Particulate fouling, due to the deposition of particles and colloids onto RO membranes, is one of the persistent problems, resulting in higher energy consumption and more frequent membrane cleaning and replacement. Hence, a reliable method to accurately predict particulate fouling is very important to optimize the performance of RO systems. Therefore, this PhD research aimed to further develop and apply the MFI-UF method to more accurately predict particulate fouling in RO systems.
Firstly, MFI-UF calibration was studied examining standard polystyrene and dextran particles. The study concluded that a heterogenous mixture of polystyrene particles is considered promising for MFI-UF calibration. Moreover, real surface water was also used to validate MFI-UF measurements. Results demonstrated that the MFI-UF method can be used to measure different levels of particulate fouling under a wide range of testing conditions (at fluxes of 20-200 L/m2.h using 5, 10, 50 and 100 kDa UF membranes).
Furthermore, the effect of surface porosity of MFI-UF membranes was investigated, quantified and corrected using a new approach. Firstly, the surface porosity of MFI-UF membranes (5–100 kDa) was determined using ultra-high resolution SEM. Thereafter, the MFI-UF was measured using suspensions of standard polystyrene particles that were pre-washed to remove particle fractions smaller than the pores of MFI-UF membranes, thus ensuring that variations in measured MFI-UF values were due to the membrane surface porosity and not membrane pore size. The results showed that the lower the MWCO of MFI-UF membrane, the lower the surface porosity, and the more non-uniform the distribution of pores is on the membrane surface. This results in lower effective membrane area and higher local fluxes, which eventually lead to overestimation of the measured MFI-UF. Consequently, correction factors of 0.4-1.0 were proposed for the MFI-UF measured with 5-100 kDa membranes, respectively. Correcting the surface porosity improved fouling prediction in a full-scale RO plant by 50-60%.
Subsequently, the improved MFI-UF method was applied in two full-scale RO plants. The study showed that particulate fouling predicted based on the 10-100 kDa membrane had the best agreement with the actual fouling observed in the studied RO plants.
Finally, the research concludes with the introduction of a complete testing protocol for the MFI-UF method, providing a reliable tool to accurately measure and predict the rate of particulate fouling in RO systems, which can support engineers, operators and researchers to design, operate and monitor RO systems more effectively and sustainably.