GNGTS 2019 - Atti del 38° Convegno Nazionale

584 GNGTS 2019 S essione 3.1 Because of the above mentioned factors afflicting the non-uniqueness of the tomographic solution, a strong integration with all available well-information in the estimation of the velocity model was mandatory as well. In particular, well tops markers data are generally another valuable constraint for the final updated model. Indeed, at the final stage of the velocity model building process, a statistical analysis of the mis-ties between the seismic imaged depths and the well-tops of the main regional horizons was done in order to scale the seismic velocity model and adjust the final depth image. In extensively explored area like the NorthAdriatic Sea, a significant and homogeneous distribution of more than 90 wells out of the 300 with velocity measurement was selected for this purpose. Figure 3 (a) represents a cluster of 28 wells that have been used for the calibration of the top Pliocene regional horizon in one of the project sub-areas. The majority of the well is located along the trend of the overthrust belt in order to better constrain the velocity model in this complex geological area. Figure 3 (c) and (d) show the histograms of mis-ties between the wells markers depths and the relative seismic image. It can be observed that mis-tie mean value decreased from -23 m (depth image shallower on average than the top marker) to -0.23 m, which represents a 99% reduction of mean depthing errors at wells locations. The resulting velocity scaling factor distribution to correct for the depth positioning error of the top Pliocene was homogenous, showing also a good correlation with the underlying geological setting, meaning that the previous tomographic iterations were stable and did not introduce unintended velocity anomalies. The velocity needed to be decreased of about -3% (blue color represents negative scaling values) in the whole foreland and increased (red color represents positive scaling value) in the foredeep region (Fig. 3-b). The final calibrated velocity model was then used to produce the final PSDM volume. The full-fold migration area involved 10 amplitude and phase matched surveys, covering an area of approximately 8000 km 2 with a maximum depth of 12 km. Results and conclusions. The final volume exhibits good image quality, with a good correlation with the key formation top data, good reflectors continuity and an improved S/N ratio if compared with the vintage Post-stack time migration volume. Foredeep thrusts structures with their complex geometries are now well-imaged (fig.1), pull-down effects are fully resolved with the layer geometries properly restored, resulting in a more reliable technical and economic assessment of new prospects and producing fields (fig.2). Despite the significant size of the processing area, high-performance computing capabilities of the Eni Green Data Center, allowed to complete the project in a very competitive time-frame. As part of the multi- disciplinary integrated G&G workflow, several studies have been carried out during and after Fig. 2 - Comparison between the vintage PoSTM volume and the new Kirchhoff PSDM with super-imposed the updated velocity model. The pull-down effects due to the presence of gas on the PoSTM volume have been fully recovered in the depth image and continuity and the geometries of the mineralized layers restored.