204 Seismic Imaging terms of geometry and mechanical properties. In the first field example, the hybrid seismic imaging tool showed that seismic data derived from traditional refraction acquisition is valuable for obtaining information about the reflectivity for targets located in the near and/or very near surface. After first break pickings, a P-wave tomography inversion was performed to obtain a depth velocity model. The processing of reflection events, present in the seismic refraction survey, was carried out by a standard procedure. However, particular attention was focused on isolating the reflection waves. Finally, after a careful adjustment of the results obtained from the processing of refraction and reflection waves, they were gathered to produce an extended time reflectivity section starting from the surface. In the second example, a refraction tomography algorithm has been applied to the first-arrival times picked manually on the shot gathers collected in a hydrothermal area. A P-wave velocity model was obtained that presented values in the range 100 – 2,000 m/s and a low velocity layer at the surface with thickness around 5 m. The processing of surface waves, extracted from the seismic survey, was performed in the f-k domain with SWIP, an open-source MATLAB-based package. The inversion of the dispersion curves produced a set of 1D models of S-wave velocity with an estimated depth of investigation of around 10 m. The final result was a pseudo 2D model obtained by merging all of the best fitting 1D models. The final S-wave velocity model showed strong lateral variations that were not visible on the P-wave velocity model, probably due to strong saturation variations. This information was used to estimate the Poisson’s ratio. The distribution of this parameter, more particularly its contrasts, clearly highlights gas pathways in the subsurface that are consistent with the degassing observed at the surface. Thus, these positive results open up new perspectives for several applications of more hybrid seismic methods. Finally, chapter 7 presents a field study at a site that has been extensively studied by the French national radioactive waste management agency (Andra). We showed how the integration of seismic data (3D survey and VSP), logging data (acoustic logging), and core measurements, combined with a succession of specific and advanced processing techniques, enabled the development of a 3D high-resolution geological model in depth. We demonstrated the benefit of geostatistical processing, both to quantify the quality of the seismic amplitude (SQI) and to perform depth conversion (Bayesian Kriging). The Q factor of geological formations can be obtained from VSP data. We used the fact that attenuation introduces dissipative dispersion, which can be measured from the frequency-dependent phase velocity of the VSP down-going wave. The methodology has been extended from well data to surface seismic data. For this purpose, a high-resolution velocity model is required. This was obtained from elastic inversion of the seismic data and by conversion of acoustic impedance Ip in velocity Vp. The procedure can be used to build a geomodel in depth defined by mechanical and hydro-geological parameters: velocities (Vp, Vs), density, Q factor, porosity, specific surface, and index of permeability. The seismic procedure was extended to compute a seismic pseudo gamma ray (GR-Seis). A high porous layer was detected at the top of the Dogger and lateral variation of the shale content can be seen in the Callovo-Oxfordian.
RkJQdWJsaXNoZXIy NjA3NzQ=