Seismic anisotropy in siliciclastic reservoir rocks

The interpretation of geophysical field measurements of seismic anisotropy is presently limited by our knowledge of the controls of the elastic anisotropy of sedimentary rocks in the subsurface. Traditionally, laboratory ultrasonic velocity measurements have been used to provide important information on bulk aggregate seismic anisotropy, however, they do not allow the discrimination of the contribution from the various microstructural parameters (e. g., crystallographic lattice preferred orientation (LPO), preferentially aligned porosity, aligned fractures and the non-random spatial distribution of mineral phases). In this study the results from scanning electron microscope-electron backscattered diffraction (SENI-EBSD), quantitative X-ray diffraction (QXRD), image analysis, ultrasonic velocity measurements, palaeomagnetism, anisotropic magnetic susceptibility, and numerical modelling are combined to elucidate the controls of the elastic anisotropy of siliciclastic sedimentary rocks from an oil reservoir. SEM-EBSD was used to measure both the overall and individual constituent mineral phase LPO (Maddock et al. 2004). As phyllosilicates are both very fine-grained, with a high aspect-ratio and low crystallinity, their LPO contribution was established via a combination of image analysis and numerical modelling (Bingham approximation). These analytical and predictive methods for determining phyllosilicate fabric intensity produced consistent results. For the first time, the azimuthally preferred orientation of elongate grains within sedimentary rocks was determined using anisotropic magnetic susceptibility of ferrous minerals and were compared to those predictions obtained using EBSD. The strength of the fabric-texture (J), as determined by EBSD, is proportional to the maximum compressional and shear-wave anisotropy, as calculated from the Christoffel equation, by taking a Hill average of the bulk aggregate elastic constants. The quartz and feldspar velocity maxima aligned in a constructive fashion throughout most of the samples. It is possible that the preferred alignment of crystals detected by EBSD reflects the palaeoflow direction. The predicted symmetries of velocity anisotropy ranged from orthorhombic in the phyllosilicate-free, well-sorted, mature sandstones to strong vertical transverse isotropy in the unfractured phyllosilicate-rich mudstones. Vertical transverse isotropy is predicted to be oriented, such that, the plane of azimuthal isotropy is aligned parallel to bedding i. e., parallel to the horizontally aligned clays and micas. Similarly, orthorhombic symmetry is predicted to be oriented, such that, one plane of symmetry is aligned approximately parallel to bedding whilst the other symmetry plane is aligned parallel to the single most dominant fracture set. The results from this study provide the input needed for a general mathematical model for the reservoir allowing the prediction of seismic anisotropy for any rock in the reservoir given accurate modal proportions. The resulting model is an advance on the empirical correlations that are usually used to determine how seismic velocities are affected by factors such as clay content and porosity. In particular, the bulk aggregate elastic stiffness tensor obtained during this study can be integrated with high-pressure ultrasonic measurements to enable the prediction of the additional contribution from grain-scale effects such as shape-preferred orientations, and grain boundary compliances (Hall et al. 2007). The results from this study have also provided the basic data to allow field seismic data to be inverted to obtain estimates of in situ fracture density and orientation (Kendall et al. 2006). In summary, analysis of a suite of siliciclastic hydrocarbon reservoir rocks has shown that the LPO of constitutive minerals can offer information about the nature of a reservoir. The results suggest that seismic anisotropy is not only indicative of lithology but can also be an indicator of reservoir quality and palaeoflow direction.