| OCR Text |
Show The plasma gun system used in the radical injection process is described by Zhou et al.. In the hot injection study, this system was replaced with an apparatus for heating and injecting hot ammonia and argon (as the carrier gas) into the combustor, as shown in Figure 2. Argon was passed through a 1 ()() cm stainless steel tube whose temperature was controlled by varying the OC current supplied to the tube for direct Joule heating. The hot argon gas then passed into an injector designed to mimic the geometry, and therefore the swirling flow conditions, of the plasma gun. Cold NH3 was introduced tangentially into the injector, where it mixed with the hot argon gas before being injected into the combustor. The injector was positioned approximately 8 cm from the combustor, as was the plasma gun in the previous experiments. An electrically insulated nozzle connecting the injector to the combustor minimized the amount of ambient air entrained by the hot jet. In the plasma experiments, a larger guard box produced a similar effect. Finally, the entire hot tube and injector were insulated to minimize heat loss. The achievable temperatures of the gas exiting the hot tube apparatus were dictated by the limitations of the materials used. At peak operating conditions, the hot tube reached 1150 K with 740 W supplied by the electrical source. Only a portion of this power was actually used to heat the injected gases, with the remainder being lost thermally and radiatively. It was of concern whether the power input to the injection gases in the hot tube expe~ents was similar to the power input to the injection gases in the plasma experiments. In the subset of data selected from the plasma experiments for comparison, the power supplied to the plasma gun was 400 W. However, this power included losses through the cooling system, EMF losses, and others. At peak power in the hot tube experiments, the argon/ammonia stream entered the combustor at about 850 K. Based on this temperature and a mean specific heat and density of argon, the power actually carried into the combustor by the input gases was about 215 W. In the plasma experiments, the temperature at the injection section was reported as 920 K (Zhou, 1991), which would correspond to a power input of about 240 W. Thus, the power supplied to the injected gases in both experiments was comparable. Results Throughout the experiments, little or no difference was detected between the NO and NOx measurements, which is consistent with commercial measurements indicating that 900/0 or more of NOx emissions from utility boilers is NO. Since the combustion gases in the present experiments were doped with NO, emission reductions are reported as the percent NO reduction relative to a baseline of no ammonia injection. In reporting NO reduction performance, "flue gas temperature" is the temperature of the combustion flue gas at the centerline of the combustor, as measured approximately 5 cm below the ammonia injection port. 4 |
