OCR Text |
Show of NO, which was introduced only through the inner tube. Conversely, in Configuration D, NO was added into the oxidant through only the outer annulus. Oxidant, simulating flue gas, was uniformly distributed throughout the entire burner. The diameter of the inner tube was carefully selected such that the streamline starting at its outer edge reached the outer edge of the flat flame, just before the flame began to curl upwards (Configurations A through D employed no annular quenching of the reaction zone). Little of the NO introduced in Configuration D, therefore, penetrated into the flat region of the diffusion flame, and in this arrangement, any reduction of NO occurred outside the stretched flat flame. The influence of the flat flame alone on NO reduction could be inferred from the difference between the results of Configuration A and those of Configuration D. To produce a truly flat diffusion flame (Configuration E), both upper and lower burners consisted of coaxial tubes. Fuel and oxidant were introduced only through the inner tubes, while curtain nitrogen flowed through the outer annuli, to quench reactions before the flame had a chance to curl upwards. A circular sheet of flat, laminar, counter flow diffusion flame was thus produced around the center axis. In this configuration, NO was doped into and uniformly distributed in the oxidant stream in the inner tube of the lower burner. Experimental Conditions In all the results presented below, the flame was operated at an overall stoichiometric ratio of 1..2, i.e. under oxidizing conditions, where rebuming is normally not effective in premixed systems. The oxidant consisted of 41.8% O2 ,12.1 % CO2, and 46.1 % N2, doped with various amounts of NO, and entered from the bottom burner, at a (nominal, cold) velocity of 0.0542 mls. The fuel consisted of 17.3% CH., and 82.7% N2, and entered from the top burner at a (nominal, cold) velocity of 0.0544 m/s. The burner spacing was 0.0227 m. Sampling and Analysis To evaluate overall reburning effectiveness, cup-mixed, integral samples were taken from the exhaust. In addition, axial species profiles, within the flame, were measured by withdrawing samples at different axial locations between two burners, close to, but not at the center streamline. The in-flame sampling was accomplished by use of a quartz probe mounted on the 2-D positioner, and multiple samples withdrawn at various radial locations, ensured that the flame was flat where the axial profiles were measured. The quartz sampling probe was made sufficiently small (0.00075 m in diameter), and sampling rates were sufficiently close to being isokinetic with respect to the (calculated) radial velocity, that there was no visual disturbance on the flame while sampling. After withdrawal, sample gas was immediately sent through an ice bath to remove moisture content before analysis. Any N02, present in the sample, was converted back to NO in a Molybdenum converter. NO was measured by chemiluminescence. NH) and HCN were captured by de-ionized water in the bubbler and subsequently analyzed by Orion 95-12 ammonia gas electrode and 95-06 cyanide ion electrode, respectively. Oxygen, hydrogen, nitrogen, carbon monoxide, carbon dioxide, methane and C2 hydrocarbons were analyzed using a Hach Carle 01311-SP Gas Chromatograph, equipped with thermal conductivity and flame ionization |