Publications

We demonstrate the success of combining wavelet shift tomography, full waveform inversion (FWI) and separated wavefield imaging (SWIM) for pre-stack depth migration (PSDM) velocity model building and imaging. The data under investigation were acquired in 2009 using dual-sensor cables over the Utsira High area offshore Norway, where thin target sands span hundreds of square kilometers. Shallow overburden velocity variations and structural complexity at the target lead to large uncertainty in estimating oil reserves. The aim of this study, a research collaboration between PGS and Lundin Norway, was to achieve better imaging of the target sands and more accurate depth predictions. Wavelet shift tomography, operating on reflection arrivals, was used to update the shallow overburden in a legacy PSDM velocity model. FWI updates based on matching modelled versus observed refraction arrivals were subsequently able to resolve high resolution velocity variations associated with channels, pockmarks and gas pockets in the shallow overburden. Additionally, gathers from SWIM were generated using up- and down-going separated wavefields uniquely provided by the dual-sensor data. The SWIM approach exploits the greater illumination of the near surface inherent in the multiple arrivals. SWIM image gathers were used to validate the longer wavelength features not seen by FWI. The additional illumination provided by SWIM is of particular importance since a shallow wedge covers large parts of the field, and the associated long wavelength velocity variations have a significant impact on the vertical position and structural shape of the target sands in respect to the oil water contact. The study illustrates how reflection, refraction, and multiple arrivals from dual-sensor data can contribute towards high-fidelity model building and imaging. Resolution of the complex shallow overburden leads to more accurate positioning and depth predictions for the reservoir, directly impacting estimation of reserves.