Integration of Qualitative and Quantitative Approaches for 3D Geostatistical Modeling of the Ciénaga De Oro Formation, Southern Sinú-San Jacinto Basin, Colombia

This study develops a three-dimensional (3D) geostatistical model of the Ciénaga de Oro Formation in the southern Sinú–San Jacinto Basin (Colombia), integrating structural, sedimentological, and petrophysical data to identify new hydrocarbon storage-prone zones. The structural model was constructed from seismic interpretation, well log correlation, and velocity models derived from VSP and check shots. Sedimentological models were generated by means of facies definition through field—outcrops description, well-log analysis, integrating computed tomography and digital rock analysis (Digital SCAL), complemented by automatic facies classification through a multi-layer perceptron (MLP) neural network. In this framework, Petrophysical properties, including porosity, permeability, density and clay volume, were interpolated using geostatistical sequential Gaussian simulation (SGS) and kriging, accounting for directional anisotropy (N45W), using the previously defined structural model as a basis. Analysis of the ANH-SSJ-La Estrella-1X and ANH-SSJ-Nueva Esperanza-1X wells revealed reservoir variability: clean sandstones associated with distributary channels exhibited the highest quality (Φ > 20%, K > 1000 mD), while heterolithic sandstones linked to delta-front mouth bars were identified as new secondary reservoir-prone zones (Φ > 10%, K > 10 mD). The proposed methodology provides a robust, integrated and replicable workflow for reservoir characterization in complex sedimentary environments and reduces exploration uncertainty, supporting both prospect evaluation and development planning.