A numerical study of a cold-mode fluidized bed reactor designed for pyrolysis of plastics

Numerical simulations have been conducted for a cylindrical, laboratory-scale fluidized bed reactor, which has been designed for fundamental research of plastic pyrolysis. The Euler-Lagrange multiphase particle-in-cell (MP-PIC) approach has been used for modeling the dense particle-load flow. As the performance of the fluidized bed heavily relies on the interaction between gas and particles, the cold flow model considering only inert quartz sand particles as bed material and nitrogen as fluidizing agent at atmospheric condition has been applied in this work. The objective of this work is to study the hydrodynamic behaviors of the fluidized bed, which is of significant importance for an efficient mixing between the bed materials in terms of heat and mass transfer.
In accordance with the experiment, the simulations have shown the bubbling fluidization regime in most considered cases, showing intermittently rising bubbles from the ground of the reactor to the upper surface of the fluidized bed. The calculated pressure drop Δp yields a good agreement with measured data, which remains constant with varied superficial gas velocity uG and increases with the bed inventory mS. ... mehrThe bed height hB increases slightly with increased uG due to the formation of larger bubbles. In addition, hB increases almost proportionally with mS due to the higher amount of sand mass.
While up-scaling the reactor at constant Δp and uG, the fluidized bed is dominated by large-scale bubbles for relatively small reactor diameters at dR = 3 cm and 5 cm, where the bubbles have a similar size compared with dR. However, multiple-column bubbles rising along off-centered axes have been detected for further up-scaled reactor with dR = 10 cm. In this case, the parallelly rising large-scale bubbles results in a recirculation of the gas flow and accumulation of the sand particles along the centerline axis, leading to an improved mixing. In summary, the simulations are able to reproduce the behavior of the fluidized bed observed in the experiments, which confirms the validity of the used computational approach. The correlations between the operational characteristics and the operating parameters obtained from the simulations can be used for further optimized design of the fluidized bed, e.g., concerning scale-up of the reactor.