Modeling of a Membrane Bioreactor using Bayesian Probabilistic Networks

Authors

Christoph Thiemig, Matthias Schubert , Johannes Pinnekamp and Michael H. Faber

Abstract

Membrane Bioreactors (MBR) are applied for municipal wastewater treatment since mid 1990s. Their number increased rapidly during the last years, but there is still a very limited experience about their behaviour and treatment efficiency during critical load situations. In the present paper a methodology to predict the performance of an MBR under different load situations is presented. The data from a six month intensive measurement campaign at a municipal MBR has been statistically evaluated utilizing Bayesian Probabilistic Networks. This methodology enables the investigation of the system performance in regard to the effluent quality under different operating conditions and different influent load situations.

Published in/by

Proceedings to IWA Regional Membrane Conference, Moscow, Russia, June 02-04, 2008.