Abstract
A fundamental understanding of fouling mechanisms is critical to improving filtration operations. The performance of four parallel membrane bioreactors (MBRs) with different sludge retention times (SRTs) was monitored during long-term constant flowrate filtration. The characteristics of the membrane and transmembrane pressure (TMP) profiles obtained were studied to demonstrate fouling mechanisms. Both classical blocking models and their combined models were evaluated. The intermediate model provided very good agreement with all the TMP data. However, the combined cake-intermediate and intermediate-standard models were more effective in the description of the experimental data. Contributions analysis indicated that the cake, intermediate and standard blocking models were the dominant fouling mechanisms. Scanning electron microscopy and energy dispersive X-ray (SEM-EDX) imaging showed that cake blocking by organic matter and standard blocking by inorganic matter made the main contributions to membrane fouling. The combined cake-intermediate and intermediate-standard models may be applicable to systems where these two models are consistent with the experimentally observed fouling mechanisms in an MBR.
Adaptive model and model selection for long-term transmembrane pressure prediction in membrane bioreactors
