Insights into colloidal membrane fouling mechanisms for nanofiltration of surface water using single and hybrid membrane processes

Abstract

Colloidal membrane fouling mechanisms were modeled for experimental data provided for nanofiltration of surface water by single and hybrid membrane systems. The hybrid system is composed of coagulation/activated carbon adsorption and bare polyethersulfone or mixed‐matrix membranes. The comprehensive modeling of these membrane systems was performed by three explicit equations to evaluate the effects of combined cake‐filtration and pore‐blocking on the fouling processes. These equations were derived from Darcy's law applied under constant operating pressure. The results showed that by using the nonlinear regression method, the combined cake‐filtration/pore‐blocking models demonstrated more accurate fits to the experimental permeate volume data compared to individual blocking models. The accuracy achieved by such nonlinear fits is expressed in terms of a mean squared error of approximately 10−10 (m3/m2)2 and R2 > 0.999. Furthermore, the impact of a pretreatment stage and the type of membrane used to prevent or reduce the fouling are also investigated. Experimental results reveal that the colloidal fouling depends on the type of membrane and the characteristics of raw water and is not only related to the cake filtration.