Incorporating convection into the solution-diffusion framework enhances modeling accuracy in reverse osmosis

Abstract

Several linear correlation equations are derived from the classical solution-diffusion theory and applied to analyze over 100 experimental datasets collected under constant and sinusoidally varying feed flows and pressures. While the theory provides reasonable predictions for water flux, it fails to accurately model permeate concentration, particularly in cases of membrane degradation. Incorporating convective transport over a small portion of the membrane successfully captures the recovery rate and permeate quality across all scenarios studied (R${}^2\ge$ 0.99). This enhanced model, with only three parameters, holds significant potential for advancing future membrane transport modeling efforts.