A New Approach For The Estimation Of A Laterally Variable Anisotropic Velocity Model

A case history of anisotropic velocity model building on a West Africa (offshore Angola) dataset is proposed. The goal of the project was to improve depth imaging quality, in particular in terms of accuracy of event placement and fault positioning, and also to estimate a reliable velocity model for pore pressure prediction. Seismic data provided evidences of the presence of anisotropy in the area. In order to obtain a robust and reliable estimation of Thomsen’s anisotropic parameters, we applied two independent methods in parallel: the first one, used as reference, is based on the standard time-domain approach [1] whilst the second one works in depth domain and applies the migration velocity analysis principles [2]. The two approaches resulted in coherent estimates of Thomsen’s parameters but the depth domain approach is preferable because of its integration in the imaging workflow: the definition of the vertical velocity (and, therefore, of Thomsen’s delta parameter) at well locations is univocal thanks to the exploitation of depth stratigraphical markers. The introduction of anisotropy in the model, tightly coupled with an accurate velocity analysis, not only allowed obtaining the correct depth positioning but also a remarkable improvement in imaging quality with respect to both PSTM and the initial isotropic PSDM.