EMPIRICAL MODELING OF N2O EMISSIONS FROM CIRCULATING FLUIDIZED-BED COMBUSTION

This paper describes the development of a first-generation empirical m odel for predicting N2O emissions as a function of significant circula ting fluidized-bed combustion (CFBC) operating parameters and fuel pro perties. The relative magnitude of the parameter estimates for the sig nificant operating parameters are racked temperature > excess air > SO 2/sorbent, which is consistent with data reported in literature. The S O2 concentration in the flue gas was substituted for the limestone fee d rate as a main operating parameter for modeling. Unlike the limeston e feed, the SO2 concentration was shown to be a significant modeling v ariable. The reason for the improved correlation is probably the resul t of (1) the SO2 concentration reflecting the level of sorbent utiliza tion and (2) the homogeneous chemistry of N2O being dependent on the m agnitude of SO2 emissions in the flue gas. Ranked in order of importan ce, the most significant fuel properties affecting the net formation o f N2O emissions were char-N > moisture > oxygen > total-N. The stabili ty of the char-N may be important for mass transport; i.e., the char m ay act as a vehicle for transporting fuel-N out of the highly reducing dense-bed region of the CFBC. This mechanism may account for the grea ter N2O emissions observed in CFBCs over their bubbling fluidized-bed combustion (BFBC) counterparts. The total coal-N flux was also include d in the model and is believed to be representative of the volatile-N fuel component in the model. The effect of fuel-O on N2O emissions is uncertain but may be attributed to rank effects when considered in con cert with fuel moisture (H2O). The parameter estimate for the fuel moi sture variable showed that H2O decreased N2O emissions. The moisture f rom the fuels is believed to be a source for H and OH radicals that ma y effectively destroy N2O through homogeneous gasphase reactions.