Open AccessHYDRODYNAMIC MODELS FOR SLURRY BUBBLE COLUMN REACTORS. FINAL TECHNICAL REPORT ALSO INCLUDES THE QUARTERLY TECHNICAL REPORT FOR THE PERIOD 01/01/1997 - 03/31/1997.
Author(s) -
Dimitri Gidaspow
Publication year - 1997
Language(s) - English
Resource type - Reports
DOI - 10.2172/7939
Subject(s) - mechanics , bubble , slurry , volume (thermodynamics) , particle (ecology) , superficial velocity , computational fluid dynamics , viscosity , reynolds number , viscometer , materials science , flow (mathematics) , bubble column reactor , thermodynamics , turbulence , chemistry , physics , composite material , oceanography , geology , gas bubble
The objective of this study is to develop a predictive experimentally verified computational fluid dynamic (CFD) three phase model. It predicts the gas, liquid and solid hold-ups (volume fractions) and flow patterns in the industrially important bubble-coalesced (churn-turbulent) regime. The input into the model can be either particulate viscosities as measured with a Brookfield viscometer or effective restitution coefficient for particles. A combination of x-ray and {gamma}-ray densitometers was used to measure solid and liquid volume fractions. There is a fair agreement between the theory and the experiment. A CCD camera was used to measure instantaneous particle velocities. There is a good agreement between the computed time average velocities and the measurements. There is an excellent agreement between the viscosity of 800 {micro}m glass beads obtained from measurement of granular temperature (random kinetic energy of particles) and the measurement using a Brookfield viscometer. A relation between particle Reynolds stresses and granular temperature was found for developed flow. Such measurement and computations gave a restitution coefficient for a methanol catalyst to be about 0.9. A transient, two-dimensional hydrodynamic model for production of methanol from syn-gas in an Air Products/DOE LaPorte slurry bubble column reactor was developed. The model predicts downflow of catalyst at the walls and oscillatory particle and gas flow at the center, with a frequency of about 0.7 Hertz. The computed temperature variation in the rector with heat exchangers was only about 5 K, indicating good thermal management. The computed slurry height, the gas holdup and the rate of methanol production agree with LaPorte's reported data. Unlike the previous models in the literature, this model computes the gas and the particle holdups and the particle rheology. The only adjustable parameter in the model is the effective particle restitution coefficient