| OCR Text |
Show DISCUSSION A systematic investigation of the reasons for the significantly different behavior for the removal efficiency and H20 characteristics of urea and ammonia has not been conducted. However, in light of the differences, several possible explanations for the different behavior have been considered. One explanation that can be hypothesized is based on the difference in decomposition products between the two reagents. For urea, the primary decompositio products are thought to be HNCO annd NHi species--in roughly equal proportions. In this scheme, HNCO reacts to form an NCO, which then reacts with NO to eventually form N20. Introduction of CO into this reaction sequence could alter the path to yield the N20 and removal efficiency results observed. Ammonia, on the other hand, is not thought to decompose to HNCO, but only NHi. This would result in a different reaction path when CO is present and, ultimately, different removal efficiency and N20 characteristics from urea. Another possible explanation involves consideration of the the physical processes that lead to the release of ammonia and urea from solution. Based on earlier lab tests, it is known that urea is released from solution more slowly than ammonia. When injected into the high temperature environment, NH3 is very rapidly released to the gas phase. Urea, on the other hand, does not begin to be released until saturation is approached. This latter behavior results in the urea being introduced into the flue gases over a longer time frame. This, in turn, allows more time for the CO chemistry to intereact and, finally, yielding different removal efficiency and N20 results from ammonia. Practical Implications There are several practical implications regarding the influence of CO on the SNCR process. First, examination of Figure 2 shows that peak NO removal efficiency for urea (excluding the temperature regime below 8500C at high CO levels) reaches a maximum of 45' without added CO and decreases to about 30' as CO increases to 2000 ppm. In contrast (see Figure 6), ammonia has a peak removal efficiency of about 55-60' independant of CO level--even though the temperature window shifts with added CO. This represents a 10-30 point higher efficiency for ammonia. Urea, on the other hand, exhibits one desirable characteristic not present with ammonia: widening of the temperature window as CO levels increase. Second, comparison of N20 levels for urea and ammonia (Figures 5 and 9) show significant differences for the two reagents. For ammonia, N20 levels are typically less than 10 ppm, or 15-20 ppm lower than for urea. This advantage tends to narrow at high CO addition levels when the temperature is low, i.e., in the 750-850oC range. Third, for either reagent, it is clear that a key to maintaining low N20 is keeping CO as low as possible at the point of injection of the SNCR reagent. This has important practical implications because CO levels at SNCR injection temperatures can be an order of magnitude higher than CO levels measured at typical instrumentation locations (e.g., econimizer outlet) for control purposes. This behavior has been confirmed by actual measurements on several 11 |
