Quantifying the variability in fault density across the UK Bowland Shale with implications for induced seismicity hazard

Abstract

To date, hydraulic fracturing for shale gas extraction has been used at three wells in the UK. In each case, the resulting microseismicity exceeded the UK’s red-light threshold of magnitude 0.5. The three wells all targeted the Bowland Shale Formation, and all were located within close proximity of each other on the Fylde Peninsula in west Lancashire. Observations of hydraulic fracturing-induced seismicity (HF-IS) elsewhere shows that the prevalence of induced seismicity is highly spatially variable. Hence, it is by no means clear whether hydraulic fracturing elsewhere in the Bowland Shale would be likely to generate seismicity at similar levels. In this study we examine the geological and geomechanical conditions across the Bowland Shale with respect to their potential controls on induced seismicity. The abundance of pre-existing faults is likely to play an important control on the generation of HF-IS. We use an automated fault detection algorithm to map faults within a selection of 3D reflection seismic datasets across the Bowland Shale play. For the identified faults, we compute the effective stresses acting on these structures in order to identify whether they are likely to be critically stressed. We find that the Bowland Shale within the Fylde Peninsula contains a significant number of critically stressed faults. However, there is significant variation in the density of critically stressed faults across the play, with up to an order of magnitude reduction in fault density from the west (i.e., the Fylde Peninsula) to the east. We use these observations to inform a seismic hazard model for proposed hydraulic fracturing in areas to the east of the Bowland Shale play. We find that the occurrence of felt seismic events cannot be precluded, however their likelihood of occurring is reduced.