Seismically active volume not a good indicator of the stimulated reservoir: evidence from Paralana
Abstract
Hydraulic stimulation is routinely carried out in EGS projects in order to engineer the geothermal reservoir, triggering large amounts of seismicity. There is broad consensus that hydroshearing, the shear failure of existing fractures caused by pore pressure increasing above a critical value, is the mechanism primarily responsible for the seismicity. It is also often assumed to be the primary mechanism behind permeability enhancement and thus reservoir creation, although several other physical processes, such as thermal and chemical effects, and tensile failure, can also contribute. An extension of this assumption is that the seismically active volume reflects the extent of the stimulated reservoir. This may not always be the case. Hydraulic stimulation aims to improve injectivity, which we define as the ratio of injection rate to wellhead pressure. During the Paralana-2 EGS stimulation, undertaken in South Australia in 2011, injectivity increased linearly over time, implying permeability enhancement. Concurrently, 4754 micro-earthquakes were detected and located. We have spatially averaged hypocentre positions in order to image the spatio-temporally evolving seismicity cloud and its main features. Using an established correlation from a previous EGS experiment, we use hypocentre density as a proxy measure of pore pressure increase. Knowing then how pressure and injectivity evolve over time, we develop an inversion procedure based on reservoir simulation to infer how permeability has evolved around the well. Our results indicate that, although there is significant permeability enhancement close to the wellbore, this is largely uncoupled from the seismically active volume that extends much further away. In other words, for the Paralana stimulation, the seismicity cloud is a poor indicator of the stimulated reservoir volume. It further implies that hydroshearing itself is not the main cause of permeability enhancement at Paralana. Our study suggests that past stimulations may need to be revisited and a stronger case made for estimates of the stimulated reservoir.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2017
- Bibcode:
- 2017AGUFM.H42D..06R
- Keywords:
-
- 1847 Modeling;
- HYDROLOGY;
- 1855 Remote sensing;
- HYDROLOGY;
- 1859 Rocks: physical properties;
- HYDROLOGY;
- 1873 Uncertainty assessment;
- HYDROLOGY