Abstract
Membrane bioreactors (MBRs) have been widely used for wastewater treatment. Although complete solid–liquid separation can be achieved using membrane, MBRs are subject to membrane fouling. To enable chemical cleaning to be performed at an appropriate time, fouling must be predicted in the long–term. Fouling prediction corresponds to transmembrane pressure (TMP) prediction under a condition of constant–rate filtration. One of the reasons to make TMP difficult to predict is a TMP jump. After the long-term operation of MBR under the condition of constant–rate filtration, TMP increases rapidly, which is called a TMP jump. We therefore have been developing both a TMP prediction model and a TMP jump prediction model. A TMP prediction model can predict future TMP with high accuracy in the long–term. A TMP jump prediction model can accurately predict timing of TMP jumps. By using our proposed models, we can arrange a schedule of chemical cleaning and optimize operating conditions and water quality that can prevent MBR fouling.
