OCR Text |
Show Introduction Almost twenty years after Lyon's patentl was issued defining the conditions under which NO can be reduced selectively to N2 by ammonia (excess air conditions with temperatures ranging from approximately 850 to 1100 °C), there continues to be industrial interest in the use of selective non catalytic reduction (SNCR) of NOx' Its potential as a low cost, effective and retrofittable NOx control technique has lead to many applications in utility boilers, waste incinerators and other stationary combustion equipment. However, the NOx removal efficiency -__ is equipment specific and depends on the time available for contacting the selective reducing agent within the appropriate temperature window. In addition to field demonstrations .of SNCR there has been considerable work at laboratory and pilot scale to identify the parameters controlling the SNCR process. Muzio et al.2 were the first to demonstrate that urea and other amines could be effective as selective reducinf. agents. The temperature range for selective reduction under fuel rich conditions was identified ,4. Perry and Siebers5 reported that cyanuric acid could be an effective selective reducing agent at temperatures as low as 600°C, although this effect was later attributed to catalytic activity involving the experimental reactor6• Chen et al.7 suggested that a hybrid NOx control scheme is possible by combining the use of selective reducing agents with combustion modifications such as reburning. Lyon8 evaluated the interaction of both hydrogen and CO with the selective reduction of NO by ammonia. Teixeira et al.9 evaluated the effect of trace combustion species on SNCR perfonnance and there have been several studies on the use of proprietary compounds to improve NO reduction in practical applications lO• Under well mixed plug flow conditions SNCR is capable of achieving very high reduction efficiencies8, in excess of ninety percent at NH3INO ratios of less than 1.2, but this level of perfonnance has not been achieved in practical applications. Under practical conditions, W,ical reduction efficiencies are less than fifty percent for a range of boiler types and fuels 11,12, I . The selective reduction process is effective over a very narrow temperature range (or "window"). Reduction at temperatures higher than the window is poor because the reducing agent creates NO by oxidation. If the temperature is less than the optimum window the selective reduction reactions are too slow. In existing combustion systems the spatial location of the optimum temperature window may vary with operating conditions or it may occur in regions of large thennal gradients typical of convective heat exchangers. This places severe design constraints on the reagent injection system which must disperse the reagent throughout the entire combustion product stream and mix the reagent with the NO while the gases are within the appropriate temperature window. . ' . . ' Although the importance of reactant mixing and the need to optimize injection location has been stressed8,IO,12, very few studies have quantified the impact of'mixing limitations on the perfonnance of the SNCR process. Muzio et al. 14 carried out a pilot scale study to quantify the effects of residence time on the SNCR process. They recognized that three characteristic times need to be considered: 1) mixing of the reducing reagent with the combustion products; 2) release of the reactive nitrogen compound from the reagent (which may be in aqueous solution), and 3) the chemical reaction times. Muzio et al. maintained constant mixing conditions and the observed residence time effects were attributed to the aggregate of the characteristic time for Page 2 |