Evaluation of membrane fouling potential by multiple membrane array system (MMAS): Measurements and applications

A new approach to evaluate the fouling potential of feed waters for reverse osmosis (RO) and nanofiltration (NF) practices was systematically investigated. A multiple membrane array system (MMAS) was developed, in which microfiltration (MF), ultrafiltration (UF), and nanofiltration (NF) membranes were connected in series. This system was designed to separate the target foulants in the feed water and evaluate the fouling potential of a targeted foulant. More specifically, particulate, colloids, and organic matters were separated by MF, UF, and NF membranes in a consecutive manner, and the modified fouling index (MFI) was measured during each separation. The resulting three MFI values are Particle-MFI, Colloid-MFI, and Organic-MFI, and presented schematically in a triangle manner. The results demonstrated that fouling potentials of various feed waters (e.g., seawaters pretreated by various pretreatment options as well as raw seawater) evaluated by the MMAS were much more accurate and informative compared to those predicted by conventional silt density index (SDI) and single MFI. The fouling potentials determined by the MMAS precisely reflected changes in feed water quality by different pretreatment options (i.e., sand filtration, microfiltration, and ultrafiltration), while SDI measurements were not sensitive enough to detect these variations. It was also shown that the fouling potentials evaluated by the MMAS were well correlated with the actual flux-decline rates determined by lab-scale RO fouling experiments simulating seawater desalination, suggesting that the MMAS could provide a better way to predict fouling potential and guidance to select proper pretreatment processes.

Research highlightsIn this research, multiple membrane array system (MMAS), for the first time, has been developed for the precise measurements of membrane fouling index. This system was designed to separate the target foulants in the feed water and evaluate the fouling potential of a targeted foulant. The fouling potentials determined by the MMAS precisely reflected changes in feed water quality by different pretreatment options (i.e., sand filtration, microfiltration, and ultrafiltration), while SDI measurements were not sensitive enough to detect these variations. It was also shown that the fouling potentials evaluated by the MMAS were well correlated with the actual flux-decline rates determined by lab-scale RO fouling experiments simulating seawater desalination, suggesting that the MMAS could provide a better way to predict fouling potential and guidance to select proper pretreatment processes.