(282e) Modeling and Simulation of Membrane Biological Reactors Using ANNs

Membrane biological reactor is a recent technology for wastewater treatment and water reuse. MBR combined settling tank and filtration unit in one single operating unit. However, the high efficiency of MBR in producing high quality disinfected water is adversely affected by the high potential of the membrane to retain foulants, which results in membrane fouling. Fouling is the accumulation of foulants on or within the membrane pores, which causes pore blocking and permeate flux decline. Furthermore, the various interactions among the process variables such as substrate concentration, biomass, aeration rate, dilution rate, operation mode, and membrane backwash scheduling increase the complexity of the system. In this paper, a mathematical model for a membrane biological reactor (MBR) is developed and further validated with experimental data. An artificial neural network model is then developed to alleviate the complexity of the mathematical model. The ANN model was shown to capture well the MBR process dynamics. ANN model is used for control of MBR in attempt to reduce fouling and optimize process operation.