1. Introduction
Fluid flow through porous media can alter the void space geometry theough chemical erosion, which dissolves and then shrinks solid grains, or mechanical erosion, which remove solid grains. The impacts of these distinct erosion mechanisms on two-phase flow behavior remain poorly understood, despite their broad relevance to subsurface processes raning from geological energy storage to groundwater remediation. Beyond void space geometry, two-phase flow behavior is also govern by the capillary number and wettability. The combined influence of void space geometry, capillary number, and wettability on two-phase flow therefore represents a multidimensional problem that has not been systematically studied.
In this study, we use pore-scale numerical simulations to systematically and quantitatively investigate, for the first time, the impacts of void space geometry induced by chemical and mechanical erosion on two-phase flow behavior across a wide range of capillary numbers and wettability conditions. This study provides mechanistic insights into how erosion-induced void space evolution governs subsurface two-phase transport, with direct implications for applications including groundwater remediation and geological carbon sequestration.
2 Geometry generation
The original geometry represents a natural geological porous medium, extract from a two-dimensional slice of an X-ray micro-computed tomography scan of a quartz sand pack. The computational domain measures 15.4 mm in the streamwise direction and 5.85 mm in width. Because this study focuses on the subsequent impacts of void space geometries on two-phase flow rather than the dynamic erosion process itself, simplified yet physically meaningful morphological algorithms were employed to mimic the erosion mechanisms. For chemical erosion method, dissolution was simplified as a uniform surface retreat, where the solid phase surfaces were uniformly retracted. Conversely, mechanical erosion method was conceptualized as fine particle migration, implemented via a morphological procedure where solid grains were systematically removed based on a threshold size of their equivalent diameters. Notably, to isolate the purely void space geometry effects, the porosity of both eroded porous media was strictly constrained to an identical value of 0.63.
3. Numerical implementation
The typical Volume of Fluid (VOF) hydrodynamics solver, interFoam, within the OpenFOAM (ESI, v2306) framework, is adopted to simulate two-phase flow in porous media. The governing equations are discretized using a finite-volume scheme, which applies the integral form of conservation laws over each cell of the computational domain and evaluates fluxes across the cell faces, thereby ensuring local conservation over each cell in complex geometries. To ensure both numerical stability and computational efficiency, we adopted an adaptive time-stepping strategy. In this approach, the time step is dynamically adjusted during the simulation to maintain a Courant number below 0.2 throughout the entire computation domain.
An immiscible two-phase flow system was considered in this study, where water serves as the invading phase and oil acts as the the defending phase. The densities of both fluid phases are ρip = ρdp = 1000 kg/m3. The dynamic viscosities are prescribed as μip = 0.001 Pa·s and μdp = 0.02 Pa·s. The subscripts ip and dp denote the invading phase and defending phase, respectively. To systematically investigate the joint influences of wettability and hydrodynamic conditions, both the contact angle and the injection velocity are varied. The contact angle θ is varied across from four values 15°, 45°, 75°, and 105°, providing a broad spectrum that spans from a strongly water-wet to weakly oil-wet conditions and encompasses a wide range of environmentally and industrially relvant scenarios. Meanwhile, the velocity is quantified by the dimensionless capillary number (Ca). Simulations were performed at six inlet injection velocities to precisely yield logCa of -5.9, -5.5, -4.9, -4.5, -3.9, and -3.5.
