Prediction of Relative Permeability and Fast Wettability Assessment Using Digital Rock Physics: An Operational Study on a Reservoir Sandstone

With the anticipated decline in oil demand over the coming decades, fast appraisal of oil fields becomes crucial for energy companies. In this regard, accelerating petrophysical synthesis plays a significant role in the fast appraisal and development of oil fields. Digital rock physics (DRP) simulation offers a cost-effective and rapid approach to compute relative permeability curves for rock/fluid systems. It also presents an opportunity to provide reservoir engineers with additional data, thereby enhancing the quality of petrophysical input for their simulations. Previous research demonstrated the predictive capabilities of DRP simulation coupled with a wettability anchoring experiment for a mixed-wet (MW) Bentheimer formation (Regaieg et al., 2022). In this study, we applied TotalEnergies’ DRP simulation workflow, coupled with a wettability anchoring experiment, in an operational context on a reservoir sandstone sample. The study was conducted as a blind test prior to the special core analysis (SCAL) measurements. Initial images representing large volumes with low resolution (LR) were enhanced using Enhanced Super-Resolution Generative Adversarial Networks (ESRGAN) to obtain high-resolution (HR) images. Subsequently, a pore network was extracted, and a parallel pore network simulator was utilized for multiphase flow simulations, incorporating the constraints derived from the anchoring experiment to minimize uncertainty. The obtained results were then compared against an in-house SCAL experiment to assess the predictive power of our DRP workflow and the accuracy of the wettability anchoring experiment. Furthermore, new simulations were conducted on different facies using a new sample without the need for a new anchoring experiment. Finally, the extrapolated simulation results were compared to an in-house SCAL experiment, allowing for an evaluation of the extrapolation capabilities of DRP simulation.