Advanced
Axial and Radial Gas Holdup in Bubble Column Reactor
Axial and Radial Gas Holdup in Bubble Column Reactor
Bulletin of the Korean Chemical Society. 2014. Jun, 35(6): 1703-1705
Copyright © 2014, Korea Chemical Society
  • Received : September 23, 2013
  • Accepted : February 16, 2014
  • Published : June 20, 2014
Download
PDF
e-PUB
PubReader
PPT
Export by style
Article
Author
Metrics
Cited by
TagCloud
About the Authors
Sameer M. Wagh
Mohashin E Alam Ansari
Pragati T. Kene
G.H. Raisoni College of Engineeting and Technology for Women, RTMNU Nagpur, India

Abstract
Bubble column reactors are considered the reactor of choice for numerous applications including oxidation, hydrogenation, waste water treatment, and Fischer-Tropsch (FT) synthesis. They are widely used in a variety of industrial applications for carrying out gas-liquid and gas-liquid-solid reactions. In this paper, the computational fluid dynamics (CFD) model is used for predicting the gas holdup and its distribution along radial and axial direction are presented. Gas holdup increases linearly with increase in gas velocity. Gas bubbles tends to concentrate more towards the center of the column and follows a wavy path.
Keywords
Introduction
Bubble column reactors are of considerable interest in industrial processes and various biochemical processes. They are used for processes involving oxidation, polymerization, waste water treatment, Fischer-Tropsch (FT) and environ-mental friendly biomass/coal/gas-to-liquid fuels production. For reactions occurring in gas-liquid/slurry reactors, the gas liquid mass transfer rate is important. The mass transfer depends largely on the gas holdup and the interfacial area, which are also inter dependent. Effective mixing as well as high gas holdup and interfacial area between the phases leads to improved heat and mass transfer characteristics.
Bubble columns are cylindrical vessels filled with water, wherein gas is sparged at bottom via a distributor, in the form of bubbles, into liquid or liquid-solid suspension (slurry bubble columns). The liquid inside column begins to expand as soon as gas is introduced. Bubble column offer advantages of good heat and mass transfer characteristics. Ease to construct and operate, absence of moving parts, low maintenance cost and relatively cheap to install are some of the factors that render the bubble columns an attractive choice as reactors for the described processes.
The potential of Computational Fluid Dynamics (CFD) for describing the hydrodynamics of bubble column has been established by several publications in the past (Baten and Krishna, 2005; Wang and Wang, 2007). What happens quantitatively, when fluids flow, how fluid flow, how are the velocity vectors, how the pressure contours are predicted by the CFD simulations. Thus CFD can successfully be used to study the hydrodynamics of bubble column reactor. In the present paper CFD simulations have been carried out to predict gas holdup and its distribution along the column in axial and radial direction.
Development of CFD Model
Geometry. FLUENT preprocessor GAMBIT is used as geometry and mesh generator. A two dimensional bubble column of length 1 m and diameter 0.1 m is used which gives a length to diameter (L/D) ratio of 10. Mesh size of 0.005 m × 0.001 m is used in L/D directions. Bottom face contain 0.002 m diameter inlet for air at the center. Velocity inlet and pressure outlet boundary condition is used. The computational grid details are shown in the Figure 1 .
Solver Settings. Commercial CFD software package FLUENT is used to solve the equations of continuity, turbulence and volume fraction. Pressure velocity coupling is obtained using SIMPLEC algorithm.
Simulations were carried out for 0.003 m bubble diameter with air water system, operating at superficial gas velocities ranging from 0.001 m/s to 1 m/s. Initial volume fraction of water was taken zero for all the simulations. Rake was defined from 0.5 m to 1 m at the interval of 0.01 m counting to 50 in numbers. Computational grid details are shown in Figure 1 . The time stepping strategy used in all simulations is 5000 time steps with 0.0001 time step size, next 2000 steps at 0.001 and remaining steps at 0.01 until steady state was achieved. The steady state was achieved in about 9000 time steps, but simulations were run until 60 seconds i.e. 12750 time steps. Steady state was indicated where all the variables remained constant. 2D simulations were carried out on Windows 7 platform, running on Intel Core (TM) I3 processor with 3.2 GHz CPU frequency. Each simulation was completed in about a day.
PPT Slide
Lager Image
Computational grid details.
Equations. The governing equations are given below:
1. Continuity equation for gas phase
PPT Slide
Lager Image
2. Continuity equation for liquid phase
PPT Slide
Lager Image
3. Volume fraction sum
PPT Slide
Lager Image
4. Conservation of momentum for gas phase
PPT Slide
Lager Image
5. Conservation of momentum for liquid phase
PPT Slide
Lager Image
6. k-€ Turbulence model
PPT Slide
Lager Image
PPT Slide
Lager Image
The Eulerian multiphase model (two phase) available in Fluent 6.3 was chosen to carry out computer simulation. Heat and mass transfer were not activated. The equation requires time averaged continuity and conservation equations for each phase.
Volume fraction of gas and liquid phase is calculated from Eqs. (1) and (2) respectively, with the condition that the volume fraction of phases sums to one. Eqs. (4) and (5) describe momentum balance for gas and liquid phases respectively. The k-€ turbulence model is used and solves the standard k-€ equations. The standard k-€ model is a semi empirical model based on model transport equations for turbulent kinetic energy (k) and the dissipation rate (€). The mixture density and velocity, ρm and ⊽ m and are computed from:
PPT Slide
Lager Image
PPT Slide
Lager Image
Turbulent viscosity μ t,m in Eqs. (6) and (7) is written as:
PPT Slide
Lager Image
It is obtained from the prediction of the transport equations for the k and € given in Eqs. (6) and (7). Gk is the rate of production of turbulent kinetic energy. The model constants are Cμ, C 1 €, C 2 €, σk and σ€ and their respective values are: 0.09, 1.44, 1.92, 1.0 and 1.3.
Results and Discussion
Overall Gas Holdup. The cross sectional area averaged and volume averaged gas holdup values is shown in Figure 2(a) and (b) for low uperficial gas velocity (0-0.01 m/s) and for higher superficial gas velocity (0-1 m/s) respectively. The gas holdup increases almost linearly with the superficial gas velocity. Similar results have been reported by Baten and Krishna (2005), Cachaza et al . (2009). As the gas velocity increases, the amount of gas introduced per unit time increases and so also the gas holdup increases.
Radial Distribution of Gas Holdup. The radial distribu-tion for time averaged local gas holdup for varying super-ficial gas velocity at a height of 0.5 m from the gas distributer is shown in Figure 3 . With increasing gas velocities, gas holdup increases. This is because gas bubble tend to concen-trate more and more in the central core of the column (Baten and Krishna, 2005). This concentration of bubbles in the central core causes a substantial increase in the liquid circulation.
PPT Slide
Lager Image
(a) Cross sectional area averaged and column volume average gas holdup as a function of low superficial gas velocity. (b) Cross sectional area averaged and column volume average gas holdup as a function of high superficial gas velocity.
Gas Holdup Distribution along Column Height. Cross sectional area averaged values of gas holdup at different heights in the column is shown in Figure 4 and Figure 5 . For low gas velocities (Ug = 0.001 m/s to 0.01 m/s), ( Figure 4 ) gas holdup remains constant with increasing height. For higher gas velocities (Ug = 0.1 m/s to 1 m/s), ( Figure 5 ), upto 0.8 m (initial part of graph) height the gas holdup remains constant but with further increase in height gas holdup starts increasing with increase in gas velocity.
PPT Slide
Lager Image
Radial distribution of gas holdup. Values measured at a height of 0.5 m from the gas distributor.
PPT Slide
Lager Image
Gas Holdup Distribution along Column Height, values measured from a height of 0.5 m from gas distributer (For lower gas velocities).
PPT Slide
Lager Image
Gas Holdup Distribution along Column Height, values measured from a height of 0.5 m from gas distributer (For higher gas velocities).
PPT Slide
Lager Image
Radial distribution of gas holdup at varying heights.
Predicting Wavy Nature. Figure 6 shows the radial distribution of gas holdup at different heights. It is useful in predicting the wavy flow nature of gas. The peaks of the each curve are not situated along a straight line but these peaks of each curve are displaced in radial direction. This indicates the movement of the central core of gas. So it is clear that the gas follows a certain wavy flow pattern. Similar results are also reported by Joshi et al . (2002).
Conclusions
A CFD model is developed to describe the hydrodynamics of air-water bubble column reactor. The studies for overall gas holdup and its distribution along the radial and axial direction are discussed. Cross sectional area averaged gas holdup increases with increase in gas velocity. From the radial distribution of gas holdup it is observed that the gas bubble tend to concentrate more and more in the central core of the column. For low gas velocities the gas holdup remains constant with increase in height. With increase in height the radial distribution of gas holdup follows a shift in peaks which indicates the movement of central core of the gas along a wavy path.
Acknowledgements
Publication cost of this paper was supported by the Korean Chemical Society.
References
van Baten J. M. , Krishna R. 2005 Chem. Eng. Sci. 60 6857 - 6867    DOI : 10.1016/j.ces.2005.05.044
Cachaza E. M. , Díaz M. E. , Montes F. J. , Galán M. A. 2009 Ind. Eng. Chem Res. 48 8685 - 8696    DOI : 10.1021/ie900314s
Joshi J. B. , Vitankar V. S. , Kulkarni A. A. , Dhotre M. T. , Ekambara K. 2002 Chem. Eng. Sci. 57 3157 - 3183    DOI : 10.1016/S0009-2509(02)00192-6
2006 Fluent 6.3 User’s Guide, Lebanon NH, 23.5 Euler Model Theory Fluent Inc
Rampure M. R. , Mahajani S. M. , Ranade V. V. 2009 Ind. Eng. Chem. Res. 48 8186 - 8192    DOI : 10.1021/ie8018593
Szafran R. G. , Kmiec A. 2004 International Journal of Chemical Reactor Engineering 2 A30 -
Wang T. , Wang J. 2007 Chem. Eng. Sci. 62 7107 - 7118    DOI : 10.1016/j.ces.2007.08.033