@article {Stensel:2000:1938-6478:633,
	author = "Stensel, H. David and Horne, Garth",
	title = "Evaluation Of Denitrification Kinetics At Wastewater Treatment Facilities",
	journal = "Proceedings of the Water Environment Federation",
	volume = "2000",
	number = "13",
	year = "2000",
	abstract = "The use of anoxic-aerobic activated sludge systems has steadily increased due to the ability to provide economical nitrogen removal and/or the ability to produce sludge with consistently good thickening properties. The anoxic reactors that precede the aeration tanks may be single or multiple-staged reactors, and their design includes mass balances on nitrate-nitrogen and terms that account for the biological denitrification rate. Two methods have been used to quantify the denitrification rate. The first is an empirical method that relates the specific denitrification rate (SDNR) to the BOD F/M loading to the anoxic reactor. The second approach uses mechanistic modeling based on Monod or Michaelis-Menten kinetics that relates biological growth rates and nitrate reduction rates to the reactor readily biodegradable COD (RBCOD) concentration. In this approach a term η is incorporated in the model to account for the fraction of heterotrophic biomass that is capable of nitrate reduction. A model default value of 0.37 has been suggested for the IAWQ activated sludge model. In this research a field experimental reactor was used to evaluate SDNRs at four different municipal activated sludge plants as a function of different anoxic zone hydraulic retention times (HRTs), activated sludge solids retention time (SRT), biodegradable COD loading, and reactor RBCOD concentrations. For each site the active biomass fraction of the MLVSS was estimated from the site operating conditions and wastewater characteristics. In addition laboratory batch endogenous respiration tests were carried out to observe differences in equivalent oxygen consumption rates under aerobic and anoxic conditions in an attempt to quantify η values. The fraction of active biomass in the mixed liquor for each plant was determined and used to develop a relationship between SDNR and RBCOD concentration. A Monod model half-saturation coefficient and maximum specific growth rate coefficient for heterotrophic bacteria growth on RBCOD was determined as 8.6 mg/L and 3.6 g/g-d, respectively. The estimated fraction of active biomass capable of denitrification varied and was a function of the plant design, degree of nitrate removal, and relative amount of soluble substrate removed with nitrate as the electron acceptor in place of oxygen. The η values for three of the plants (0.85, 0.67, and 0.52) were higher than the IAWQ model default value of 0.37. The plant with the lower value (0.30) was operating with minimal nitrate removal. An empirical equation relating SDNR to anoxic tank F/M loadings could provide conservative estimates of SDNRs at lower loadings but the equations are of limited accuracy.",
	pages = "633-654",
	url = "http://www.ingentaconnect.com/content/wef/wefproc/2000/00002000/00000013/art00050",
	doi = "doi:10.2175/193864700784607910"
}