In 2005 a 3D high density wide-azimuth seismic survey was shot over the Irharen area (south east of Timimoun, Algeria). This acquisition was designed to better image the structure but also to improve fracture characterization and mapping, recognized as key in the description of this tight sand reservoir. The use of the offset vector binning concept, introduced by Cary (1999) and Vermeer (2002), on dense wide azimuth datasets is well suited for preserving the azimuthal information. An azimuthal residual velocity analysis based on semblance optimisation by scanning isotropic and anisotropic components of a parabolic elliptical model is proposed and tested successfully. The significant imaging improvement achieved with this land data case demonstrates the accuracy of the extracted azimuthal parameters for a better post-PSTM azimuthal residual moveout. After this correction, Common Image Gathers are ready for an azimuthal analysis of amplitudes. Taking advantage of the good sampling in offset and azimuth directions, the present study investigates the potential of Fourier analysis for describing the azimuthal variations of seismic amplitude.
The description of fracturation is a key information to optimize tight or carbonate reservoirs production, as a consequence many studies have been undertaken to get this information from seismic. In particular, anisotropic models have been proposed to link the fracturation to seismic anisotropy, Hudson (1980,1981), Schoenberg et al. (1980, 1995). P wave velocity is then depending on fracture orientation, which may cause azimuthal variation in the seismic velocity field. On the other hand, coefficients of anisotropy may be estimated from azimuthal variations in reflected P wave amplitude (Rüger, 2001).
Wide azimuth seismic acquisitions, which provide a full azimuthal illumination of the structures for a more accurate image, may additionally offer information on fractures orientation through the azimuthal analysis of the seismic response.
To meet these objectives, it is essential that the 3D wideazimuth character of the data be preserved throughout the processing sequence in order to get full benefit from the use of true 3D algorithms. In this perspective, the use of the offset vector binning concept, introduced by Cary (1999) and Vermeer (2002), on dense wide azimuth datasets is advantageous as it preserves offset and azimuth information, even after migration. In the present paper, we show how the Common Offset Vector (COV) processing is used to point out azimuthal variations in the velocity field and analyze the azimuthal variations of the reflected amplitude at a given horizon.