Automation-driven multi-well reservoir characterization of cenomanian–eocene clastic reservoirs in the offshore Douala Sub-Basin, Cameroon

Abstract

A formation evaluation was carried out on Cenomanian–Eocene sandstone reservoirs in three offshore exploration wells (N1, N2, N3) of the Douala Sub-Basin, Cameroon. A fully reproducible Python-based workflow was designed to assess reservoir quality, applying transparent Jupyter routines for electrofacies classification, shale-volume correction, and petrophysical property estimation (ϕ, ϕe, Sw, k). The workflow integrates GR–RHOB–NPHI–Rt pattern recognition, multi-log cross-plots, and shaly-sand saturation models to differentiate fluids and quantify net pay, while maintaining a complete digital audit trail that allows rapid reprocessing when new data become available. Results show that reservoir quality and fluid content vary significantly across wells and intervals. Several units exhibit good to very good porosity and permeability yet fail pay criteria due to high water saturation. The only interval meeting all cutoffs is A3 in Well N2, which demonstrates favorable petrophysical characteristics (ϕe ≈ 10.5%, Sw ≈ 59%, k ≈ 686 mD), establishing it as the principal pay zone. Reservoir heterogeneity is strongly facies controlled: blocky to cylindrical motifs interpreted as channel/bar sands exhibit high net:gross, strong resistivity separation, and movable hydrocarbons, whereas serrated heterolithic intervals contain elevated shale content and capillary-bound water that reduce effective reservoir performance. Cross-well electrofacies mapping highlights consistent trends across the margin, showing that depositional architecture, shale distribution, and facies-dependent diagenesis dictate reservoir behavior. These insights refine net-pay prediction, help avoid misclassification of water-wet intervals, guide perforation strategies, and identify where advanced logs or testing can reduce uncertainty in future well decisions across the Douala/Kribi Campo margin.