Modeling hydrodynamics of gas–liquid airlift reactor

Abstract

This article deals with the physical modeling and numerical simulation of two-phase bubbly flow in an airlift internal loop reactor. The objective is to show the ability of computational fluid dynamics (CFD) to correctly simulate hydrodynamics and axial dispersion in such a bubbly reactor. The modeling of two-phase bubbly flow is based on the so-called two-fluid model derived from Reynolds-averaged Navier-Stokes equations in two-phase flow. From the local perspective, CFD leads to the distributions of phases, interfacial area, and velocity field in the whole volume of the airlift. Numerical simulations are discussed after comparison with experimental data. Sensitivity analysis is then presented to highlight the main parameters that must be taken into account in two-fluid modeling, especially in terms of interfacial transfer of momentum and turbulence. Once local hydrodynamics has been discussed and validated, the axial dispersion is then addressed. The axial dispersion coefficient is estimated from simulation of transport equation of salt concentration. Given the time evolution of the concentration, measured by a ''numerical'' probe located in computed airlift reactor, it is possible to numerically estimate the axial dispersion in the airlift, in the same way as in the experiments. The axial dispersion coefficient determined after simulation is compared with the experimental one and the ability of CFD to simulate mixing time and axial dispersion is shown. In addition, a physical analysis of axial dispersion is proposed.

A new numerical method for axial dispersion characterization in microreactors

“…This post-processing can lead to significant loss of information. Talvy et al (2007) proposed an alternative method based on the expression of the different contributions to axial dispersion as function of the velocity and the concentration fields (i.e. molecular diffusion, temporal dispersion and spatial dispersion) (Talvy et al, 2007;Le Moullec et al, 2008).…”

“…Talvy et al (2007) proposed an alternative method based on the expression of the different contributions to axial dispersion as function of the velocity and the concentration fields (i.e. molecular diffusion, temporal dispersion and spatial dispersion) (Talvy et al, 2007;Le Moullec et al, 2008). Therefore, axial dispersion coefficient can be estimated from simulations solving the continuity equation, Navier-Stokes equations and the transport equation of a tracer.…”

Experimental measurement and CFD simulation on the hydrodynamics of an internal-loop airlift reactor

“…As of current, no universal set of closure models have been concluded for bubbly flow in an airlift reactor although a diverse set of closure models have been proposed across literature [19][20][21][22]. These closure models which composed of interfacial momentum forces can be divided down further as drag and non-drag forces.…”

“…Hence, comparison studies have been conducted over the decade to evaluate the sensitivity of these drag models on local hydrodynamics. A study carried out in an internal-loop airlift reactor have shown superior performance through the drag model Karamanev and Nikolov (1992) as a function of Reynolds number when compared with other models governed by Eötvös number [19]. Another study showed that the combination of both Reynolds and Eötvös number (i.e.…”

New Approaches for Theoretical Estimation of Mass Transfer Parameters in Both Gas-Liquid and Slurry Bubble Columns

“…For bubble sizes above 3×10 -3 m they have reported correction factors above 1.1. Talvy et al (2007b) have shown that the ellipsoidal shape of the bubbles seems to be significant in estimation of the drag coefficient. When these correlations are improved and become applicable at various temperatures, then the dimensionless temperature ratio in Eq.…”