OCR Text |
Show chemical kinetics are faster in the presence of CO such that the increased mixing caused by higher injection momentum was beneficial. In the absence of CO, increases of injection momentum with the accompanying diluent nitrogen lead to reductions in efficiency as the reactants pass through the temperature window before they can react. Figure 9 presents measured and predicted ratio of final to intial NO for SNCR in the UFurnace with initial conditions of 500 ppm NO, NH3INOj = 2.0 and various levels of CO. The combined turbulent mixing/chemical kinetic model can accurately capture the effect of CO on NO reduction by SNCR. As shown in Figure 9, the addition of CO to the reagent mixture initially increases the reduction of NO and subsequently reduces the reduction of NO (as indicated by higher NOp'N0i levels) as additional CO is injected. Ammonja Slip Figure 10 presents a comparison of measurements and JASPER model predictions of ammonia slip for conditions of low quench rate (-400Kls), an injection temperature of 1267 K, an initial NO level of500 ppm and an initial ammonia level of750 ppm. In this figure, NH3 slip is plotted versus the amount of CO that was injected with the reagent. The model predicts the data very well except for conditions in which very little CO is injected. It is believed that the measured levels ofNH3 slip for these conditions is too low since NO reductions were quite low and since these were the first measurements made with the FTIR whose sampling train could have inadvertently captured NH3 in the initial measurements. The ammonia slip predictions are therefore quite good. Injection Temperature and Quench Rate The JASPER model can also capture the effects of variable reagent injection temperature and variable quench rate on the effectiveness of the ASR process. Figure 11 presents data and model predictions for two different reagent injection temperatures (1300 K and 1265 K) and two different quench rates (-400 Kls and -2400 K/s) for conditions of an initial NO level of 500 ppm and an initial NH3 level of 750 ppm. Variations were made in the reagent jet to product stream momentum ratio by addition of various amounts of nitrogen to the reagent jet, and, in all cases, no CO was injected with the reagent. The JASPER model predicts the N0f'N0j levels and trends very well. In each case the reduction efficiency drops as the reagent jet momentum increases. The best reduction is achieved for the high reagent injection temperature and low quench rate cases with decreases in the injection temperature and increases in the quench rate leading to higher fmal NO levels. This is due to the fact that overall conditions (considering the quench rate and mixing delay) for these measurements and predictions are on - the low temperature side of the temperature window for effective reduction of NO presented in Figure 2. A surprising result of the experiments indicated that the effect of momentum ratio was the opposite for cases in which CO was injected with the reagent stream. When CO was present, increases in reagent injection momentum increased the NO reduction efficiency (as opposed to decreasing NO reduction as indicated in Figure 8). This is due to two factors. The presence of CO effectively increases the rates of SNCR reactions by two mechanisms: increasing local temperatures and increasing net production of OH radical which is necessary for the initiation of SNCR chemistry. The first effect is due to the exothermicity of the main CO burnout reaction, Page 7 |