A single-fibre microfiltration system was employed to investigate the importance of various operating and sludge property parameters to the membrane fouling during sludge filtration. The sludge was obtained from a submerged membrane bioreactor (SMBR). A series of comparative and correlative filtration and fouling tests were conducted on the influence of the operating variables, sludge properties and the liquid-phase organic substances on the membrane fouling development. The test results were analysed statistically with Pearson's correlation coefficients and the stepwise multi-variable linear regression. According to the statistical evaluation, the membrane fouling rate has a positive correlation with the biopolymer cluster (BPC) concentration, sludge concentration (mixed liquor suspended solids, MLSS), filtration flux and viscosity, a negative correlation with the cross-flow velocity, and a weak correlation with the extracellular polymeric substances and soluble microbial products. BPC appear to be the most important factor to membrane fouling development during the sludge filtration, followed by the filtration flux and MLSS concentration. The cross-flow rate also is important to the fouling control. It is argued that, during membrane filtration of SMBR sludge, BPC interact with sludge flocs at the membrane surface to facilitate the deposition of the sludge cake layer, leading to serious membrane fouling.
Keywords: biopolymer clusters, membrane bioreactor, membrane fouling, microfiltration, multi-variable linear regression