(313d) A Comprehensive Hydrodynamic Assessment of an Agitated Fermenter

Fluid mixing is a fundamental operation in wide range of the industries as biochemical and chemical industries. The mixing is fundamental for some process such as fermentation, wastewater, crystallization and polymerization. Stirred tanks are used for the homogenization of single or several phases. The gas-liquid system has received considerable interest in the last decades. Both the mechanical agitation and sparging play an important role in the design of this kind of systems.

To produce the required degree of uniformity inside the bioreactor the sparging and the agitation they should be the best possible. In the literature, many types of impellers have been reported for different applications. These studies demonstrate that the traditional Rushton turbine impeller has some advantages over other impellers for the gas-liquid mixing operations, although it has also shown some drawbacks such as the formation of dead regions [1]. This indicates that radial impellers provide poor pumping characteristics, which can be overcame using axial-flow for providing downward flow. However, these impellers showed other disadvantages such as torque instabilities and more power demand [2, 3].

In this study we analyzed the hydrodynamic performance of a new impeller called axial-radial impeller (ARI) in a stirred tank containing a Newtonian fluid, which was designed in our research group [4]. We do a comprehensive hydrodynamic assessment through of global and local parameters. We used the non-intrusive of PIV technique for local parameters (flow patterns, turbulent intensity, vorticity) and a torquemeter together with an electrode oxygen dissolved for global parameters (power consumption and volumetric mass transfer coefficient).

The hydrodynamic performance of the ARI impeller was compared to the performance of the radial-flow Rushton impeller and the axial-flow pitched blade turbine.

We used an experimental setup, which consists basically of a glass vessel of diameter T of 190 mm and liquid height H of 305 mm and an operating volume of 0.008 m³. The vessel is a cylinder with torispherical bottom and we put four baffles having a width J of 15.8 mm equally spaced with the aim of minimizing vortexing in the free-surface. Also, we put a ring sparger to supply air to the vessel at a distance of 59 mm measured from the bottom of the tank. We used sets of three impellers with a diameter Da of 63 mm coaxially placed along the shaft, namely: Rushton turbines (RT), 6-pitched blade turbines (PBT) and axial-radial impellers (ARI). They were driven with a 373 W (0.5 hp) DC motor (Baldor Reliance CDP330), whose speed was carefully controlled with a KBMD Baldor control and monitored with a tachometer (Delta CTA4). We mounted the first impeller on the shaft at a distance of 31.6 mm (C₁) measured from the sparger and the second impeller at a distance of 63.3 mm (C₂) measured from the first impeller and the third impeller at a distance of 63.3 mm (C₃) measured from the second impeller. With the aim of minimizing air intake to the working liquid, the distance between the third impeller and the free-surface we placed it to 92 mm (C₄). For a complete gas dispersion and with the aim of avoiding flooding as a function of volumetric gas flow rate from 0.125 to 2 vvm, the minimum rotation speed was set to 1.326 m/s, equivalent to N = 400 rpm.

Inside of the vessel, we poured the working fluid, which was made up with sodium alginate dissolved in distilled water at a concentration of 1 g/L, whose liquid density and dynamic viscosity was ρ= 997 kg‧m^-3 and µ= 6 mPa‧s, respectively.

Figure 1a) and Fig. 1b) (first row) shows the results of the power drawn by the impellers and the mass transfer, which were determined using the methodology described elsewhere [5, 6]. The power number (N_pg) vs flow rate number (Fl_g) are used to present the results of the power consumption and volumetric mass transfer coefficient (K_La) vs volume of air per volume of liquid (vvm) are used to present the results of the mass transfer.

As expected, the power number decreases as the flow number increases; however, for Fl_g higher than 0.05, the power number remains practically constant, which indicates that the turbulent regime has been reached. In the case of the Rushton turbines, the power drops quickly as the flow number increases; however, once the turbulent regime has been reached, the Rushton turbines require less power compared with the two other impellers. As reported elsewhere, this is due to the large cavities formed behind the blades and the higher frequency of bubble breakage [7-9]. It is noticeable that hybrid impellers (ARI) demand higher power. In such impellers, straight blades are attached to pitched blades like winglets. As these impellers rotate, small cavities are formed behind the straight blades and more energy would be required for bubble breakage, which could explain the higher power demanded. On the other hand, the figure 1 (b) shows the volumetric mass transfer coefficient for the impellers investigated as a function of the aeration rate in a vvm range from 0.25 to 2 min^-1. As expected, K_La increases with increasing the aeration rate, although the mass transfer values are highly dependent on the impeller used. As this figure shows, Rushton turbines exhibit the highest mass transfer capacity, while the lowest one is observed with the PBT at low vvm. However, this trend changes for vvm values higher than 0.5 min^-1, so that ARI impellers show the lowest mass transfer capacity.

Figures 1c), 1d), 1e), 1f), 1g), 1h) show the flow patterns obtained in the X-Y plane from the PIV measurements for the Rushton turbine (Fig. 1c) and Fig. 1f), the pitched blade turbine (Fig. 1d) and Fig. 1g), and the axial radial impeller (Fig. 1e) and Fig. 1h) for single-phase (second row) and gassed phase (third row) conditions. For all cases, the Reynolds number is Re=33502 and for gassed conditions the gas flow rate is 0.25 vvm. The color scale represents the velocity field magnitude normalized by V_tip and the impellers contour is illustrated with continuous lines and the horizontal and vertical dimensions (X and Y, respectively) are normalized using the radius of the tank (T/2).

As can be seen, the Rushton turbine exhibits more homogenous flow patterns in the whole tank and with more regions where the flow is faster for both conditions (single-phase and gassed phase). Besides, the flow pattern in the gas phase changed considerably probably due to the strong interaction between the turbine and gas bubbles dispersed in the flow, which are concentrated between the second and third impellers. For the pitched blade turbine and the axial radial impeller, the flow patterns do not present changes significant for both conditions (single-phase and gassed phase) due to the bubbles that do not interact with the impeller blades as for the case Rushton turbine. Nevertheless, we can see that the axial-radial impeller (ARI) creates a more complex flow pattern compared with the one created by the pitched blade turbine, in which more high-velocity regions are observed close to the shaft.

In this study, the hydrodynamic performance of a hybrid impeller was analyzed and compared to a Rushton turbine and a pitched blade turbine. The most relevant conclusions are:

Rushton turbines demand lower power and provide higher mass transfer capacity. Also, more homogeneous flow patterns were observed with these impellers as well as more high-velocity regions. On the other hand, PBT and ARI turbines showed similar hydrodynamics characteristic.

Figure 1- Global and local parameters for RT, PBT and ARI: a) Power consumption; b) Volumetric mass transfer coefficient and typical flow patterns for a X-Y plane for: RT, c) single-phase, f) 0.25 vvm; PBT, d) single-phase, g) 0.25vvm; ARI, e) single-phase, h) 0.25 vvm

References:

1. Nienow, A. W. "Gas-Liquid Mixing Studies-A Comparison of Rushton Turbines with Some Modern Impellers." Chemical engineering research & design4 (1996): 417-423.
2. McFarlane, Caroline M., and Alvin W. Nienow. "Studies of high solidity ratio hydrofoil impellers for aerated bioreactors. 1. Review." Biotechnology progress6 (1995): 601-607.
3. Nienow, A. W. "Hydrodynamics of stirred bioreactors." (1998): 3-32.
4. Ascanio, Gabriel, Stephane Foucault, and Philippe A. Tanguy. "Performance of a new mixed down pumping impeller." Chemical Engineering & Technology: Industrial Chemistry‐Plant Equipment‐Process Engineering‐Biotechnology 26.8 (2003): 908-911.
5. Ascanio, G., B. Castro, and E. Galindo. "Measurement of power consumption in stirred vessels—a review." Chemical Engineering Research and Design9 (2004): 1282-1290.
6. Garcia-Ochoa, Felix, and Emilio Gomez. "Bioreactor scale-up and oxygen transfer rate in microbial processes: an overview." Biotechnology advances2 (2009): 153-176.
7. Ozcan-Taskin, N. G., A. W. Nienow, and S. Chatwin. "Les characteristiques de puissance des lntermigs et des turbines Rushton dans les fluides viscoelastiques a viscosite constante." Agitation Industrielle, J(1992): 11-10.
8. Kipke, K. D. "Fluid loadings on impellers and tanks in gassed liquids." Fourth European Conference on Mixing. 1982.
9. Derksen, Jos, and Harry EA Van den Akker. "Large eddy simulations on the flow driven by a Rushton turbine." AIChE Journal2 (1999): 209-22