The influence of hydraulic retention time on cake layer specifications in the membrane bioreactor: Experimental and artificial neural network modeling

The fouling control mechanisms were elucidated in the membrane bioreactor (MBR) by investigating the cake layer specifications in the different hydraulic retention times (HRTs). In this study, petrochemical wastewater was used. The sludge particle size distribution (PSD), excitation-emission matrix (EEM) fluorescence spectra, compressibility cake layer, Fourier transform infrared spectroscopy (FTIR) profile and extracellular polymeric substance (EPS) were measured to determine cake layer characteristics. The results showed that the particle size in the cake layer decreased with reduction in HRT while EPS concentration and transmembrane pressure (TMP) slope increased by time. The EEM florescence spectra of the cake layer showed the existence of two obvious protein-like substance peaks at the wavelength of Ex/Em of 290/355 and Ex/Em of 230-240/355 nm at different HRTs. Furthermore, a feed forward artificial neural network (ANN) was trained using back propagation algorithms for prediction effluent chemical oxygen demand (COD) and TMP. The best structure was a trainlm network with two layers including 17 and 2 neurons in the hidden layer and output layer, respectively. Sensitivity analysis showed that the most and the least sensitive parameters on TMP were mixed liquor suspended solid (MLSS) and time, respectively.