Membrane Bioreactor Process Modeling and Optimization: Ulu Pandan Water Reclamation Plant

Abstract

The operating energy efficiency of the Ulu Pandan membrane bioreactor plant is optimized by artificial neural network (ANN) and bioprocess models. The ANN model mines historical plant data to uncover optimal operating settings. Historical plant data indicate that adjusting the membrane scouring aeration cycle will lead to direct energy savings. The ANN model concurs and shows the same correlation. Changes to plant operations carry substantial risks to the stability of the plant and place limitations on the range of operational variations. A plant risk assessment is conducted to ascertain the risk proposition for the adjustment of operating parameters. The bioprocess model investigates the underlying biological treatment mechanisms to identify the impact of the solids retention time on the volatile suspended solids, soluble microbial products, and endogenous decay coefficient. Results of the modeling show qualitatively good agreement with measured operating data. The concentrations of volatile suspended solids, soluble microbial products, and endogenous decay coefficient are shown to be sensitive to the system solids retention time.