NEURAL-NETWORK ANALYSIS OF THE DIFFUSIONAL LIMITATIONS IN ACTIVATED-SLUDGE FLOCS

This work deals with the development of a comprehensive but simplified version of floc model for the activated sludge process. The rate expr ession of the type used by the IAWPRC Task Group was employed in this study to represent the reaction kinetics inside the floc matrix. The i ntraparticle mass transport was then incorporated to account for the e ffects of the diffusional limitations. Solution to the flee model was obtained by orthogonal collocation technique. Since the actual reactio ns are not only related to the sludge floc size and the bulk character istics but ave also associated with the floc size distribution, the co ncept of overall effectiveness factor was therefore used. Evaluation o f overall effectiveness factors using neural network analysis was carr ied out and potential control strategies were discussed. Copyright (C) 1996 Elsevier Science Ltd.