Modelling of 3D physical property intra-unit heterogeneity of the UK East Midlands for the BGS PropBase project
Kingdon, Andrew; Williams, John D.O.; Williamson, J. Paul. 2010 Modelling of 3D physical property intra-unit heterogeneity of the UK East Midlands for the BGS PropBase project. [Lecture] In: 2010 Geological Society of America Annual Meeting, Colorado, USA, 31 Oct – 3 Nov 2010. (Unpublished)Before downloading, please read NORA policies.
3D geological models have replaced maps as BGS’s primary output. They show the intersection of geology and the surface plus subsurface geometry constrained by wells and geophysics. They give a simplified view of defined lithostratigraphic units attributed only with names and enveloped by surfaces. Properties are implicit through descriptions; internal variability is largely ignored. PropBase populates 3D models by extracting physical properties from BGS databanks and mapping their variation through 3D volumes with no seismic framework. We present a populated grid of the Triassic Sherwood Sandstone Group (SSG) near Nottingham, the principal UK carbon storage reservoir candidate. A 45 X 65 km test model, built in GOCAD® from surface linework and stratigraphy from 578 wells, was selected because of its relatively simple geology; Carboniferous, Permian and Triassic rocks outcrop progressively eastwards and dip ~1° east. Wireline logs are the sole pervasive property data- but, as they target the underlying Coal Measures, few sample the entire SSG. 58 logs were used to calculate porosity proxies from both sonic (Raymer-Hunt-Gardner method) and density data. A proportional 3D stratigraphic grid (Sgrid) was generated with 200 x 200 x 2 m cells. Log-derived porosities were upscaled as points into the Sgrid; all cells cut by a log were attributed with the mean porosity. Data outliers (spikes; high clay content) were omitted as artefacts before interpolation, preventing distorted results in areas of poor data control. Values from both density and sonic data were initialised across separate Sgrids using Discrete Smooth Interpolation, so each cell contained a porosity attribute. Two main features emerged: • a high porosity zone close to the SSG outcrop caused by surface water flushing • porosity decreasing eastward with depth, due to compaction A zone of higher porosity between two faults in the model’s east is less clear, especially on the sonic grid, and poorly constrained, but may indicate fault-induced fractures. Ever greater environmental and resource pressure on the geosphere requires ever more realistic subsurface simulation. Limited 2D and scarce 3D seismic data and a lack of suitable property data hinders upscaling across much of the UK. Though imperfect, the SSG grid makes best use of available data in Nottinghamshire.
|Item Type:||Conference or Workshop Item (Lecture)|
|Programmes:||BGS Programmes 2010 > Geoscience Technologies|
|NORA Subject Terms:||Earth Sciences
Data and Information
|Date made live:||13 May 2011 14:00|
Actions (login required)