Fouling characteristics and flux prediction model of reverse osmosis membrane based on hydrophobic fractions in reclaimed water

During the operation of the reverse osmosis (RO) system, the identification of key fouling components and the prediction of flux changes based on these components are crucial for minimizing the costs associated with flux recovery. In this study, the DAX-8 resin fractionation method was employed to separate the organic matter in secondary effluent (SE) into hydrophobic acid substances, hydrophobic base substances, hydrophobic neutral substances (HON), hydrophilic substances, and hydrophobic substances (HOS). The primary objective was to discern critical membrane fouling components. The results revealed that HOS constituted 50.0%-60.0% of the dissolved organic carbon concentration in the SE, emerging as the pivotal fraction responsible for RO membrane fouling, contributing to 90.0% of the flux drop induced by SE. Additionally, a fouling prediction model employing two parameters quantified the fouling characteristics of distinct hydrophilic/hydrophobic fractions in RO treatment. Model fitting results indicated a flux decrease of 61.3%, 54.3%, and 46.7% caused by SE, HOS, and HON, respectively. Notably, 90.0% of the decrease in RO membrane flux induced by SE was attributed to the presence of HOS, with HON accounting for about 90.0% of this decrease. The results of the extended DerjaguinLandau-Verwey-Overbeek theory demonstrated higher adhesion energy and cohesion energy between HOS and RO membrane surfaces compared to hydrophilic substances. This finding suggests that hydrophobic pollutants are more closely bound to RO membrane surfaces. Subsequently, an RO membrane flux prediction method based on HOS in reclaimed water was established. Predicting the RO membrane fouling characteristics of SE became feasible by measuring the concentration of HOS in the influent. Furthermore, resin pre-treatment mitigated the membrane fouling by removing HOS, and the HOS concentration was proposed as a new influent quality index for RO system design.