Effects of the Choice of Gas on the Hydrodynamics of Fluidized Beds

Journal article, 2019

Using a cold-flow model, this work examines how the hydrodynamics in the dense region of a fluidized bed are affected when the gas acting as the fluidizing agent is changed. While the focus here is on the use of a gas other than that suggested by scaling laws in cold-flow models, the study also has relevance for fluidized bed processes in which different fluidization agents can be used, such as drying or coating. For cold-flow models, the scaling criteria for the hydrodynamics of the fluidized bed prescribe certain properties of the gas to be used as the fluidizing agent. In certain scenarios, air does not match the suggested properties, and other gases must be used, which increases the complexity and cost. In the worst-case scenario, no gas will fulfill the prescribed properties, and experimentation under strictly scaled conditions will be impossible. Assessing the impact of not using the correct gas allows researchers to evaluate the reliability of their findings when there is no compliance with the scaling laws. In this work, experiments were carried out in a hydrodynamically down-scaled model of a 100 kW chemical looping combustion (CLC) unit, which under hot conditions contains metal oxide particles fluidized with steam. If the hydrodynamic properties are to be resembled at ambient temperature with the same solids, a gas lighter than air, e.g., helium, must be used according to the scaling laws. This entails an experimental setup with gas recycling, adding cost and complexity to the system in comparison to using air. This work investigates how fluidization with air, instead of helium, affects the hydrodynamics of the cold-flow model based on two different approaches: (i) maintained superficial gas velocity and (ii) maintained decay constant in the splash zone. The results show that air does not adequately substitute for helium in a bubbling bed with respect to the following key hydrodynamic properties: pressure fluctuations were 30% lower; the bubble fraction was up to 36% smaller; bubble frequency was up to 17% lower; and the solids concentration was up to 10% higher. It was also found that the use of air yields a poorer horizontal distribution of the gas injected from the reactor side-walls, which affected the cross-sectional distributions of the solids concentration, bubble fraction, and bubble frequency.