4 mines. Common scientific targets among them are the stress state and the microstructure in the seismogenic zone. In addition to these targets, specific targets in individual mines are detailed below. A M5.5 earthquake occurred beneath the MK mine on 5 August 2014. The hypocenter of this event was 5km depth from the surface. In contrast to the normal faulting of induced earthquakes in mining horizons (<4km depth), the M5.5 event was a strike-slip one with an N-S striking, sub-vertical nodal plane along which aftershocks aligned. Aftershocks extend up to 3.5km depth. We established a drilling site at 2.8km depth in the mine, from where two boreholes 800m-long penetrate into the areas of high and low aftershock densities. Targets of these drilling are 1) to investigate a depth variation in the stress state from the normal faulting to the strike-slip one, 2) to know what controls the spatial variation in the aftershock activity, and 3) to explore a limit of deep life that might be trapped in Archean sediments. Our site in the TT mine is 50m under the hypocenter of a M3.2 earthquake which occurred on 28 January 2017 at 3.6km depth. Although aftershock activity recorded by the seismic network operated by the mine is low, the source fault looks to extend along or parallel to a pre-existing, N-S striking fault. Three boreholes go through the fault at the hypocenter and the northern and the southern margins of the fault to compare the stress states and the microfracture distributions. Further, monitoring of microseismicity down to M -4 and geochemistry is planned to evaluate how much is a ratio of microseismicity associated with creation of new fractures. In the C4 mine, there was the site of a previous project, in which the microseismicity monitoring and the stress measurement by the CCBO technique were carried out. A M2.8 earthquake occurred 1 year after the CCBO and its hypocenter was only 100m away from the CCBO site. Due to little mining activity in the source region, the stress state just after the M2.8 event should be preserved. We will measure the stress again. Damage zones that evolved quasi-statically were seen by the microseismicity monitoring. Drilling into these zones would provide a clue to see a difference between faults evolved dynamically and quasi-statically.