The importance of characterizing uncertainty in controversial geoscience applications: induced seismicity associated with hydraulic fracturing for shale gas in northwest England

Abstract

Westaway, R. (2016) The importance of characterizing uncertainty in controversial geoscience applications: induced seismicity associated with hydraulic fracturing for shale gas in northwest England. Proceedings of the Geologists' Association, 127(1), pp. 1-17.There may be differences between this version and the published version. You are advised to consult the publisher's version if you wish to cite from it.http://eprints.gla.ac.uk/119683/ AbstractFracking of the Preese Hall-1 well in 2011 induced microseismicity that was strong enough to be felt.This occurrence of 'nuisance' microearthquakes, unexpected at the time, was a major factor resulting in adverse public opinion against shale gas in the UK and was thus of significant political importance. Despite this, and notwithstanding the technical importance of this instance of induced seismicity for informing future shale gas projects, it has received little integrated study;contradictory results have indeed been reported in analyses that lack integration. This instance therefore provides a case study to illustrate how a small but significant multi-disciplinary geoscience dataset may be put to best use, including how best to quantify uncertainties in key parameters, which may themselves be relatively poorly quantified but whose values may significantly affect the ability to understand the occurrence of induced seismicity. The best-recorded event in this induced microearthquake sequence (at 08:12 on 2 August 2011) is thus assigned an epicentre circa British National Grid reference SD 377 358, south of the Preese Hall-1 wellhead, a focal depth of ~2.5 km, and a focal mechanism with strike 030°, dip 75°, and rake -20°, this NNE-striking nodal plane being the inferred fault plane. Like other parts of Britain, this locality exhibits high differential stress, with maximum and minimum principal stresses roughly north-south and east-west. This instance indeed fits an emerging trend of the occurrence of relatively large induced earthquakes in localities with high differential stress; such an association was predicted many years ago on the basis of experimental rock mechanics data, so observational confirmation might well have been anticipated and should thus not have been unexpected. Many steep faults, striking NNE-SSW or NE-SW, mostlyCarboniferous-age normal faults, are present; the stress field is favourably oriented for their leftlateral reactivation, southward leakage of fracking fluid into one such fault having presumably caused the induced seismicity. Given the pervasive presence of similarly-oriented faults, future occurrences of similar induced seismicity should be planned for; they pose a significant technical challenge to any future UK shale-gas industry.

Preese Hall-1 bedding dip

“…∼S70 • E. This orientation of the wellbore is consistent with published depictions of the well track (such as those by Clarke et al, 2014, andWestaway, 2016a) and confirms that Harper (2011) interpreted this imaging tool dataset correctly. The version reported by de Pater and Baisch (2011) is likewise correct: according to this dataset the bedding dips at ∼30 • towards an azimuth of ∼290 • or ∼N70 • W. Publication of this excerpt from the imaging tool dataset therefore does not add anything to the information already available, as Smythe (2016) has claimed; it confirms the existing interpretation of these data.…”

Some additional thoughts on Preese Hall

“…Second, the geological structure at Preese Hall-1 and neighbouring boreholes, presented in Fig. 1 of Clarke (2016), differs dramatically from that which has featured in all previous literature on this topic, including the illustrations in my own recent outputs (Westaway, 2015(Westaway, , 2016a(Westaway, , 2016b. He now reports that the Emstites leion (Cravenoceras leion, or E1a1) marine band, which defines the boundary between the Visean and Namurian stages of the Carboniferous, at ∼2500 m depth (MD) in the Preese Hall-1 borehole.…”

“…Third, the excerpt from the 3-D seismic section that was originally published by Clarke et al (2014) and was re-published by Westaway (2016aWestaway ( , 2016b) with stratigraphic labelling, shows the component of section-parallel bedding dip changing downward from westward, across a zone of deformed bedding, to eastward in the deepest ∼200 m of the Preese Hall-1 well. In contrast, measurements made from the borehole image log, reported by Harper (2011), indicate that the bedding in this depth interval dips WNW at ∼30-40 • , steepening downward to ∼70-80 • .…”

“…One reason why this wellbore deformation has not hitherto been associated with the induced seismicity was that Clarke et al (2014) located the seismicity so far from the wellbore. This argument was superseded by the realisation of the mistakes in their paper and the resulting adjustment of the position of the seismogenic fault much closer to the wellbore (Westaway, 2016a). Another reason has been because de Pater and Baisch (2011) were unable to identify a clear con-…”

Repurposing of disused shale gas wells for subsurface heat storage: preliminary analysis concerning UK issues

“…This depiction, which is not to horizontal scale, has been annotated to indicate the meaning of the parameter L used in the calculations (equation (12)). It is noted in 25 passing that, while the present manuscript was in preparation, it became apparent that considerable uncertainty exists regarding the published depiction of stratigraphic and structural data pertaining to the Preese Hall-1 well (Westaway 2016a). Furthermore, while the manuscript was under review, new information pertaining to these issues was placed in the public domain by Clarke (2016) and Kingdon (2016) as part of a wider online discussion.…”