InterPore2025

Numerical modeling of Complex Fluid injection for LNAPL displacement: Insights from 1D Column Experiments

22 May 2025, 10:35

1h 30m

Poster Presentation (MS18) Innovative Methods for Characterization, Monitoring, and Remediation of Contaminated Soils and Aquifers Poster

Speaker

Dana Sapobekova (National Laboratory Astana - Nazarbayev University)

Description

Contamination of soils and aquifers by light non-aqueous phase liquids (LNAPLs) poses significant risks to environmental sustainability and public health. Conventional in-situ LNAPL remediation techniques often encounter high costs and limited efficiency challenges, leaving residual hydrocarbons trapped within soil pores. As an alternative, unconventional in-situ flushing using complex shear-thinning fluids — such as polymers, foam, emulsions, etc. — has emerged as a promising approach. These advanced fluids enhance contaminant recovery through their high viscosity and non-Newtonian shear-thinning properties, ensuring stable and uniform displacement within porous media.
This study, which investigates the performance and flow behavior of aqueous biopolymers in porous media through a series of one-dimensional (1D) sand-packed column experiments, provides a solid foundation for further numerical modeling efforts. The primary objective is to assess the potential and feasibility of using reservoir simulators for soil remediation in LNAPL-contaminated environments while exploring upscaling approaches to implement these techniques effectively at the field scale.
The model was developed using the Builder package in the CMG reservoir simulator, with numerical computations performed via the IMEX and STARS simulation packages. It employs the black oil model equations to describe a two-phase flow system, where the LNAPL is defined as the oil phase, and the polymer is represented within the aqueous solution.
One-dimensional (1D) column experiments served as the foundation for modeling bio-polymer injection. These experiments thoroughly assessed the polymer's recovery efficiency in displacing LNAPL (here diesel fuel) from unconsolidated, homogeneous porous media. The model was meticulously calibrated using experimental data by incorporating parameters such as porous media characteristics, dimensions, boundary conditions, and the properties of the injected and displaced fluids, ensuring accurate replication of the actual model conditions. A key challenge is adapting the simulator for the application in highly permeable porous media.
The modeling process revealed that the polymer behavior can be effectively adjusted by calibrating the injection fluid's endpoint mobility. This adjustment accurately represented the reduced mobility ratio observed during the experiments. As a result, the model demonstrated a stable displacement front, with flow propagation as a function of injected pore volumes (PV) closely matching the experimental data. Furthermore, the model achieved a high recovery yield, successfully replicating the experimental outcomes and validating its accuracy in simulating polymer-assisted remediation processes.
The two-phase flow model demonstrated great agreement with experimental observations, validating its accuracy in representing the real case. This confirms the model’s reliability as a tool for simulating polymer-assisted remediation processes. These results led us to upscale the process that can be applied to field case problems.