Development of a new technique to predict reverse osmosis fouling

In this paper RO colloidal fouling was predicted using an on-line Feed Fouling Monitor (FFM) combined with a monitor based On ultrasonic time domain retlectometry (UTDR) under constant flux filtration. The FFM incorporated the relevant cross-flow hydrodynamics and detected the development of foulant load and resistance from a continuous sample passing over a small UF membrane. A UF membrane was used in the FFM, rather than a RO membrane, because it was more sensitive to fouling resistance and this decreased monitoring time and increased accuracy. The UTDR was coupled to a RO membrane cell to monitor the rate of cake thickness increase, which is the required information for the estimation of cake-enhanced osmotic pressure (CEOP). A model was developed to predict fouling in RO using data from the two monitors and combined both the cake resistance and the CEOP effect due to cake formation. RO and FFM fouling experiments were performed at different constant fluxes using the same feed solutions (200 mg/l SiO2, 2000 mg/l NaCl) and the same cross-flow velocity (0.1 m/s). The results showed that higher fluxes caused an increased fouling rate for both RO and the FEM. The foulant layer thickness measured by UTDR in the RO experiments increased faster at higher applied fluxes and the estimated CEOP effect was also strongly influenced by the flux. The results show that the model, combined with the FFM and UTDR measurements, can provide a good estimation of the RO fouling profile over a range of applied fluxes. This study also underscores the importance of CEOP in contributing to the increase in transmembrane pressure due to colloidal fouling in RO of saline feeds. (C) 2013 Elsevier BY. All rights reserved.