Minimum miscibility pressure (MMP) is recognized as a key design parameter in miscible gas injection because it governs the pressure at which injected gas and reservoir oil develop dynamic miscibility and, consequently, strongly influences displacement efficiency, process feasibility, and expected oil recovery. In this review, the diverse methods used to determine MMP in gas–oil systems are critically examined to provide a unified, application-oriented framework for method selection. Conventional and emerging experimental techniques, including slim-tube, rising-bubble, vanishing interfacial-tension, core-flood, microfluidic, and nanofluidic approaches, are assessed alongside empirical correlations, equation-of-state-based models, numerical simulation methods, and machine-learning workflows. It is shown that experimental methods remain the benchmark for MMP determination, with the slim-tube experiment being the most representative of multicontact miscibility, while faster screening alternatives, such as rising-bubble and vanishing interfacial tension methods, offer lower cost and shorter testing time, but reduced physical realism. Nonexperimental methods are shown to provide rapid and economical prediction approaches, although their accuracy is strongly affected by thermodynamic assumptions, data quality, and transferability. More than three principal method families and multiple subcategories are synthesized within a single comparative framework, and recent machine-learning models are shown to achieve high predictive accuracy when trained on representative data sets. By comparing these approaches within a single framework, the review clarifies their governing principles, strengths, limitations, and best-use domains, and highlights the need for standardized benchmark data sets, uncertainty-aware validation, and tighter integration among laboratory measurements, compositional modeling, and data-driven prediction. Hence, this framework supports more defensible MMP evaluation for research, screening, and field-oriented decision-making.
Al-Hakami et al. (Sat,) studied this question.