SEDIMENTOLOGICAL, STRATIGRAPHIC, AND PETROPHYSICAL CHARACTERIZATION OF RESERVOIR FORMATIONS FOR HYDROCARBON EXPLORATION

Abstract

Accurate reservoir characterization forms the cornerstone of modern petroleum exploration, field development, and production optimization. Conventional evaluation workflows have historically relied on isolated petrophysical assessments or simplified geological models that often fail to capture the complex heterogeneity of the subsurface, resulting in significant uncertainties in resource estimation. This review synthesizes advanced interdisciplinary workflows that bridge the gap between microscopic pore geometries and macroscopic reservoir performance by systematically integrating depositional sedimentology, high-resolution sequence stratigraphy, post-depositional diagenetic history, and quantitative rock typing. Depositional systems determine the initial mineralogical composition and structural architecture of both clastic and carbonate reservoirs, whereas sequence stratigraphic cycles regulate sand-body geometries through variations in the accommodation-to-sediment supply (A/S) ratio. Superimposed on these depositional frameworks, diagenetic processes modify the primary pore network through destructive mechanisms such as mechanical compaction and quartz cementation, as well as constructive processes including mineral dissolution. At the reservoir scale, quantitative rock typing employs empirical Winland R₃₅ equations and physics-based Flow Zone Indicator (FZI) relationships to classify distinct hydraulic flow units. Furthermore, this review highlights recent advances in Digital Rock Physics (DRP), which utilizes multi-scale three-dimensional X-ray computed tomography imaging to construct high-fidelity digital twins of reservoir rocks. Simultaneously, subsurface characterization has entered a new era driven by big data analytics and artificial intelligence. Modern workflows now incorporate unsupervised clustering algorithms, Physics-Informed Neural Networks (PINNs), multimodal subsurface foundation models, and autonomous agentic systems. These advanced AI architectures process massive geological datasets to perform real-time well-log interpretation, accurate rock typing, and automated risk assessment. Ultimately, the integration of traditional reservoir characterization techniques with Digital Rock Physics and machine learning provides geoscientists with a robust predictive framework for evaluating heterogeneous reservoirs under complex, multi-scale geological uncertainty.