Network model evaluation of the effects of pore and particle size distributions on dynamic straining-dominant deep bed filtration in porous media

Abstract

A straining-dominant filtration model based on a 2D square network is proposed. In this model, the colloid particle can be captured in a node only. The effects of network parameters (node size distribution parameters μn and σn and pore throat size distribution parameters μn and σb), particle concentration of influent (C0), and particle size distribution parameters (μr and σr) on straining-dominant deep bed filtration simulation are investigated. Four parameters (C0, μn, μb, and μr) significantly affect the simulated pressure drop and the normalized particle concentrations of effluent (Ce/C0). The increase in parameters μn and μb decreases pressure drop and increases Ce/C0, whereas the increase in parameters μr and C0 increases pressure drop and decreases Ce/C0. With the experimental condition and a 5% variation in parameters, parameters σn, σb, and σr are lowly sensitive to Ce/C0 and pressure drop. When the variation in σr is sufficiently large, the model simulation is significantly affected. The simulated normalized effluent concentrations and pressure drop are consistent with the experimental data under appropriate simulated conditions.