| Description |
A significant amount of data has been collected on sulfur capture by calcium-based sorbent injection based on the reaction: 2Ca0 + 2S02 + 02 • 2CaS04. The most promising method for extrapolating these data to determine performance in any one specific combustion scheme seems to be through the application of a predictive computer model. Based on the analysis of available data, the controlling mechanisms of S02 removal by dry sorbent injection have been identified to be the turbulent dispersion rate of the particle-laden jet, and the diffusion controlled heterogeneous rate of reaction. A grain model has been developed and coupled with S02 reaction and particle dispersion sub models, which were then integrated into an existing turbulent reaction flow model. This model was then used to analyze mixing and reaction data from a US EPA bench-scale reactor. The model is used to predict several variables based on two different particle injection geometries, straight and venturi throats. The process gas flow rate was varied from 30 to 100 Umin in the two geometries in order to see the effects of mixing and turbulence on S02 conversion. The reaction tube was modeled using a 40 x 36 grid, with increased resolution near the injection port. The modeled reactor was assumed to be adiabatic with constant gas temperature of 1323 K and sorbent entering at 296 K. The results predicted by the model were found to be within experimental error of the laboratory data. Differences in conversion rate between the venturi and straight throat designs are most easily attributed to different rates of sorbent mixing in the apparatus. The application of this predictive computer model to a bench-scale reactor demonstrates its usefulness in examining the controlling mechanisms involved in sulfur capture. The comparison made with measured data demonstrates that a reasonable description of the important mechanisms that control S02 removal by sorbent injection include the diffusion limited reaction rate and the turbulent particle dispersion rate. The value of such a model lies; in its ability to use fundamental information gathered in laboratory systems and apply the results to various full scale combustion applications. |
