Methods for evaluating transport parameters of low-salt-rejection reverse osmosis (LSRRO) membranes

Low-salt-rejection reverse osmosis (LSRRO) technology for treatment of high-salinity brines requires membranes of varying solute rejection to be efficiently implemented. Nanofiltration membranes, a class of low-salt-rejection membranes toward certain solutes, may offer a viable option to design LSRRO trains and treat high-salinity industrial waste streams. However, effective process design necessitates accurate solute and water transport modeling that can predict filtration performance under the operating conditions of the LSRRO train. The Spiegler-Kedem-Katchalsky (SKK) framework has typically been used to model transport across low-salt-rejection membranes, where a reflection coefficient is added into the transport equations to account for the imperfect rejection of solutes by the membrane. Hence, the nonlinear SKK transport equations require evaluation of three coefficients to fully describe solute and water permeation. This work compares and models the performance of three commercial nanofiltration membranes, NF90, TS80, and NFX, in crossflow filtration experiments using sodium chloride solutions ranging from 25.0 to 53.9 g/L, to evaluate their feasibility for implementation in LSRRO systems. Through thorough analysis of the filtration data obtained, experimental and transport modeling methods for evaluating low-salt-rejection membranes are reviewed and expanded upon, identifying limitations that may lead to erroneous conclusions on membrane behavior in high-salinity applications. A computational approach is proposed to harmonize the determination of transport coefficients from experimental data, and experimentation guidelines are revisited revealing research gaps that must be addressed to accurately describe transport in low-salt-rejection membranes.