Comparison of reverse osmosis membrane fouling profiles from Australian water recycling plants

Fouling profiles for reverse osmosis membranes were compared for four municipal and one industrial effluent samples collected from water recycling schemes located on the east coast of Australia. The dissolved organic carbon (DOC) of the four municipal effluents ranged from 5 to 8 mg L-1, while the industrial effluent contained 59 mg L-1. The hydrophobic/hydrophilic content and molecular size distribution of the effluent organic matter (EfOM), post biological treatment and membrane filtration, were determined using rapid resin fractionation and size exclusion chromatography. Thin film polyamide reverse osmosis (RO) membranes were operated in cross flow to a recovery of 87%. A good correlation was observed between flux decline on the municipal effluent and the overall organic loading (R-2 = 0.957 and 0.929) and the percent fraction of hydrophobic acids (HPhoA, R-2 = 0.987). A similar correlation was observed for the hydrophilic fraction (HPhi, R-2 = 0.912), however, the rate of fouling was independent of the percentage of transphilic acids (TPhiA, R-2 = 0.029). Further support for the role of the humic and neutral fractions in membrane fouling was found in changes in the surface chemistry of the polyamide membranes as reflected in change in contact angle and zeta potential. While it is tempting to use EfOM characterisation as a predictor of RO fouling, data from the performance of membranes on the industrial stream suggest that the data should be interpreted cautiously and that consideration should also be given to inorganic fouling. In this case, although the industrial effluent contained more than six times the amount of DOC, which consisted of more than 80% humic substances, scanning electron microscope-energy dispersive X-ray analysis indicated that the dominant foulant on the industrial effluent was inorganic deposits. Consequently, detailed organic characterisation, while useful is site specific and is no substitute for a comprehensive assessment of all potential foulants. (C) 2012 Elsevier B.V. All rights reserved.