Abstract:
Hydrodynamic behavior in small-scale circulating fluidized beds (CFBs)was studied using the Isothermal Flow Analysis Program (IFAP), which has beendeveloped since 1980s based on K-FIX program. The program used finite differencetechnique to solve the mathematical models.In 1993 and 2001, Gidaspow and Therdthianwong [3] and Pantarak [4] usedthis program to simulate the hydrodynamic behavior of SO2 sorption and ozonedecomposition in CFBs reactors respectively. The flow pattern in both cases has beensuccessfully predicted. In this study, the two-dimensional closed-loop and open-loopconfigurations were applied to simulate the experimental works in small-scalecirculating fluidized beds [6] [7] [8] to verify the accuracy of the mathematicalmodels used in IFAP. In addition, the models have been used to study the effects ofrestitution coeffkient, gas phase turbulence, boundary condition of the particlestangential velocity, finite difference grid size and simulation time step.The predicted results fiom the closed-loop configuration showed anasymmetry of flow pattern relative to the experimental measurements due to twodimension.With this effect, it can be concluded that this configuration is not suitableto simulate small-scale CFBs. For the open-loop configuration, the predicted resultsare different. The core and annulus flow pattern, which is the phenomena observed inthe real CFBs, can be predicted. Nevertheless, the computed solid concentration atthe bottom region of CFB risers was about haft of the experimental values.It was also found that an increasing of the restitution coefficient in a rangeof 0.9 to 0.9999 enhances the segregation of particles near the wall. The gasturbulence term added to the model dissipated the accumulation of particles withinthe core region of CFBs riser and promoted the core-annulus flow pattern. Theboundary condition of the particles tangential velocity did not have a significanteffect on the flow behavior. Finally, the finite difference grid size and the simulationtime step were observed to have significant effects to the accuracy of simulatedresults. It is strongly recommended to verify these two parameters before beingapplied to simulate multiphase flow phenomena.