Directional petrological characterisation of deep-marine sandstones using grain fabric and permeability anisotropy: Methodologies, theory, application and suggestions for integration

Different methods for quantifying preferred grain orientation (or grain shape fabric) in sandstones are reviewed and their value for the reconstruction of palaeocurrent direction and preferred direction of pore fluid flow is reassessed critically. Based on existing knowledge from literature and new grain fabric data, it is shown that petrographic analysis has distinct merits, because grain fabric can be determined with low equipment costs and population statistics are based on the orientation of individual grains. Several recommendations are given how to standardise procedures for acquisition of grain fabric through microscopy. These recommendations include: (1) the careful selection of cross-section orientations to acquire apparent long grain axes that are representative of true long grain axes; (2) the preferred measurement of the coarsest grains in cross-section; and (3) a new statistically founded method to select a subset of grains that is representative of the total population of grains in a sandstone sample. Magnetic techniques that use anisotropy of magnetic susceptibility (AMS) as a proxy for grain orientation, require specialised equipment and are more restricted in statistical evaluation, but these methods benefit greatly from being able to measure a large number of grains in a three-dimensional space, in a short amount of time and with a lower sensitivity to user bias. Of the two magnetic techniques discussed in this paper, the new enhanced AMS method, in which rock magnetic properties are enhanced by introducing magnetite into the pores and as precipitates on grain surfaces, is regarded to be an improvement over the traditional natural AMS method, in which only the natural assemblage of magnetite particles present in the rock is analysed. Enhanced AMS measures the preferred orientation of essentially all grains in a rock sample, thus providing the shape fabric of a hydrodynamically more accurate representation of framework grains than natural AMS. Through a small technical modification, the enhanced AMS method can be used to obtain a magnetic cast of the pore network of a rock sample, so that the resulting AMS is controlled by the shape anisotropy of the individual pores. It is commonly assumed that long axes of sand grains in sedimentary deposits parallel the direction of the flows that formed the deposits. This concept is challenged by a new systematic study of grain fabric in deep-marine turbidites. It indeed appears that grains in ripple cross-laminated divisions of turbidites dominantly attain a preferred orientation parallel to the palaeoflow direction, but grains in plane-parallel-laminated divisions have an additional flow-perpendicular mode, which suggests a contribution of rolling grain transport. Three different fabric modes occur in massive structureless sandstones and massive divisions in turbidites. Apart from flow-aligned and flow-perpendicular modes, there is a statistically significant, but hydrodynamically enigmatic, flow-oblique mode with an average azimuthal deviation of 40 degrees. This large variation of fabric types suggests that a similar variation exists in preferred paths of pore fluid flow within turbidites, controlled to at least some degree by depositional process. This is confirmed by independent laboratory measurements of permeability anisotropy on samples for which grain fabric data are available, and thus defines grain fabric analysis as a viable alternative to traditional permeametry techniques. (C) 2007 Elsevier B.V. All rights reserved.