Model evaluation of simultaneous nitrification and denitrification in a membrane bioreactor operated without an anoxic reactor

Sarioglu M., INSEL G., Artan N., ORHON D.

JOURNAL OF MEMBRANE SCIENCE, vol.337, pp.17-27, 2009 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Article
  • Volume: 337
  • Publication Date: 2009
  • Doi Number: 10.1016/j.memsci.2009.03.015
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Page Numbers: pp.17-27
  • Istanbul Technical University Affiliated: Yes


This study was conducted to evaluate and optimize a simultaneous nitrification/denitrification process in a membrane bioreactor operated with no separate anoxic volume. The membrane bioreactor was fed with strong domestic sewage and operated at steady state at a sludge age of 36 days and a corresponding suspended solid levels in the range of 17,500-21,000 mg L-1. The wastewater temperature in the MBR ranged from 20 to 25 degrees C. Despite the strong influent and moderate air supply, a dissolved oxygen concentration of approximately 1.8 mg O-2 L-1 was recorded in the bulk liquid. Simultaneous nitrification/denitrification was sustained due to limited diffusion of oxygen into the floc. Nitrification was only partial and the ammonia nitrogen profile always remained above 50 mg L-1 and essentially followed the fluctuating trend of the influent TKN concentration. The available nitrate was depleted with full denitrification, which enabled an average of 57 mg N L-1 of the influent nitrogen to be removed in the membrane bioreactor without the need for a separate anoxic volume. The adopted suspended growth model could be calibrated with the same set of stoichiometric and kinetic coefficients for all parameters. The model accounted for the diffusion limitation and the resulting simultaneous nitrification/denitrification in terms of the high half saturation constants and yielded a value of 1.75 mg O-2 L-1 for K-OH and 2.0 Mg O-2 L-1 for K-OA. The model calibrated for the experimental data was then used to optimize the MBR operation for nitrogen removal. The evaluation identified an MLSS threshold level of around 16,000 mg L-1, below which nitrogen removal was essentially controlled by denitrification and above which the rate limiting mechanism shifted to nitrification. The evaluation also indicated that controlling the bulk dissolved oxygen concentration at around 1-5 mg O-2 L-1 in the MBR would enable nitrogen removal efficiencies in the range of 85-95% for typical domestic sewage. (C) 2009 Elsevier B.V. All rights reserved.