Numerical Investigation of LNAPL Displacement by Complex Fluids: Colloidal Gas Aphrons in One-Dimensional Porous Columns

Abstract

Unintentional industrial releases of light non-aqueous phase liquids (LNAPLs) have led to contamination ofsoils and aquifers, posing serious risks to ecological sustainability and public health. Conventional remediation techniques, such as pump-and-treat systems, are commonly applied to address this problem. However,they often exhibit limited efficiency, as substantial fractions of residual hydrocarbons remain trapped withinthe pore space due to capillary forces and subsurface heterogeneity. These limitations highlight the need foralternative remediation strategies to improve LNAPL displacement efficiency in porous media. In this context,the use of viscous, shear-thinning fluids during in situ remediation has emerged as a promising approach, astheir non-Newtonian behavior enables improved mobility control, a more uniform sweep, and enhanced hydrocarbon mobilization.This study investigates the potential of unconventional in-situ flushing using complex shear-thinning fluids,with a particular focus on colloidal gas aphrons (CGA), to enhance LNAPL recovery. CGA fluids are gasin-liquid dispersions stabilized by a polymer–surfactant system and exhibit non-Newtonian shear-thinningrheology, characterized by elevated apparent viscosity at low shear rates. In this work, the CGA formulation was prepared using biopolymer xanthan gum (XG) as the viscosifying agent and sodium dodecyl sulfate(SDS) as the surfactant, yielding a stable fluid with favorable rheological properties for controlled injectionand transport in porous media. The performance of CGA-based flushing was investigated through laboratoryscale one-dimensional (1D) sand-packed column experiments, supported by numerical modeling.1D column experiments demonstrated high diesel recovery using CGA injection, reaching approximately 98%.The CGA formulation exhibited stable flow behavior, characterized by piston-like displacement throughoutthe injection period. A numerical model was developed using Computer Modelling Group (CMG) STARSand calibrated against experimental observations, including sand-pack geometry, porosity, permeability, fluidproperties, and injection conditions, with explicit incorporation of the shear-thinning behavior of the CGA.The results demonstrate that the numerical model successfully reproduces the experimentally observed stabledisplacement front and LNAPL recovery, with predicted recovery deviating from experimental values by lessthan 1%. These discrepancies are mainly associated with simplified assumptions regarding pore-scale heterogeneity and CGA rheology. Based on this validation, the model was extended to two-dimensional (2D) simulations to investigate the influence of flow geometry on the local apparent viscosity of the shear-thinningCGA. These simulations revealed spatial variations in apparent viscosity and flow structure that cannot becaptured by one-dimensional models. Overall, the combined experimental–numerical framework provides arobust basis for evaluating CGA-based flushing strategies. Future work will extend this approach to threedimensional (3D) heterogeneous systems and near-field conditions, with particular emphasis on optimizinginjection strategies, CGA slug design, and breakthrough control for sustainable LNAPL remediation.

More by Sagyn Omirbekov

• Data-driven preprocessing for prediction of the dynamic α₀-factor in full-scale WWTP aerationResults in Engineering · 2026
• Foam-assisted (bio)remediation of petroleum-contaminated soil: effects of surfactant formulation on foam behaviour, interfacial properties, and bioavailability2026
• OPTIMIZING SURFACTANT FOAM FORMULATIONS FOR ENHANCED (BIO)REMEDIATION OF PETROLEUM-CONTAMINATED SOILHAL · 2026
• Foams for enhanced remediation of NAPL-contaminated soils and aquifers: A review of foam formation, stability, and environmental applicationsFigshare · 2026