An Analysis of the Results of Studying the Separation of Particles in the Upper Part of CFB Apparatuses

Abstract

Matters concerned with the motion of solid particles in circulating fluidized bed (CFB) apparatuses are considered. It is shown that a significant portion of particles does not enter the cyclones and returns into the furnace, thereby producing an internal circulation loop. The particle capturing efficiency, i.e., the downward to the upward flow rate ratio, depends on the geometrical parameters at the furnace outlet to the cyclone and on the physical characteristics that can be expressed in terms of the Stokes number. The ratio of the outlet port area to the furnace or CFB reactor area is the most important parameter. It has been shown that in analyzing and comparing data, it is also necessary to take into account other geometrical indicators, such as possible influence of density and mass flows, as well as methods for determining them. Measured upward flow rate values, as for example those obtained in the authors’ own experiments, can be the most representative ones. The experimental data obtained at the All-Russian Thermal Engineering Institute (VTI) under different flow outlet conditions from furnace models were analyzed. The upward and downward mass flow rates near the outlet port and the flowrate of circulating material captured in the separator are compared. The authors’ own studies have shown that the average size of sand particles increases to some degree with increasing the air velocity. Therefore, all previously obtained results were recalculated in this study with taking into account the updated upward flow rate and decrease in the particle sizes at low gas velocities, and compared with the experimental and calculated data reported in the literature sources. The values the authors obtained on the test bench facility with a cross section of 0.4 × 0.4 m lie by approximately 20–30% below the calculated values. The same also relates to other conditions of flow outlet from the furnace, except for the data on the boiler and data at which the calculated values are very close to the experimental ones. A correction of the calculation relationship is proposed.