Research On Membrane Fouling Of Short-cut Nitrification In The Intermittently Aerated Membrane Bioreactor

Membrane bioreactors(MBR) have many advantages, but menbrane fouling remains a major obstacle to their wide application. This paper attempts to adopt the idea of combining the shortcut nitrification - denitrifying theory with IAMBR process, and study the membrane fouling behavior under the operating conditions of low C / N and efficient and stable nitrogen removal in the continuous operation of IAMBR. Through the determination of membrane fouling factors and the establishment of membrane resistance distribution, furthermore, performing the membrane cleaning experiment, then membrane fouling was able to be analyzed.The main experimental result and conclusion are as follows:(1) Darcy's Law was used to analyze the resistance distribution on the membrane surface. The resistance of membrane accounted for 13.56%, the cake resistance accounted for 82.61%, the reversible irresistance accounted for 3.46% and the irremovable accounted for 0.37%. The formation of fouling cake layer is the major reason that hinders the practical application of membrane bioreactor systemsys.(2) Membrane fouling on the membrane surface was observed and analyzed with SEM, FTIR and XRF. The result show that the organic pollutants were essentially composed of protein and polysaccharide, and the inorganic pollutants were scales mainly composed of Ca, P, S and Si compounds on the surface of menbrane.(3) Adopting the cleaning method combining physical cleaning and chemical cleaning can effectively remove the contaminants on the membrane surface and recover membrane flux. In the experiment the membrane flux can recover 99.63% of the full flux after rinsing with pure water followed by soaking for 5 hours with NaClO solution at a concentration of 3000mg/L.(4) After the membrane fouling behavior characteristics is analysed on the different working conditions, we found that cake layer is the main cause of the membrane pollution resistance which occupys 85%. We use MATLAB to build an artificial neural network based on the mechanism of cake-layer pollution to forecast the change rule of the membrane fouling over time on the constant flow rate(Q=150mL/min), and choose MLSS, aeration strengh, A/O, system running time t, initial transmembrane pressure TMP₀ as input parameters, t-test moment transmembrane pressure TMP_t for the objective vector. After the net training, the result shows that BP artificial neural network has a good approximation on the training sample function, and possessed a good extrapolating characteristics as well.