2011 Volume 10 Issue 4 Pages 131-140
Membrane bioreactors (MBRs) have been widely used to purify wastewater for reuse. However, MBRs are subject to fouling, which is the phenomenon whereby foulants absorb or deposit on the membrane. When MBR filtration is operated at a constant permeate flow rate, the transmembrane pressure (TMP) and the energy required to maintain the permeate rate increase with time. To enable chemical cleaning to be performed at an appropriate time, one must be able to predict membrane fouling in the long-term. There has been research on correlations among fouling phenomena, water quality variables, and operating conditions. Therefore, in this paper, we aimed to construct a statistical model between the increase in TMP and MBR parameters such as water quality variables and operating conditions and to use this model to predict TMP (Figure 1). In our study, two methods are used to construct regression models. One is a partial least-squares method and the other is a support vector regression method. We analyzed two data sets measured in a real industrial MBR plant and then confirmed that the constructed model could predict TMP over time with high accuracy (Figures 4, 5).