OCR Text |
Show INTRODUCTION The research described here is directed towards understanding the mechanisms through which nitric oxide is destroyed in both laminar flames and, possibly, in laminar flamelets, during reburning[l]. Specifically, we wish to explore the hypothesis that diffusion flame fronts sweeping through NO laden flue gases can destroy NO more efficiently than would be possible under well mixed conditions. This hypothesis is supported by the field data(2] shown on Fig. 1 , which indicated that, contrary to research results from rapidly mixed laboratory combustors[3,4] (solid line), it was possible to achieve significant NO destructions in full scale boilers (data points), under overall fuel lean conditions. The difference between full scale and pilot scale data is hypothesized to be that large scale flames in boilers produce reaction zones which are not premixed, but which, rather, consist of turbulent diffusion flames, which may, in turn, consist of stretched diffusion flamelets(5]. It was further hypothesized that the heterogeneous characteristics of diffusion flames can be effectively exploited, in order to achieve NO destruction by reburning under overall lean, and therefore, furnace benign, conditions. The current research is concerned with the destruction of NO in a laminar counter flow diffusion flame, a schematic of which is shown on Fig. 2. Of interest is the-destruction of NO contained in flue gases arising from a primary flame zone. Diffusion flames, such as the one shown on Fig. 2, are non-isotropic, and, therefore, contain a range of oxidizing and reducing environments. Even when the overall stoichiometry is fuel lean, there will exist local environments that are indeed fuel rich (Fig. 2). The key issue addressed was to determine whether the presence of these local rich regions, or inhomogeneities, could be exploited in order to allow substantial NO destruction under overall fuel lean diffusion flame environments. As shown on Fig. 1, previous bench scale[l], and pilot scale[4], research has indicated that NO destruction under reburning conditions can occur, in significant amounts, primarily when the overall stoichiometry is fuel rich. In diffusion flames, the overall stoichiometry has very little to do with local NO destruction processes. A counter-flow diffusion flame that is experimentally realized in the laboratory, is somewhat different from an hypothetical stretched flamelet. First, both streams originate from two parallel plates a fmite distance apart, as shown in Fig. 3, rather than infinitely far apart. This can, and has been, incorporated in a model of the opposed jet flame configuration by Corley and Wendt[6]. Second, the actual experimental burners are not infinitely wide, but have a finite diameter, beyond which the flame, if it exists, is no longer flat. Therefore, anyone - dimensional model based on infinitely wide burners would account neither for edge effects, nor for any non- one-dimensional effect present in the experiment. Initially, it was thought that an important parameter determining the net destruction of NO in the laboratory flame, would be the fraction of NO that was transported into the reaction zone. Clearly, it is unlikely that the portion of NO, which is transported away from the reaction zone, without having an opportunity to diffuse unto it, will be destroyed. Therefore, it seemed reasonable to explore the consequences of varying how the NO in the oxidant (simulating flue gas) contacts the diffusion flame front. |