| dc.description.abstract | Fluidized bed (FB) combustion has as main feature its capacity to reduce the release of pollutants in relation to conventional burning methods of pulverized fuel. Brazil has a potential in energy generation with FB through the combustion of coal, given the number of available reservoirs, of which 99% are located in the Southern states of Rio Grande do Sul and Santa Catarina. Among the available technologies for combustion and gasification on FB, we can highlight that of Circulating Fluidized Beds (CFB), given its use in Integrated Gasification Combined Cycles (IGCC). In the project, design and operation of FBs, the understanding of the gas-solid multiphase flow is highly important. Glicksman’s scaling laws provide the guidance needed for building beds in scale with fluid dynamics similarity, allowing the reproduction in pilot or even laboratory level of the fluid dynamics of a bed in industrial level. Along with that, Computational Fluid Dynamics (CFD) has established itself as a powerful tool in the simulation of CFB processes. Therefore, the aim of this paper is to develop a computational model for the simulation of CFBs fluid dynamics, using the MFIX code and to apply this model to the validation of scaling rules through the numerical modeling of a CFB in real scale with experimental validation and a bed in reduced scale according to a reduced set. For that to happen, a bed in fully correspondence wtih the reduced set, six scaled beds with alterations in their operational parameters, and a bed scaled by the full-set of Glicksman’s scaling laws have been developed. The computational model is based on the Euler-Granular Approach, in which the solid and gas phases are considered as interpenetrating continua. The solid phase is modeled as a fluid whose tensors are built in order to describe the flow of the granular phase according to the kinetic theory of granular flows (KTGF). The approximate physical model for the validation of this simulation to real scale was the Third Challenge held by NETL and PSRI. A comparison was made among the scales, one of the horizontal and vertical profiles of gas volume fraction; horizontal of vertical dimensionless speed of solids and dimensionless mass flux of solids; vertical of dimensionless pressure drop and the temporal evolution of the average gas volume fraction. In this study, the results allowed to verify that, in the numeral modeling of a reduced scale based on the reduced set of Glicksman’s scaling laws, the average of Relative Error (RE) considered over all the analyzed profiles showed a 14.2% value in relation to the real scale, which is acceptable for this kind of system. It has also been verified that the reduction in diameter of particles which were not suitable with the scaling laws, in some of the analyzed profiles, resulted in a reduction of RE when compared to that obtained through the results of the reduced set, due to a larger approximation of the particles diameter to the value determined by the full-set. Regarding the comparison of the results obtained through the reduced and full-set of Glicksman’s scaling laws, the most expected was confirmed – a larger matching for the bed scaled through the full-set, highlighting the correct prediction of the horizontal profile of the dimensionless mass flux of solids, which was not predicted by the reduced set. Thus, the reduced set of Glicksman’s scaling laws provides, within its inherent limitations, a good approximation for the scaling of CFBs through the Euler-Granular numerical simulation. | en |
