OCR Text |
Show Figure 5 shows the results of combustion tunnel experiments in which methanol was injected in to flue gas produced by burning either natural gas, or natural gas doped with SDe, or coal, and the conversion of NO to NOe was measured. It is to be remembered that in these experiments methanol was used in 111 ratio to NO, i.e. no methanol was supplied for the concurrent reduction of 5~. COMPUTER MODELING RE5ULTS Figure 6 shows the results of computer·.Modeling calculations for the oxidation of NO by CH~OH in the absence of 503 while figure 7 shows the computer model's predictions for the reduction of S03 by methanol in the absence of NO for varying temperatures and reaction times. Figure 8 shows the model's prediction for the relative rate of 503 reduction NO oxidation, i.e. Figure 8 is a log log plot of S03 remaining versus NO remaining. Discussion The results in Table 2 clear show that the prediction's of the computer model were correct: methanol does indeed rapidly and selectively reduce 503 to 50e • This success of the modeling in predicting a previously unknown reaction is both scientifically gratifying and potentially of practical importance. In the Thermal DeNOx process NH3 is injected into flue gas at the point in the boiler or furnace at which the temperature is approximately 9500C and NO is reduced to Ne via a homogeneous gas phase reaction (16). An analogous process with metha~ol seems entirely possible, a process in which 503 is reduced to SOe, and NO converted to NOe • The NOe could then be removed by the same scrubbers which are used for SOe control and the corrosion problems associated with 503 would be avoided. Methanol injection could also be used downstream of the Thermal DeNOx process, or the very similar Urea Injection process, to prevent fouling due to N~HSO~. NO which escaped reduction in the Thermal DeNOx process would be converted to NOe and removed by the SOe scrubber, thereby providing a very efficient method of NOx control. The results obtained in the combustion tunnel i.e. in a pilot plant scale combustion system, support the potential importance of methanol injection. There is, however, an apparent conflict between the laboratory results and those obtained in the combustion tunnel: while the former showed no effect on NO to NOe conversion from the presence or absence of SOe, the latter shows a limited but real decrease in NO conversion when the fuel contains sulfur. The explanation for this seeming conflict is, of course, that the combustion tunnel experiments were done at a CH3 0H/NO ratio of 1. Combustion of a sulfur containing fuel produces a flue gas that contains both SOe and SOa and the competition between NO and SOa for the limited supply of methanol resulted in decrease NO conversion. The model redicts that there will be a range of temperatures about 2000C wide within which the NO to NOe and 50e to 503 conversion can occur and that this "temperature window· will move to higher temperatures as the reaction time is decreased. While the experimental data confirm this with respect to NO, the model shows the temperature window at 0.1 seconds reaction time to be centered at 8500 C while experimentally the 0.1 second window is centered at 725o C. Why this difference of 125o C? Conceptually one can divide the model,i.e. the reaction mechanism, into two parts, one part being a simple non-branching chain reaction consisting of the following steps: CH~OH + OH = CHeOH + HeO CHeOH + De = CHeO + HOe HOe + NO = OH +NOe HO. + 50a = HSOa + De |