Development of a decision tree for the integrated operation of nutrient removal MBRs based on simulation studies and expert knowledge

Autores: Dalmau, M.; Rodriguez-Roda, I.; Ayesa Iturrate, Eduardo José; Odriozola González, Juan Vicente; Sancho Seuma, Luis; Comas, J.
ISSN: 1385-8947
Volumen: 217
Páginas: 174 - 184
Fecha de publicación: 2013
Lugar: WOS
The combination of simulation studies and expert knowledge allows the development of a decision tree for the integrated operation of nutrient removal as well as filtration processes in membrane bioreactors (MBRs). In order to do it, a model-based methodology, including model development and use of system analysis tools, was followed to identify the best operational strategies for good filtration performance and efficient biological nutrient removal at lower costs in a University Cape Town (UCT)-MBR pilot plant. Activated sludge no. 2 and a series resistance model were used to simulate the biological and filtration processes, respectively. A sensitivity analysis was carried out for the identification of the most sensitive operational parameters, and then, their best setpoint ranges/values were identified through a scenario analysis. Simulation results were supplemented with expert process knowledge to overcome the current limitations of biological and filtration models. The results were organized in the form of a decision tree. The most sensitive parameters, and thus the first branch to explore, were operational parameters related to the filtration processes (i.e. flux and relaxation time). Based on expert knowledge, membrane-air scour was the following control action to check. Then, the decision tree evaluates the biological nutrient removal processes and modifies the necessary operational parameters, ranked according to their sensitivity. The most sensitive were aerobic and membrane aeration, aerobic and anoxic recirculations and carbon dosage. The last branch takes into account the operational cost, directly related to aeration and carbon addition. (C) 2012 Elsevier B.V. All rights reserved.