Maximizing Mediterranean Gas Exploration Success Rate and Revolutionize Subsurface Imaging of The Reservoir Heterogeneity with Artificial Intelligence-Assisted Workflow

Abstract A successful gas exploration/appraisal drilling campaign of the Pliocene-aged complex deep-water incised canyon reservoir in the Mediterranean Basin is achieved with outstanding results through applying an innovative integrated workflow that firmly assisted with the artificial intelligence (AI) algorithms and technologies. The convolutional neural network (CNN) algorithm is the core building block of the workflow, which enabled us to recognize and delineate the internal depositional architectural elements of the incised canyon reservoir (i.e. 4th/5th order channels) directly from the 3D seismic data with high-level of accuracy. Six 4th/5th order channels have been successfully delineated within the deep-water incised canyon of interest. Three discrete depositional facies have been robustly distinguished for each 4th/5th order channel through implementing Kohonen self-organizing mapping (Kohonen SOM) algorithm. This enabled for confident high-resolution assessment of the reservoir facies and properties heterogeneities. A robust 3D reservoir model has been built utilizing the outcomes of the AI-assisted workflow, which then used to optimize the subsurface locations of the exploration and appraisal wells targeting the reservoir sweet spots. An appraisal well has been optimized and successfully drilled that encountered 102 m of net pay thickness. The long production test of the well has been shown 38 MMscf/d production rate, which is a very good commercial rate. The workflow is a major step in maximizing the exploration success rate of the Mediterranean Basins through accurate delineation of the deep-water turbidite reservoir heterogeneities and complexities, as well as in other basins where similar settings exist. By applying the workflow, the subsurface complexities were revealed with the artificial intelligence algorithms, uncertaintie s were captured, risks were reduced and project commercial value was uplifted.