Thursday 18th April, 2013
Lecturer: Mairéad Nic Bhloscaidh (University of Ulster)
Tea/Coffee from 7:00pm
Lecture commencing at 7:30pm
Venue: Geological Survey of Ireland, Beggars Bush, Haddington Road, Dublin 4
Technologies such as Hydraulic Fracturing or Carbon Capture and Storage, which involve the injection of fluid into the earth’s crust, have been associated with increased rates of seismicity. One issue of concern in assessing the risks associated with these technologies is the possible magnitude of the induced earthquakes; although most of the observed earthquakes are small, there are cases where fluid injection has been linked to the triggering of large, potentially damaging events, for example the 2011 M5.7 Oklahoma event, thought to be the result of injection of waste water from fracking. The energy released in an induced earthquake is not limited to the energy introduced into the crust by the injection. Rather, the injection acts as a trigger for the release of energy that is already stored. The magnitude of the induced earthquake depends on properties of the existing faults such as size, location, state of stress and frictional behaviour. As with tectonic earthquakes, most of this information is not known: it is currently difficult, therefore, to provide meaningful constraints on the maximum probable magnitude at an injection site. Improving the quality of hazard assessment from induced seismicity relies, therefore, not only on advances in modelling fluid induced stress changes, but also developing physics based, stochastic models to explore uncertainties associated with unknown details of existing fault networks.In this talk I review some of the research the UU Geophysics group has carried out in these areas, including insights from numerical modelling of injection induced stress changes and rupture propagation under heterogeneous conditions of fault stress and friction, and also statistical investigations of earthquake populations associated with stress triggering in tectonic environments. Finally, we describe work in progress to develop a model for induced seismicity, which includes time dependent stress changes from injection and anisotropic fluid flow on stochastically generated, complex fault networks, which evolve as information is accumulated from developing catalogues of induced seismicity. We include probabilistic models for stress triggering and earthquake magnitude, informed by both tectonic and induced earthquakes, and also numerical simulations. By comparison of these stochastically generated earthquake sequences with existing real data from injection sites, we hope to gain insight into both our understanding of the processes and the adequacy of the statistical methods currently in use, particularly with respect to assessing the likelihood of damaging earthquakes during fluid injection.