Detection of Reservoir Sorting Characteristics from Seismic Data

It is usually the case that rock permeabilities do not correlate especially well with elastic properties, so seismic based methods for estimating permeability are not conventionally viewed with much optimism. However, there are certain kinds of permeability variation that are driven chiefly by variations in sorting, and these sorting or grain-size distribution variations have a much more direct impact on elastic properties. We present an overview of modeling work applied to some deepwater field data where the turbidite reservoir facies has substantial grain size polydispersity, and this is shown to have significant seismic and permeability expressions. The effect of poor sorting on elastic properties is well captured by a simple model that takes into account the effect of non-load bearing (or 'floating') grains in the rock fabric. This model has ample corroboration from well logs, core permeability measurements, experimental laser-grain studies and numerical rock-assembly models. It is implemented as part of an open source stochastic seismic inversion. The statistical outputs of the inversion show that the sorting parameter is detectable in the field data.