OCR Text |
Show Given the general mechanism for solids combustion as shown in Fig. 2~ given the wide variety in fuel particles and consequent reactivity as shown in Table 1, and given the typical furnace dimensions for low grade solid fuel systems, the approach taken was to utilize a thermodynamic model as the basis for SFCOMB development. Kinetic models were considered to be inapplicable to this effort. Specific mechanistic pathway models were then evaluated as they described the formation of particulates [16, 60], oxides of nitrogen [4, 11, 12, 31, 57, 63], PAR'S [7, 35, 59], chlorinated dibenzo-dioxins and chlorinated dibenzo-furans (COD and COF) [1, 3, 9, 21, 23, 33, 34, 49, 58] and trace metals [2, 8]. Such models were analyzed for consistency with the general combustion approach shown in Fig. 2, and with particular attention to applicability to grate frred systems. This analysis led to the adoption of the particulate formation mechanism shown in Fig. 4, which is an adaption of the Friedlander [16] and Tuttle [60] analyses. Particulate can come from inerts carried away in the gas stream, from carbonized fine fuel particles (char), or from condensation of long chain volatile pyrolysis products. The NO. formation mechanism adopted is shown in Fig. 5, as adapted from Beer et al. [4]. While it recognizes the potential for thermal NOx according to the Zeldovich mechanism, this approach places considerable emphasis on conditions leading to oxidation of reduced nitrogen bound in the fuel. The dioxin mechanisms considered to be most relevant include the formation of COD and COP species include the reaction of precursor compounds, typically downstream of the combustion system and in the presence of flyash [3, 21,45], as shown in Fig. 6. Conditions which favor dioxin formation include the presence of significant quantities of chlorine in the gas phase, and also significant quantities of flyash to promote the necessary reactions. Increasing concentrations of flyash tend to increase the quantities of COD and COP compounds formed [4, 21]. Pathways describing the fate of trace metals have been well delineated by Barton et. al. [2], and by Brittain, Lamoreaux, and Lau [8]. For all metals, combustion temperature detennines fate -- specifically whether it remains in the bottom ash or leaves with the gaseous products of combustion. 2.2. Overall Approach to Model Development Given the mechanisms considered most appropriate, an overall approach was constructed for the development of SFCOMB. This approach involved selection of a base model, and identification of methods for adapting that model to the problem of low grade solid fuels. Base Model Selection. The model chosen as the starting point was the Gordon-McBride (G-M) model described by S. Gordon et. al. [20]. This model calculates flame temperatures and products of combustion assuming that the entire set of reactions goes to equilibrium. The approach taken is Gibbs Free Energy minimization. The data base in the model is the JANAF tables. Essentially the model tears down all reactants to atoms, and then constructs the products of combustion. The G-M model has been shown to have wide applicability. Ellis et. al. [15] have demonstrated its applicability to the combustion of hazardous wastes. Tillman and Anderson [54] demonstrated its utility in analyzing the combustion of wood wastes, the most dominant fonn of biomass used for energy. The model constructed by Stanford University [8] utilizes a similar approach. 5 |