Multi-Scale Numerical Simulation On Hydrodynamic And Reaction Characteristic In Fluidized Bed Reactor

Gas-solid two-phase flow is widespread in nature and industrial application involving energy, chemical industry, metallurgy, materials and other fields. Due to the effects of heterogeneous meso-scale bubble/cluster structures in fluidized beds, the gas-solid two-phase flow behaviors and the change of state parameters become very complex. Fundamental knowledge of gas-solid flow dynamic behavior and multiphase reactive mechanism in fluidized beds is essential for the optimization and prediction of industrial problems.With the development of computer technology, numerical simulation has become one of the most promising approaches to study the gas-solid flow. To establish a reasonable and comprehensive mathematical model is a key to perform the simulation.On the basis of multi-scale structures of the dense phase in the form of clusters and the dilute phase in the form of dispersed particles in the grid cell, the relationship between accelerations and local structure parameters in the dense phase and the dilute phase is established and the non-uniform structure formation condition of the minimum of energy dissipation consumed by multi-scale drag force is proposed. Based on the bivariate extreme value theory, a cluster structure dependent(CSD) drag model is developed, which incorporates the effect of gas pressure gradient in the grid cell. Within the CSD drag model, the local structure parameters are obtained. Considering the effect of meso-scale structures on the hydrodynamics, chemical reactions, heat transfer and mass transfer, a muti-scale gas-solid flow-chemical reaction coupled model is developed under the framework of the two-fluid model.Hydrodynamics characteristics of gas-solid two-phase in high solid flux and low solid flux risers are simulated by means of the CSD drag model. The core-annular structure of particle flow and the heterogeneous structure characterized by clusters and dispersed particles are captured. The clusters are observed to form, move, aggregate and disperse into single particle. Predicted velocity and concentration of particles agree better with experimental results. Flow behaviors of clusters are also analyzed. With consideration of wall friction, gas-wall and particle-wall frictional pressure loss is incorporated into momentum conservation equations and the CSD drag model is modified.The results indicate that a higher wall friction will weaken the dissipation of fluctuation energy and promote the solid volume fraction near the wall. In addition, the accelerations are significant at a low solid volume fraction and the maximum amplitude is an order of magnitude greater than the gravitational acceleration. Hence, the accelerations can not be ignored during the solution of the CSD drag model.Considering the impact of clusters during chemical reactions, a multi-scale chemical reaction model coupling heat transfer and mass transfer is developed, which is incorporated into the multiphase reactive fluid dynamic model. The local structural parameters are provided using the CSD drag model. The regeneration reaction of Ni-based oxygen carriers in riser reactors is simulated.The results indicate that the clusters lead to a clear discrepancy in gas species concentrations between the dense phase and the dilute phase. The oxygen molar fraction in the dense phase has a significant decrease along the axial direction compared to that in the dilute phase. As solid volume fraction increases, the reaction rates in the dense phase and the interface appear an obvious growth and the change of reaction rate in the dilute phase is not evident.The mass transfer and heat transfer between the dense phase and the dilute phase have a significant impact on the distribution of reactant concentration. With no consideration of mass transfer and heat transfer between phases, the difference of reactant concentrations between the dense phase and the dilute phase becomes significant. The discrepancy of the temperatures in the dense phase and the dilute phase is not obvious.Chemical looping combustion processes in a dual circulating fluidized bed(DCFB) reactor are investigated by means of the computational fluid dynamics(CFD) method. The muti-scale gas-solid flow-chemical reaction model is employed to account for the effect of meso-scale heterogeneous structures. The simulations are performed to predict the gas-solid flow behaviors and reactive characteristics in the DCFB reactor. By comparisons of the outlet gas components and the axial pressure profile with the measured data, the model can obtain a reasonable prediction. The influence of frictional stress model on flow of particles is compared. The effects of solid inventories and the position of the upper loop-seal on the solid distribution are investigated. At a large solid inventory, the particles spread over the whole height of the reactor. When the solid inventory is at a low level, no real circulating regime is reached for the fuel reactor. Meanwhile, the impact of clusters on gas composition and temperature field is analayzed.Bubbles are the typical meso-scale structures in low-velocity gas-fluidized beds, and play similar roles as clusters in circulating fluidized beds(CFB). To describe the effects of such meso-scale structures in low-velocity gas-fluidized beds, a bubble structure-dependent(BSD) drag model is developed and incorporated into the CFD code. In this model, the effects of the solid pressure in the emulsion phase and the meso-scale added mass force are taken into account. The simulations of gas-solid flow behaviors in bubbling fluidized beds are performed and the model is evaluated by comparisons with experimental data.The solid pressure gradient and accelerations as a function of solid concentrations are analyzed. To further verify the model, three-dimensional simulations of bubbling fluidized beds are carried out. The electrical capacitance tomography(ECT) system is employed to measure the lateral distribution of solid volume fraction in the bed. By comparisons with measured data, the BSD drag model can obtain a better prediction than the conventional drag model.By means of the BSD and CSD drag models integrated with chemical kinetic, chemical looping reforming(CLR) processes in an interconnected fluidized bed system are simulated. The solid circulation pattern in the reactor is obtained and the effects of bubbles on flow behaviors, the distribution of gas compositions and temperatures are investigated. The results indicate the mesoscale-based drag model can predict a more reasonable result with experimental data. In addition, the influence of operating parameters involving temperature, velocity and operating preesure is also evaluated. To some extent, the hydrogen production can be promoted by reducing the operating temperature and increasing the fuel inlet velocity. A higher H2O/CH4 molar ratio promotes the water gas shift reaction, which is beneficial for the hydrogen production. A higher pressure produces an increase in the inlet gas flow, which reduces fuel conversion and hydrogen yield.