Deformation bands and their significance in porous sandstone reservoirs
Geologists and engineers in the oil industry are becoming increasingly aware of the importance of small-scale (sub seismic) faults and fractures in many reservoirs, but modelling of such structures is currently hampered by our limited understanding of these structures. The many fault parameters that, directly or indirectly, should be input to advanced reservoir simulation models include the number, distribution, orientation, geometry, displacement, length, and petrophysical properties of small-scale faults and deformation bands. In porous sandstone reservoirs, which are the focus of the current contribution, deformation bands with displacements less than a few tens of centimetres are generally developed instead of slickenlined fault surfaces typical for less porous lithologies. Deformation bands may show reduction in permeability by up to three orders of magnitude (Antonellini & Aydin 1994). This parameter needs to be estimated for each field or reservoir level, as it will depend on degree of lithification during deformation (for example, increasing depth during faulting implies increased influence of cataclasis and therefore stronger reduction of permeability across the deformation bands). Although of less importance at the exploration stage (Harper & Lundin 1997), deformation bands that significantly reduce permeability may be very important with respect to production from sandstone reservoirs. Displacement, as an isolated parameter, is of less importance, as it generally is too small for deformation bands to cause sealing by juxtaposition of sand and shale layers. Field studies have shown that deformation bands generally can be modelled as straight structures that are particularly numerous around larger faults, but may also occur as scattered structures between seismically resolvable faults (e.g. Antonellini & Aydin 1995). Much work is left to explore the relation between the distribution of micro faults and larger structures. However, the question of distribution must always be addressed in each case, for example by use of attribute maps and core data (e.g. Hesthammer & Fossen 1997). In this work we will discuss the length vs. displacement of small-scale structures (deformation bands) in porous sandstones. It has been suggested that length and displacement can be scaled down from the size range covered by seismic data to the sub seismic domain (e.g. Yielding et al. 1992). The underlying assumption is that displacement and length are related through a power law and thus are fractal by nature. Down-scaling of fault length and displacement has already been applied as input to sandstone reservoir simulation models (e.g. Gauthier & Lake 1993). We here present new results from a population of deformation bands and discuss why properties of the seismically mappable part of a fault population cannot necessarily be extrapolated far into the sub seismic domain.