VSP Quantitative Diagnosis of Seismic Velocity Model for 3D VSP Imaging

Velocity model building in the Gulf of Mexico is very challenging due to complex structures and salt resulting from poorly illuminated and noisy data. The predominately layered sand and shale geology introduces anisotropic effects which need to be included in the model building. Many combinations of velocities and anisotropy parameters can yield flat common image gathers after depth migration, and the ‘flat criteria’ is a non-unique solution assessing the quality of a velocity model. Due to uncertainties in the velocity models, we had poorly illuminated surface seismic images, especially in the sub-salt regions. Thus, 3D VSP are introduced intending to provide high-frequency/resolution images to complement the surface seismic. In this paper, we used multifarious VSP data in three wells to quantify the quality of surface seismic velocity models. Examples in the Gulf of Mexico are discussed where multiple surface seismic velocity models have been built, and several VSP data were used to quantify differences in those velocity models. The distributions of multi-VSP sources and multi-receivers also allow us to diagnose azimuth variations of the seismic velocity models. Our post-survey modelling practices have proven that multifarious VSP diagnosis method is a useful tool to guide the seismic velocity model updating.