| OCR Text |
Show Figure 7 shows a comparison between the mean flue gas temperature profile in the regenerator for baseline and reburning conditions. The difference between the baseline and rebuming temperatures is smaller at the regenerator entrance and increases as it reaches the exit. The averag~ increase is expected to be approximately 250°F. Exhaust gas residence time in the checkers is about 2.5 seconds. The effects of temperature elevation and alternating oxidizing and reducing gases on refractory lifetime are a practical consideration for reburning design retrofits . Near the top of the checkerwork, brick temperatures increase by about 100°F due to reburn. Like the gas temperatures, the magnitude of the change in temperature increases further down in the checkerwork. At the top elevation, only the flISt 3.5 feet of regenerator packing exceed the baseline maximum temperature. NOx Control The reburning gas path temperature profile developed in the thermal analysi~ was used to estimate the potential NOx control achievable with gas reburning using a kinetic model of the reburning process, in the absence of mixing processes. For the container glass furnace, the key model inputs were: • • • • • Primary NOx Reburning Fuel Injection Temperature Reburning Zone Residence Time Reburning Zone Stoichiometry Burnout Air Injection Temperature 7lb/ton (1193 ppm, wet, 0 %02) 2850°F 0.7 seconds 0.90 2783°F The gas temperature at the point of rebuming fuel injection and the gas temperature at the burnout air injection point were based upon the thermal model results. To calculate the reburning fuel average residence time, it was assumed that the volume occupied by the combustion gas was equal to the port flue volume plus one half of the regenerator crown volume (to one foot above the checkerwork). This assumption was based upon the results of a two-dimensional computational fluid dynamics model which indicate that major stream tubes occupy about half of the regenerator crown and that a significant recirculation zone occupies the remainder of the headspace. For the container glass furnace, Figure 8 presents NO, HeN, NH3 concentrations in the reburning and overfire air zones as a function as a function of time. During rebum fuel injection, NOx emissions decrease rapidly, and then slower as the initial hydrocarbon radical pool is consumed. Some HeN, NH3 remains following the reburning zone. However, at these temperatures, NH3 or HeN do not survive the burnout air injection step. For nominal conditions, the idealized kinetic model predicts that NO emissions will be reduced from 1193 to 116 ppm, a 90 percent reduction. Sensitivity studies, summarized in Figure 11, show the impact of rebuming zone residence time and overfrre air injection temperature on final NOx' As the reburning zone residence time increases from 0.5 to 0.9, the estimated NOx reduction efficiency improves slightly from 90 to 91 percent. As the temperature at the overfue air injection point increases from 2700°F to 2800°F, the NO reduction efficiency falls from 92 percent to 90 percent. Rebuming Fuel mixing is a potentially limiting factor, and the reductions estimated by the kinetic model are usually optimistic. By comparing kinetic modeling predictions to bench and full-scale da~ it is possible to adjust the kinetic predictions for the impacts of reburning fuel and overfire air mixing. Based upon this comparison, it was estimated, conservatively, that 80 percent reduction 7 |
