CFD Modeling of a Bubble Column Reactor Carrying out a Consecutive A → B → C Reaction

In this paper, we develop a CFD model for describing a bubble column reactor for carrying out a consecutive first‐order reaction sequence A → B → C. Three reactor configurations, all operating in the homogeneous bubbly regime, were investigated: (I) column diameter D_T = 0.1 m, column height H_T = 1.1 m, (II) D_T = 0.1 m, H_T = 2 m, and (III) D_T = 1 m, H_T = 5 m. Eulerian simulations were carried out for superficial gas velocities U_G in the range of 0.005–0.04 m/s, assuming cylindrical axisymmetry. Additionally, for configurations I and III fully three‐dimensional transient simulations were carried out for checking the assumption of cylindrical axisymmetry. For the 0.1 m diameter column (configuration I), 2‐D axisymmetric and 3‐D transient simulations yield nearly the same results for gas holdup ?_G, centerline liquid velocity V_L(0), conversion of A, χ_A, and selectivity to B, S_B. In sharp contrast, for the 1 m diameter column (configuration III), there are significant differences in the CFD predictions of ?_G, V_L(0), χ_A, and S_B using 2‐D and 3‐D simulations; the 2‐D strategies tend to exaggerate V_L(0), and underpredict ?_G, χ_A, and S_B. The transient 3‐D simulation results appear to be more realistic. The CFD simulation results for χ_A and S_B are also compared with a simple analytic model, often employed in practice, in which the gas phase is assumed to be in plug flow and the liquid phase is well mixed. For the smaller diameter columns (configurations I and II) the CFD simulation results for χ_A are in excellent agreement with the analytic model, but for the larger diameter column the analytic model is somewhat optimistic. There are two reasons for this deviation. Firstly, the gas phase is not in perfect plug flow and secondly, the liquid phase is not perfectly mixed. The computational results obtained in this paper demonstrate the power of CFD for predicting the performance of bubble column reactors. Of particular use is the ability of CFD to describe scale effects.     

Comparison of Hydrodynamics and Mass Transfer in Airlift and Bubble Column Reactors Using CFD

Computational Fluid Dynamics (CFD) is used to compare the hydrodynamics and mass transfer of an internal airlift reactor with that of a bubble column reactor, operating with an air/water system in the homogeneous bubble flow regime. The liquid circulation velocities are significantly higher in the airlift configuration than in bubble columns, leading to significantly lower gas holdups. Within the riser of the airlift, the gas and liquid phases are virtually in plug flow, whereas in bubble columns the gas and liquid phases follow parabolic velocity distributions. When compared at the same superficial gas velocity, the volumetric mass transfer coefficient, k_La, for an airlift is significantly lower than that for a bubble column. However, when the results are compared at the same values of gas holdup, the values of k_La are practically identical.     

Simulations of Gas Distributors in the Design of Shallow Bubble Column Reactors

Computational fluid dynamics (CFD) was used to simulate the effect of sparger construction in gas holdup and liquid axial velocity in a shallow bubble column reactor for the air‐water system. Model parameters were evaluated in 2‐ and 3‐D simulations by using a two‐fluid model and the standard k‐? turbulence model. The Eulerian‐Eulerian approach was employed to predict the height of column that is affected by the sparger. It was found that increasing the number of orifices in the sparger increases the total gas holdup. Moreover, each orifice causes an increase in the circulation and mixing of liquid in the column. The results of the simulations follow the trends observed in the findings of Dhotre and Joshi [1].     

Large Bubble Sizes and Rise Velocities in a Bubble Column Slurry Reactor

The results are reported of an experimental study of the gas holdup, ?_G, large bubble diameter, d_Lb, and large bubble rise velocity, V_Lb, in a 0.1 m wide, 0.02 m deep and 0.95 m high rectangular slurry bubble column operated at ambient temperature and pressure conditions. The superficial gas velocity U was varied in the range of 0–0.2 m/s, spanning both the homogeneous and heterogeneous flow regimes. Air was used as the gas phase. The liquid phase used was C₉‐C₁₁ paraffin oil containing varying volume fractions (?_S = 0, 0.05, 0.10, 0.15, 0.20 and 0.25) of porous catalyst (alumina catalyst support, 10 % < 10 μm; 50 % < 16 μm; 90 % < 39 μm). With increasing slurry concentrations, ?_G is significantly reduced due to enhanced bubble coalescence and for high slurry concentrations the “small” bubbles are significantly reduced in number. By the use of video imaging techniques, it was shown that the large bubble diameter is practically independent of the gas velocity for ?_S > 0.05 and U > 0.1 m/s. The measured large bubble rise velocity V_Lb agrees with the predictions of a modified Davis‐Taylor relationship.     

Experimental Study on Bubble Rising and Descending Velocity Distribution in a Slurry Bubble Column Reactor

To determine bubble rising and descending velocity simultaneously, a BVW‐2 four‐channel conductivity probe bubble parameters apparatus and its analysis are used in gas‐liquid and gas‐liquid‐solid bubble columns. The column is 100 mm in internal diameter and 1500 mm in height. The solid particles used are glass beads with an average diameter of 17.82 μm, representing typical particle size for catalytic slurry reactors. The effects of superficial gas velocity (1.0 cm/s ≤ U_g ≤ 6.4 cm/s), solid holdup (0 % ≤ ?_s ≤ 30 %), and radial location (r/R = 0, 0.4, and 0.7) on bubble velocity distributions are determined. It is found that increasing U_g can increase the velocity of bubbles but do not exert much influence on bubble velocity distribution. Solid holdup mainly affects the distribution of bubble velocity while the radial direction affects bubble velocity distribution only slightly. The ratio of descending bubbles to rising bubbles increases from the bubble column center to the wall. It can be proved experimentally that large bubbles do not always rise faster than small bubbles at higher U_g (for example 6.4 cm/s).     

Mass Transfer Characteristics of Syngas Components in Slurry System at Industrial Conditions

The gas‐liquid mass transfer behavior of syngas components, H₂ and CO, has been studied in a three‐phase bubble column reactor at industrial conditions. The influences of the main operating conditions, such as temperature, pressure, superficial gas velocity and solid concentration, have been studied systematically. The volumetric liquid‐side mass transfer coefficient k_La is obtained by measuring the dissolution rate of H₂ and CO. The gas holdup and the bubble size distribution in the reactor are measured by an optical fiber technique, the specific gas‐liquid interfacial area aand the liquid‐side mass transfer coefficient k_L are calculated based on the experimental measurements. Empirical correlations are proposed to predict k_L and a values for H₂ and CO in liquid paraffin/solid particles slurry bubble column reactors.     

Gas‐Liquid Mass Transfer in a Slurry Bubble Column at High Slurry Concentrations and High Gas Velocities

The volumetric mass transfer coefficient k_La in a 0.1 m‐diameter bubble column was studied for an air‐slurry system. A C₉‐C₁₁ n‐paraffin oil was employed as the liquid phase with fine alumina catalyst carrier particles used as the solid phase. The n‐paraffin oil had properties similar to those of the liquid phase in a commercial Fischer‐Tropsch reactor under reaction conditions. The superficial gas velocity U_G was varied in the range of 0.01 to 0.8 m/s, spanning both the homogeneous and heterogeneous flow regimes. The slurry concentration ?_S ranged from 0 to 0.5. The experimental results obtained show that the gas hold‐up ?_G decreases with an increase in slurry concentration, with this decrease being most significant when ?_S < 0.2. k_La/?_G was found to be practically independent of the superficial gas velocity when U_G > 0.1 m/s is taking on values predominantly between 0.4 and 0.6 s^–1 when ?_S = 0.1 to 0.4, and 0.29 s^–1, when ?_S = 0.5. This study provides a practical means for estimating the volumetric mass transfer coefficient k_La in an industrial‐size bubble column slurry reactor, with a particular focus on the Fischer‐Tropsch process as well as high gas velocities and high slurry concentrations.     

Application of Chaos Analysis for the Investigation of Turbulence in Heterogeneous Bubble Columns

The Kolmogorov entropy (KE) algorithm was applied to local gas holdup fluctuations for estimating the degree of turbulence (chaos) in a large‐scale bubble column (0.289 m ID and clear liquid height: 1.31 m). The latter was operated with nitrogen and polyalphaolefin liquid (PSS8) in the churn‐turbulent regime (0.117 ≤ U_g ≤ 0.257 m/s) at both high temperature (T = 441–468 K) and pressure (P = 1.8–2.5 MPa). Measurements in four different zones (z = 0.35, 0.67, 1.1 and 1.52 m) were performed. When the bubble coalescence is weak (local gas holdups < transitional gas holdup), the KE exhibits an initial peak (enhanced turbulence) in the first zone, then a sudden drop in the second zone and finally it increases smoothly in the upper zones. In the case of strong bubble coalescence, the KE exhibits alternating maxima and minima, i.e., instable flow patterns. It was found that at z = 1.52 m, the KE values increase with P. At higher values of T, the KE values are lower in all zones.     

CFD Simulations of a Bubble Column Operating in the Homogeneous and Heterogeneous Flow Regimes

Bubble columns are operated either in the homogeneous or heterogeneous flow regime. In the homogeneous flow regime, the bubbles are nearly uniform in size and shape. In the heterogeneous flow regime, a distribution of bubble sizes exists. In this paper, a CFD model is developed to describe the hydrodynamics of bubble columns operating in either of the two flow regimes. The heterogeneous flow regime is assumed to consist of two bubble classes: “small” and “large” bubbles. For the air‐water system, appropriate drag relations are suggested for these two bubble classes. Interactions between both bubble populations and the liquid are taken into account in terms of momentum exchange, or drag‐, coefficients, which differ for the “small” and “large” bubbles. Direct interactions between the large and small bubble phases are ignored. The turbulence in the liquid phase is described using the k‐ϵ model. For a 0.1 m diameter column operating with the air‐water system, CFD simulations have been carried out for superficial gas velocities, U, in the range 0.006–0.08 m/s, spanning both regimes. These simulations reveal some of the characteristic features of homogeneous and heterogeneous flow regimes, and of regime transition.     

Multiple effects of operating variables on the bubble properties in three-phase slurry bubble columns

Flow properties of gas phase reactants such as size, rising velocity and frequency were investigated in simulated three-phase slurry bubble column reactors. Effects of gas velocity, reactor pressure, liquid viscosity, solid content in the slurry phase and column diameter on the flow properties of a gas reactant were determined. The multiple effects of operating variables on the bubble properties were well visualized by means of contour maps. The effects of operating variables on the flow properties of bubbles changed with changing column diameter of the reactor. The size, rising velocity and frequency of reactant gas bubbles were well correlated in terms of operating variables including column diameter of the reactor. This work was presented at the 7 ^th China-Korea Workshop on Clean Energy Technology held at Taiyuan, Shanxi, China, June 26–28, 2008.     

Gas Holdup and Volumetric Mass Transfer Coefficient in a Slurry Bubble Column

The gas holdup, ?, and volumetric mass transfer coefficient, k_La, were measured in a 0.051 m diameter glass column with ethanol as the liquid phase and cobalt catalyst as the solid phase in concentrations of 1.0 and 3.8 vol.‐%. The superficial gas velocity U was varied in the range from 0 to 0.11 m/s, spanning both the homogeneous and heterogeneous flow regimes. Experimental results show that increasing catalyst concentration decreases the gas holdup to a significant extent. The volumetric mass transfer coefficient, k_La, closely follows the trend in gas holdup. Above a superficial gas velocity of 0.04 m/s the value of k_La/? was found to be practically independent of slurry concentration and the gas velocity U; the value of this parameter is found to be about 0.45 s^–1. Our studies provide a simple method for the estimation of k_La in industrial‐size bubble column slurry reactors.     

Oxygen mass transfer coefficient in bubble column slurry reactor with ultrafine suspended particles and neural network prediction

The gas–liquid volumetric mass transfer coefficient was determined by the dynamic oxygen absorption technique using a polarographic dissolved oxygen probe and the gas–liquid interfacial area was measured using dual‐tip conductivity probes in a bubble column slurry reactor at ambient temperature and normal pressure. The solid particles used were ultrafine hollow glass microspheres with a mean diameter of 8.624 µm. The effects of various axial locations (height–diameter ratio = 1–12), superficial gas velocity (u_G = 0.011–0.085 m/s) and solid concentration (ε_S = 0–30 wt.%) on the gas–liquid volumetric mass transfer coefficient k_La_L and liquid‐side mass transfer coefficient k_L were discussed in detail in the range of operating variables investigated. Empirical correlations by dimensional analysis were obtained and feed‐forward back propagation neural network models were employed to predict the gas–liquid volumetric mass transfer coefficient and liquid‐side mass transfer coefficient for an air–water–hollow glass microspheres system in a commercial‐scale bubble column slurry reactor. © 2012 Canadian Society for Chemical Engineering     

CFD simulation of effects of the configuration of gas distributors on gas-liquid flow and mixing in a bubble column

Numerical simulations of gas-liquid flow in a cylindrical bubble column of 400 mm in diameter at the superficial gas velocity were conducted to investigate effects of the configuration of gas distributors on hydrodynamic behaviour, gas hold-up and mixing characteristics. Eight different gas distributors were adopted in the simulation. The simulation results clearly show that the configuration of gas distributor have an important impact on liquid velocity and local gas hold-up in the vicinity of the gas distributor. Comparisons of the overall gas holdup and mixing time among different gas distributors have demonstrated that none of the adopted gas distributors was able to produce the highest interfacial area and also yield the shortest mixing time. The CFD modelling results reveal that an increase in the number of gas sparging pipes used in gas distributors is beneficial in improving the gas hold-up but is disadvantageous in reducing bubble size due to a decrease in turbulent kinetic dissipation. It has been demonstrated from the simulations that the appearance of asymmetrical flow patterns in the bubble column and the adoption of smaller gas sparging pipes for gas distributors are effective in improving the mixing characteristics.     

CFD modeling of flow,macro-mixing and axial dispersion in a bubble column

The flow pattern in a bubble column depends upon the column diameter, height, sparger design, superficial gas velocity and the nature of gas–liquid system. In this paper, the effect of some of these parameters have been simulated using Computational Fluid Dynamics (CFD). The relationship of these parameters with the interphase force terms has been discussed. A complete energy balance has been established. Using this methodology, the flow patterns reported by Hills (1974), Menzel et al. (1990), Yao et al. (1991) and Yu and Kim (1991) have been simulated. Excellent agreement has been shown between the CFD predictions and the experimental observations. The above model has been extended to homogenization of an inert tracer. In order to confirm this model, mixing experiments were carried out in a 200 mm i.d. bubble column. A radioactive tracer technique was used for the measurement of mixing time. Tc-99m (^99m Tc), in the form of sodium pertechnate salt, was used as the liquid phase tracer. Good agreement has been shown between the predicted and the experimental values of mixing time. The model was further extended for the estimation of axial dispersion coefficient (D_L). Excellent agreement between the simulated and the experimental values of the axial dispersion coefficient confirms the predictive capability of the CFD simulations for the mixing process.     

Prediction of The Kolmogorov Entropy Derived from CT Data in a Bubble Column Operated under The Transition Regime and Ambient Pressure

The Kolmogorov entropy (KE) algorithm was successfully applied to single source γ‐ray Computed Tomography (CT) data measured by three scintillation detectors in a 0.162 m‐ID bubble column equipped with a perforated plate distributor (163 holes × ?? 1.32 · 10^–3 m). The aerated liquid height was set at 1.8 m. Dried air was used as a gas phase, while Therminol LT (ρ_L = 886 kg m^–3, μ_L = 0.88 · 10^–3 Pa s, σ = 17 · 10^–3 N m^–1) was used as a liquid phase. At ambient pressure, the superficial gas velocity, u_G, was increased stepwise with an increment of 0.01 m s^–1 up to 0.2 m s^–1. Based on the sudden changes in the KE values, the boundaries of the following five regimes were successfully identified: dispersed bubble regime (u_G < 0.02 m s^–1), first transition regime (0.02 ≤ u_G < 0.08 m s^–1), second transition regime (0.08 ≤ u_G < 0.1 m s^–1), coalesced bubble regime consisting of four regions (called 4‐region flow; 0.1 ≤ u_G < 0.12 m s^–1), and coalesced bubble regime consisting of three regions (called 3‐region flow; u_G > 0.12 m s^–1). The KE values derived from three scintillation detectors in the first transition regime were successfully correlated to both bubble frequency and bubble impact. The latter was found to be inversely proportional to the bubble Froude number. The KE model implies that the bubble size in this particular flow regime is a weak function of the orifice Reynolds number (d_b = 7.1 · 10^–3Re₀^–0.05).     

CFD‐Facilitated Flow Field Analysis of Bubble Columns with Concentric Plates for Biological Applications

As the hydrodynamic behavior in bubble columns is difficult to characterize, computational fluid dynamics (CFD) is a useful alternative tool for research and design. An experimental and computational analysis of the macromixing and gas holdup of a bubble column indicates that 3D simulations with CFD using an Eulerian‐Eulerian approximation yield results of the overall velocity field and gas holdup distribution that are suitable for engineering design purposes. Particularly, CFD simulations uncover that the inclusion of concentric solid plates into a bubbling column increase the gas holdup by 79 % and the mixing time by 48 % when compared with a column without plates operating at similar superficial gas velocities.     

Identification of Various Transition Velocities in a Bubble Column Based on Kolmogorov Entropy

The Kolmogorov entropy (KE) algorithm was applied successfully to gas holdup fluctuations measured in a 0.102 m I.D. stainless steel bubble column equipped with a perforated plate distributor (19 holes ?? 1 mm). Nitrogen was used as the gas, while both 1‐butanol and gasoline were used as liquids. 1‐Butanol was aerated at pressures, P = 0.1 and 0.5 MPa, whereas gasoline was aerated at P = 0.1 and 0.2 MPa. Based on the peaks in the KE values under the pressures examined in both liquids, the boundaries of the following five regimes were identified: bubbly flow, first transition, second transition, coalesced bubble (4‐region flow) and coalesced bubble (3‐region flow). As the pressure increases to P = 0.5 MPa in 1‐butanol, all four transition velocities shift to higher superficial gas velocity, u_G. In addition, in gasoline at P = 0.2 MPa and u_G ≤ 0.017 m s^–1, the existence of a chain bubbling regime was detected, whereas in 1‐butanol at P = 0.5 MPa and u_G ≤ 0.02 m s^–1, both laminar and turbulent chain bubbling subregimes were identified. It was found that in 1‐butanol under ambient pressure, the second and fourth transition velocities occur earlier than in gasoline.     

Influence of solids concentration and static mixers on fluid dynamics in three-phase fluidized bed bioreactors

The influence of solids concentration and static mixers on the hydrodynamics of the gas phase was studied for a three-phase fluidized bed bioreactor (air, nutrient solution, biocatalyst Ca alginate beads). Axial gas hold-up profiles, radial gas velocity profiles, mean bubble diameter and gas/liquid interfacial area per unit volume were measured in a bubble column (D_R = 0.142 m, H_R = 1.748 m). The influence of solids concentration on the gas hold-up is insignificant; static mixers enhance the gas hold-up in the reactor volume element in which they are installed. Axial gas velocity decreases with increasing solids concentration. At high solids concentrations, static mixers exert little influence on the gas phase but, at low concentrations, they do. A model is suggested to describe the influence of solids concentration (characterized by turbulent viscosity v_t) and static mixers (characterized by profile parameter n) on the gas velocity profile.