Geometrical straining of particles in porous media is of critical importance in a broad range of natural and industrial settings, such as the contaminants transport in aquifers and the permeability decline due to pore plugging in oil reservoirs. Despite its importance, relatively few studies have been performed on particle straining under fluid-driven flows in porous media. Pore-network modeling is attractive option for predicting particle straining. However, network models often lack predictive capability due to simplification of pore-geometry and require adjustable parameters. This is especially true for particle straining. Direct numerical simulation using computational fluid dynamics (CFD) to model the fluid phase, coupled with the discrete element method (DEM) to model the particle transport is a more rigorous approach but computationally expensive. The CFD-DEM simulations are limited to systems with particle number less than 105.
Our current research has coupled a pore network model with a CFD-DEM model in a computationally-efficient framework. Particle jamming is a matter of probability. We perform CFD-DEM simulations on particle filtration by a single layer of grains and formulate the jamming probability as a function of particle/pore size ratio, particle concentration, pore throat geometry etc. The upscaled results are then implemented into the pore-network model. During a time step, the jamming probability of each pore throat is calculated based on the local particle concentration, flow rate etc. The hydraulic conductivities of pore throats are updated dynamically. The numerical results are compared to direct CFD-DEM and experiment results and good agreement is achieved.
|Acceptance of Terms and Conditions||Click here to agree|