Investigation of azimuthal anisotropy in high-fold 3D multicomponent seismic reflection

Abstract

3D P-wave seismic surveys can exhibit significant azimuthal variation in stacking velocities, and failure to allow for such variations can introduce smearing into the stacked volume. The problem is likely to be worse in the case of converted-wave (PS) reflection. Firstly, S-waves have lower velocities such that time variations are amplified. Secondly, PS rays are asymmetrical, such that anomalous features may be traversed by different wavetypes (P or S), depending on the direction of travel. Recently, a coal-scale 3D-PS trial was recorded in the Bowen Basin with the aim of providing a detailed investigation of such azimuthal variation. The survey was designed with extremely high fold (>500). This allowed good quality images to be constructed for data subsets having restricted ray azimuths. Target structures interpreted using different ray-azimuths exhibit significant timing variations (up to 30ms). The observed azimuthal variations may not necessarily indicate true azimuthal anisotropy. They can result from poor statics solutions, and PS imagery is notorious for difficult statics. On the other hand, the observed variations may be indicative of true geological anisotropy. Based on shear-wave splitting models, our PS velocity variations have been modelled in terms of elliptical variation with azimuth, and this approach predicts the orientation of the horizontal stress field. It is interesting that a majority of our data zones indicate a consistent stress orientation. Furthermore, the interpreted horizontal-stress orientation is consistent with observed reverse faulting in the area.