Electrofacies Estimation of Carbonate Reservoir in the Scotian Offshore Basin, Canada Using the Multi-resolution Graph-Based Clustering (MRGC) to Develop the Rock Property Models

Rock properties in geomechanical models depend on electrofacies. Electrofacies classification is a crucial task for generating accurate rock property models. Insufficient information about core samples and image logs at all locations is the major drawback in electrofacies classification. This study classified electrofacies by the multi-resolution graph-based clustering (MRGC) approach using the well-log data from the Scotian shelf, Offshore Canada. The unsupervised method such as MRGC, ascendant hierarchical clustering, and self-organizing map uses the K-nearest neighbors and kernel index for defining the cluster dots which characterize the electrofacies. These classified facies are incorporated with core information to establish a relationship. This relation can develop electrofacies in the non-core zone/wells. The electrofacies were predicted using the MRGC approach to generate rock mechanical properties such as Young's modulus, Poisson’s ratio, unconfirmed compressive strength, and internal friction coefficient.