Implementation of membrane bioreactor fouling models based on linear and bilinear autoregressive model structures

Abstract

The focus of this research is to create pragmatic and novel membrane bioreactor (MBR) models which can be applied for plant design, control and optimisation. Consequently, this research compares the traditional mechanistic models based on existing well known MBR filtration and biochemical processes with alternatives forms based on autoregressive input–output model formulations that in turn are based on system identification methods. Both model types are calibrated and validated using the same plant layouts and datasets derived for this purpose. This collated plant information included data obtained from carrying out standard flux-stepping experiments on a membrane filtration unit, and long term filtration experiments on a pilot MBR plant. In order to overcome the inherent deficiencies in any traditional approach, a novel alternative approach was tried in order to predict membrane filtration and fouling process for a MBR in a quick and easy manner. The rationale behind this novel approach is that it is simple to apply and that it does not require an intimate knowledge of the exact processes occurring in the MBR, so it could be applied by any non-specialist who was new to wastewater treatment modelling. This alternative approach uses linear and bilinear autoregressive model structures. Initial results from both the traditional and novel approaches indicate reasonable model predictive capabilities.

Keywords:

• Membrane
• Fouling
• Modelling
• Autoregressive

Acknowledgements

The author would like to thank and acknowledge the assistance received especially in terms of data collation from the following organisations: Process Control Water Software Systems at De Montfort University; and ITT Sanitaire (UK).

Notes

Presented at the Workshop on Membrane Fouling and Monitoring, September 21–22, 2012, Balliol College, University of Oxford, UK.