Experimental and numerical study of the effect of polymer flooding on sand production in poorly consolidated porous media

TL;DR

This study investigates how polymer flooding affects sand production in poorly consolidated porous media through combined experimental and numerical modeling, revealing increased sand production during polymer injection.

Contribution

It introduces a coupled DEM-CFD numerical model validated against laboratory experiments to analyze sand production mechanisms during polymer flooding.

Findings

Polymer injection causes significant sand production, with over 50% of sand produced after one pore volume.

No sand production observed during brine injection at various flow rates.

Numerical model shows non-uniform bond breakage and increased medium-sized particles during polymer flooding.

Abstract

Polymer flooding is crucial in hydrocarbon production, increasing oil recovery by improving the water-oil mobility ratio. However, the high viscosity of displacing fluid may cause problems with sand production on poorly consolidated reservoirs. This work investigates the effect of polymer injection on the sand production phenomenon using the experimental study and numerical model at a laboratory scale. The experiment uses an artificially made sandstone based on the characteristics of the oil field in Kazakhstan. Polymer solution based on Xanthan gum is injected into the core to study the impact of polymer flooding on sand production. The rheology of the polymer solution is also examined using a rotational rheometer, and the power-law model fits outcomes. We observe no sand production during the brine injection at various flow rate ranges. However, the sanding is noticed when the polymer solution is injected. More than 50% of cumulatively produced sand is obtained after one pore volume of polymer sand is injected. In the numerical part of the study, we present a coupling model of DEM with CFD to describe the polymer flow in a granular porous medium. In the solid phase, the modified cohesive contact model characterizes the bonding mechanism between sand particles. The fluid phase is modeled as a non-Newtonian fluid using a power-law model. We verify the numerical model with the laboratory experiment result. The numerical model observes non-uniform bond breakage when only a confining stress is applied. Alternatively, the injection of the polymer into the sample leads to a relatively gradual decrease in bonds. The significant difference in the pressure of the fluid results in its higher velocity, which causes intensive sand production at the beginning of the simulation. The ratio of medium-sized produced particles is greater than the initial ratio of those before injection.