We predict anisotropic velocity perturbations around a salt dome on the edge of the Sigsbee escarpment in the Gulf of Mexico, using a combination of 3D geomechanical modelling and the application of rock-physics predictions. The modelling predicts an increase in compressive stress in directions perpendicular to the salt face compared to a background model without salt. This stress increase causes an increase in P-wave velocity in the direction perpendicular to the salt wall. Similarly, parallel to the salt wall a decrease in compressive stress is predicted, compared to a salt-less background model. The stress decrease causes a decrease in P-wave velocity in directions parallel to the salt wall. A good match is observed between these predictions andobserved azimuthal seismic velocity variations from a multi-azimuth seismic survey. This firmly establishes a geomechanical cause for anisotropic velocity perturbations around salt.
The work of this paper contributes to the general topic of improving the seismic imaging of complex subsurface structures. Combining improved subsurface velocity models and more accurate migration algorithms, new wide-azimuth seismic acquisition schemes aim at improving our ability to image steeply dipping and overhanging salt-bodies, as well as sub-salt reflectors. In this context, these imaging applications have benefited greatly from the application of wide azimuthal raycoverage acquisition patterns and processing the data with anisotropic velocity models (Zdraveva et al., 2012). Nonetheless, given that anisotropic velocity models are parameterized with large number of variables (compared to the isotropic case), the task of attaining all parameters involved becomes much more complex. Given the inverse nature of the imaging process, the analysis of the seismic gathers may prove insufficient in some cases to efficiently reconstruct all parameters describing the subsurface velocity field.
This problem was addressed by the work of Sengupta, et al. (2008, 2009) and Bachrach, et al. (2008), where it was proposed to employ geomechanical modeling in deriving updates to subsurface velocity models in the vicinity of salt. This paper examines the application of these concepts to real field data, where apparent azimuthal anisotropy is observed surrounding a salt diapir. The aim is to provide a geomechanical basis aimed at explaining observed azimuthal velocity changes by numerically simulating the expected geomechanical conditions for subsurface sediments lying in the vicinity of salt bodies and the manner in which these would affect the velocity field. The eventual outcome of this technique should contribute to attaining improved velocity models that incorporate complex geological structures, and therefore, improve the quality of the seismic image. In this paper we show that geomechanically-derived orthorhombic velocity predictions are consistent with wide-azimuth anisotropic seismic data observations in the Gulf of Mexico.