Determining reservoir intervals in the Bowland Shale using petrophysics and rock physics models

Abstract

An evaluation of prospective shale gas reservoir intervals in the Bowland Shale is presented using a wireline log dataset from the UK's first shale gas exploration well. Accurate identification of such intervals is crucial in determining ideal landing zones for drilling horizontal production wells, but the task is challenging due to the heterogeneous nature of mudrocks. This heterogeneity leads to stratigraphic variations in reservoir quality and mechanical properties, and leads to complex geophysical behaviour, including seismic anisotropy. We generate petrophysical logs such as mineralogy, porosity, and organic content, and calibrate these to the results of core studies. If ‘reservoir quality’ is defined by combined cut-offs relating to these parameters, we find that over 100 m of reservoir quality shale is present in the well, located primarily within the upper section. To examine the link between geophysical signature and rock properties, an isotropic rock physics model is developed, using effective medium theories, to recreate the elastic properties of the shale and produce forward-looking templates for subsequent seismic inversion studies. We find that the mineralogical heterogeneity in the shale has a profound impact on modelled elastic properties, obscuring more discrete changes due to porosity, organic content and water saturation, and that the best reservoir quality intervals of the shale bear a distinctive response on rock physics cross-plots. Finally, we consider the density of natural fractures in the shale by developing an anisotropic rock physics model to reflect high-angle fractures observed on micro-imagery logs. We invert crack density using shear wave splitting well log data and find a crack density of up to 4% which correlates well with micro-imagery observations. Our work further supports previous authors’ core-based studies in concluding that the Bowland Shale holds good reservoir characteristics, and we propose that there are multiple intervals within the shale that could be targeted with stacked horizontal wells, should those intervals’ mechanical properties also be suitable and there be adequate stress barriers between to restrict vertical hydraulic fracture growth. Finally, our rock physics templates may provide useful tools in interpreting pre-stack seismic datasets in prospective areas of the Bowland Shale and picking the best locations for drilling wells.