Reservoir Modeling Overview

Principles

Definition

Reservoir modeling is a methodological approach within applied earth sciences and engineering that involves creating quantitative representations of subsurface geological formations. It integrates diverse geological, geophysical, and engineering data with physical principles to simulate fluid flow and transport processes, enabling predictive analysis for resource assessment, production optimization, and subsurface management.

Ontological type

Core Methods

Data Integration

Field Development Applications

Key developments

Key Figures

Dual‑Porosity Foundations era

Hossein Kazemi [1] was a leading figure in the era, pursuing research across major centers such as University of California, Berkeley [3] and California Institute of Technology [4]. His key contributions include the numerical simulation of water–oil flow in Naturally Fractured Reservoirs [5] and Improvements in Simulation of Naturally Fractured Reservoirs [6], which laid foundational numerical approaches for fractured-reservoir modeling in this era. John H. Seinfeld [2] contributed within leading institutions such as University of California, Berkeley [3] and California Institute of Technology [4]. His key contributions include A New Algorithm for Automatic History Matching [7] and Error Analysis in History Matching: The Optimum Level of Parameterization [8], which advanced history-matching methodologies and parameterization strategies, driving reliable calibration in this era.

Integrated Data Assimilation era

Craig Cipolla [1] stands as a central figure in the Integrated Data Assimilation era (1994–2014), engaging with cross-disciplinary collaborations across industry and academia. His key contributions include seismic-constrained geostatistical inversion and ensemble-based history matching, illustrated by What Is Stimulated Reservoir Volume? [4], Stimulating Unconventional Reservoirs: Maximizing Network Growth while Optimizing Fracture Conductivity [5], and Modeling Production and Evaluating Fracture Performance in Unconventional Gas Reservoirs [6], which advanced uncertainty-aware forecasting in shale contexts. E. P. Lolon [2] is associated with Texas A&M University [3], shaping the era’s reservoir-modeling practice through shale-gas and unconventional plays. Her contributions are reflected in What Is Stimulated Reservoir Volume? [4], Stimulating Unconventional Reservoirs: Maximizing Network Growth while Optimizing Fracture Conductivity [5], and Reservoir Modeling in Shale-Gas Reservoirs [7], where integrated data assimilation, fracturing performance evaluation, and scale-appropriate modeling improved predictive capability for reservoir development during this era.

Physics‑Informed Hybrid Modeling era

Soroosh Sorooshian [1] is a leading figure in reservoir modeling whose work in this era of physics-informed hybrid modeling is associated with the University of California, San Diego [3] and the University of California, Irvine [4]. He contributed to reservoir forecasting by integrating artificial intelligence with climate-phenomenon information to develop monthly inflow forecasts [7], and by applying a classification and regression-tree approach with a shuffled cross-validation scheme for California reservoir operation [8], advances that were pivotal for real-time, data-driven decision support in the era. Kamy Sepehrnoori [2] has affiliations with Peking University [5] and Texas A&M University [6], illustrating cross-institutional influence in this era. His key contributions include discrete-fracture modeling of complex hydraulic-fracture geometries in reservoir simulators [9], compositional simulation of CO2 Huff 'n' Puff in Eagle Ford tight oil reservoirs with CO2 diffusion and nanopore confinement [10], and the implementation of a preprocessor for embedded discrete fracture modeling in an IMPEC compositional reservoir simulator [11], collectively enabling accurate fracture-matrix representations and scalable simulations that support topology-aware surrogates and hybrid models in this era.