Tertiary oil recovery, particularly chemical flooding, has emerged as an essential strategy to enhance oil recovery from residual and remaining oil post-water flooding (Park et al., 2021; Tavakkoli et al., 2022). Among chemical flooding techniques, microemulsion flooding has gained significant attention due to its unique ability to reduce interfacial tension (IFT), alter wettability, and mobilize trapped residual oil in porous media (Qin et al., 2020; Das et al., 2021; Wu et al., 2024; Koreh et al., 2025). Microemulsions, thermodynamically stable dispersions of surfactant, cosolvent, salt, and oil, have demonstrated remarkable success in improving oil recovery.

Microemulsions are classified into three types based on their phase behavior: Winsor I (oil-in-water, O/W), Winsor II (water-in-oil, W/O), and Winsor III (bicontinuous). Winsor I microemulsions consist of nano-scale oil droplets dispersed in a continuous water phase, stabilized by surfactant molecules at the oil-water interface. Winsor II microemulsions feature water droplets dispersed in a continuous oil phase, while Winsor III microemulsions exhibit a bicontinuous structure where both oil and water phases are interconnected, forming a sponge-like network (Scriven, 1976; Zhao et al., 2025). The formation of these microemulsions is highly dependent on surfactant concentration, cosurfactant selection, salinity, and oil-to-water ratio (Budhathoki et al., 2016). Salinity and oil-water ratio scanning experiments are critical for identifying optimal conditions for microemulsion formation, ensuring maximum oil recovery efficiency (Zhou et al., 2020; Mariyate and Bera, 2022).

The unique physical properties of microemulsions, such as solubilization capacity, ultra-low IFT, wettability alteration, and viscosity modulation, collectively contribute to enhanced oil recovery (EOR). Solubilization enables microemulsions to dissolve crude oil and water, improving oil mobility (Wei et al., 2011). Ultra-low IFT reduces the oil-water interfacial energy, facilitating the detachment of residual oil (Das et al., 2021; Li et al., 2025). Wettability alteration adjusts the wetting state of rock surfaces, optimizing oil-water flow paths, while viscosity modulation enhances the sweep efficiency of displacing fluids (Yang et al., 2021). These properties make microemulsions highly effective in EOR applications.

Two primary methods are employed to utilize microemulsions for EOR: ex-situ and in situ. The ex-situ method involves injecting pre-prepared microemulsions into the reservoir (Zhou et al., 2020). While this approach has demonstrated significant oil recovery enhancement, it faces challenges such as high injection pressures and potential incompatibility with reservoir conditions (Zhu et al., 2022). In contrast, the in situ method involves injecting a microemulsion-forming surfactant system into the formation, where microemulsions form spontaneously within the pores (Mo et al., 2023). This method leverages the natural conditions of the reservoir, offering better adaptability and cost-effectiveness. However, the pore-scale mechanisms governing microemulsion-assisted oil recovery remain poorly understood, limiting the optimization of this technology for field applications.

Advanced visualization techniques, including X-ray computed tomography (Hu et al., 2020), nuclear magnetic resonance (Govindarajan et al., 2020), and microfluidics (Fu et al., 2016), have facilitated the analysis of microemulsion formation and EOR in complex porous media. These technologies provide high-precision experimental data, revealing the formation mechanisms, phase behavior, and oil displacement efficiency of microemulsions. For instance, Ott et al. (2020) employed micro-CT to investigate the in situ emulsification behavior of surfactants and crude oil during alkali flooding, providing critical insights into the changes in interfacial tension and oil-water phase distribution. Unsal et al. (2019) utilized synchrotron-based X-ray micro-tomography to study the dynamic changes in emulsification under optimal salinity conditions, highlighting the critical role of salinity in emulsification efficiency Alzahid et al. (2019) systematically investigated the pore-scale flow characteristics of surfactant flooding under varying salinity conditions, demonstrating the significant impact of salinity on emulsified phase properties and oil-water distribution. She et al. (2021) further advanced the field by employing three-dimensional visualization techniques to study the role of in situ emulsification in alkali flooding, revealing the formation, migration, and trapping mechanisms of emulsions in porous media.

Despite these advancements, visualizing the dynamic distribution of microemulsion concentrations within pores remains challenging due to the temporal and spatial limitations of X-ray CT. Two-dimensional micromodels with high-speed cameras have been employed to capture near real-time emulsification dynamics. For example, Unsal et al. (2016) used microfluidic flow experiments to study the in situ formation of microemulsions by co-injecting n-decane and surfactant solutions into a T-junction capillary geometry. Broens and Unsal (2018) investigated emulsification kinetics in quasi-miscible flow conditions, while Tagavifar et al. (2017) analyzed phase formation and spatial configurations in a 2.5-dimensional (2.5D) micromodel. Zhao et al. (2022) demonstrated in situ emulsification using on-chip experiments, and Xu et al. (2024) studied the formation and transport mechanisms of microemulsions in fractured porous micromodels. Nevertheless, current visualization techniques remain inadequate for accurately quantifying microemulsion concentrations within complex pore structures, thereby constraining the comprehensive understanding of microemulsion formation and enhanced oil recovery (EOR) mechanisms. Pore-scale numerical simulations have emerged as a promising approach for investigating oil-water interface dynamics in porous media during chemical flooding processes, including surfactant, polymer, and nanoparticle flooding. However, microemulsion flooding presents additional complexities, as it involves not only multiphase flow but also requires consideration of multiphase mass transfer processes within intricate pore networks that has received limited research attention to date.

To overcome these research gaps, the present study develops a novel pore-scale simulation framework to systematically examine microemulsion formation and EOR mechanisms in porous media. By analyzing microemulsion behavior in a simplified T-junction structure, we aim to elucidate the roles of IFT reduction, wettability alteration, viscosity enhancement, and solubilization in oil displacement. Furthermore, a real complex porous medium is reconstructed from digital rock, and the pore-scale dynamics of the emulsification process are coupled with fluid flow. This study leverages advanced imaging techniques and numerical simulations to provide a comprehensive pore-scale investigation of microemulsion-enhanced oil recovery, bridging the gap between laboratory-scale observations and field-scale applications. The findings are expected to contribute to the development of more effective EOR strategies, ultimately improving oil recovery efficiency and reducing the environmental footprint of hydrocarbon extraction.

This study employs the conventional Volume of Fluids (VoF) method to simulate two-phase displacement within porous media and accurately track the interface between water and oil. In the VoF method, an indicator function, denoted as α, is defined throughout the flow domain, representing the volume fraction of one of the fluids within each computational grid cell. Specifically, α = 1 indicates that a cell is entirely occupied by water, while α = 0 signifies that the cell is completely filled with oil. The indicator function serves as a critical tool for identifying and tracking the interface between the two phases. The governing equations for the system, which describe the behavior of two incompressible, isothermal, and immiscible fluids (such as water and oil), are presented as follows: ∂ ρ u ∂ t + ∇ · ρ uu − ∇ · μ ∇ u + ∇ u T = − ∇ p + F σ (1) ∇ · u = 0 (2) ρ = α ρ w + 1 − α ρ o (3) μ = α μ w + 1 − α μ o (4)

In the above equations, Equations 1, 2 are momentum and continuity equations of the fluid, where ρ, u , p, and F _σ are density, fluid velocity, dynamic pressure, and surface tension, respectively. t is the time. The effect of gravity is ignored in two-dimensional models. α denotes the volume fraction of the water phase and correspondingly (1−α) the oil phase in Equations 3, 4.

The surface tension F _σ in Equation 2 is modeled as continuum surface force (CSF) and calculated by F σ = σ κ ∇ α (5) where σ is the surface tension coefficient, and κ is the curvature of the water-oil interface in Equation 5, which can be approximated as follows κ = − ∇ · n (6) n is the interface normal unit vector (in Equations 6-8) given by n = ∇ α ∇ α (7)

To accurately model the wettability of the porous media, the contact angle is imposed as a wetting boundary condition. The normal unit vector of the interface at the wall boundary is adjusted according to the following equation: n = n p ⁡ cos ⁡ θ + n t ⁡ sin ⁡ θ (8) where n _p is the unit vector perpendicular to the wall, n _t is the unit vector tangential to the wall, and θ is the contact angle. In this way, the interface can form the prescribed angle θ when in contact with the wall.

In this study, we will only focus on the Winsor I (oil in water, O/W) microemulsion. The surfactant system forming the microemulsion is an aqueous solute and the microemulsion formed is treated as a solute in water.

To track the concentration of microemulsion-forming surfactant system in water, the mass conservation equation for the surfactant is calculated: ∂ α C ∂ t + ∇ · C α u = ∇ · α D c ∇ C − S c (9) where α is 1, representing surfactant in water phase, C and D _c are the concentration and diffusion coefficient of surfactant in water, respectively. The source term S _c denotes the rate of consumption of surfactant in Equation 9.

To track the concentration of microemulsion in water, the mass conservation equation for the microemulsion is calculated: ∂ α C e ∂ t + ∇ · α C e u − ∇ · α D e ∇ C e = S e (10) where α is 1, representing microemulsion in water phase, C _e and D _e are the concentration and diffusion coefficient of microemulsion in water, respectively. The source term S _e denotes the rate of production of microemulsion in Equation 10.

The evolution of water or oil phase is governed by the mass balance equation: ∂ α ∂ t + ∇ · α u = S α (11)

The source term is calculated by: S a = − k α C C e − C e ∞ δ (12) where k _α and δ are the rate of emulsification, and specific interfacial area between oil and water in Equations 11, 12.

The mathematical model has been implemented within the open-source computational fluid dynamics (CFD) platform OpenFOAM. The governing equations are discretized using a finite-volume method and solved sequentially. The pressure–velocity coupling is addressed through a predictor–corrector approach, leveraging the Pressure-Implicit with Splitting of Operators (PISO) algorithm to ensure numerical stability and accuracy.

In this study, a novel pore-scale numerical simulation model was proposed to capture the in situ microemulsion formation during surfactant solution flooding in porous media. Then, a series of numerical simulation were conducted to explore the dynamics distribution of oil and in situ formed microemulsion during microemulsion-forming surfactant flooding in a real rock medium. The influence of the wettability and viscosity of microemulsion on the enhanced oil recovery was analyzed. The primary conclusions are summarized as follows:

1. The proposed novel pore-scale numerical simulation model is able to simulate the dynamic formation process of oil-in-water microemulsions within the pore space. By considering some typical characteristics of in situ formed microemulsions, including increased viscosity, reduced interfacial tension, altered wettability, and solubilization, it is analyzed how these characteristics enhance recovery.

It is worth noting that the proposed model is based on the Navier-Stokes equations, which fundamentally restricts the analysis to conventional fluid dynamics. Consequently, the framework may not accurately capture flow behaviors in unconventional reservoirs (e.g., shale formations) where nanoscale pore structures dominate the transport phenomena. In addition, while this investigation provides valuable insights into oil-water displacement mechanisms at the pore scale and identifies key factors governing microemulsion-driven recovery enhancement, direct field application remains challenging due to the inherent scale differences between laboratory investigations and reservoir conditions.