Probability of fault reactivation in the Bavarian Molasse Basin

2019 
Abstract In the Bavarian Molasse Basin, especially in the greater Munich, geothermal exploration of the hydrothermal Upper Jurassic reservoir is rapidly expanding. Until now, little seismic reservoir response is observed, only at two out of 16 sites seismic events with M L > 2.0 were detected: at Unterhaching, in 2008, six events with M L > 2 occurred soon after the onset of circulation; at Poing, in 2016, 5 years after circulation started, two events with M L ∼2.1 occurred, both located near the injection well. The analysis of the reactivation potential allows to connect seismicity to fault structures. In the Bavarian Molasse Basin, fault structures generally exhibit low seismic reactivation potential, as long as they trend ENE-WSW. By Monte-Carlo simulation, the geological uncertainty and the sensitivity of the individual parameters are quantified. They show that critically pre-stressed fault segments, e.g. at Unterhaching, combined with minimum change of the hydraulic reservoir conditions can lead to a dramatic increase of the reactivation potential of seismicity. For uncritical fault segments, e.g. at the Poing site, two self-enforcing effects are discussed which increase the reactivation potential over time: first, stress field modification by thermo-hydraulic effects and, second, fault alteration by carbonate dissolution can reduce the fault friction and cohesion. Both effects increase the sensitivity of the reactivation potential to the fault friction and can bring previously uncritical fault segments to critical state. Finally, the possible impact of coupled thermo-hydro-mechanical and chemical processes at hydrothermal systems on the reactivation potential is highlighted: at fault segments with high reactivation potential hydro-mechanical effects may dominate whereas at low reactivation potential thermo-mechanical processes can potentially yield to a slow rotation of the stress field.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    34
    References
    10
    Citations
    NaN
    KQI
    []