2016 年 41 巻 4 号 p. 155-159
Membrane bioreactors (MBRs) have been widely used for wastewater treatment. MBRs are subject to membrane fouling. To perform chemical cleaning at appropriate timing, fouling must be predicted in the long–term. Fouling prediction corresponds to transmembrane pressure (TMP) prediction under conditions of constant-rate filtration. A reason for difficulty of TMP prediction is TMP jumps, which is rapid increase of TMP after long-term operation of MBRs under constant–rate filtration. Our laboratory therefore have been developing both TMP prediction models and TMP jump prediction models. TMP prediction models can predict future TMP with high accuracy in the longterm. TMP jump prediction models can accurately predict timing of TMP jumps. A schedule of chemical cleaning of membrane can be arranged by using those models. By analyzing TMP jump model, operating conditions such as aeration can be designed for performing energy–saving MBRs. In addition, CFD simulation can provide useful information on optimizing size of bubble and arrangement of membranes in MBR.