Mixing is a key process in the chemical, pharmaceutical and related industries. Inadequate understanding of mixing can result in unsatisfactory product quality and increased production costs.
Mixing operations often are complex and multi-faceted. They not only require an understanding of the fluid flow aspects, but also an understanding of the equipment's mechanical and power requirements. Mixing operations can involve single-phase liquid mixing, liquid,"liquid mixing, solid,"liquid mixing, gas,"liquid mixing, solid,"solid mixing and, in some cases, three-phase mixing involving solids, liquids and gases.
The primary function of a mixing vessel is to provide adequate stirring and mixing of the material. The mixing characteristics influence the product quality and efficiency of the process to a great degree.
For liquids, mixers ," or stirred vessels ," come in various shapes and sizes. The main vessel is cylindrical in shape, and the vessel bottom is often contoured. Baffles are included in the vessel to break the vortex and prevent solid-body rotation of the fluid. Draft tubes are included to direct suction and discharge streams. Dip tubes are employed to inject fluids at specific locations.
An important component of a stirred tank is the impeller. The rotating impeller imparts motion and shear to the fluid, inducing mixing. The type of impeller employed depends on the nature of the task. Often, the same stirred vessel is required to perform various duties. It is important to ensure efficient and optimum operation of the stirred vessel for a given duty. It also is necessary to create process conditions that are optimum at the lab scale, pilot scale and production scale so productivity is maximized.
A stirred vessel incorporates little or no instrumentation. The degree of mixing effectiveness is determined by product quality. To compensate for poor mixing, the mixing equipment often is over designed; however, such an over-design can be counter-productive. Excessive mixing can damage biological material and lead to high capital and operating costs.
The requirements for liquid,"liquid mixing are very different from those of liquid,"solid mixing or liquid,"gas mixing. Scale-up or scale-down of mixing processes is not easy. Scale-up of lab processes to pilot and production scale is difficult.
Scale-up problems are made worse when existing pilot and production mixers are used for new processes. Scale-up or scale-down often is carried out using trial-and-error methods based on prior experience and equipment vendor suggestions.
Time and effort spent on process scale-up/scale-down can be significant. Predictive tools to analyze existing equipment and techniques to scale a process from lab to pilot ," and ultimately to production ," are required.
This article describes analysis methods that can be used to design and analyze liquid,"liquid mixing equipment and to address scale-up/scale-down issues.
For the purposes of this article, a mixing process for liquid,"liquid blending first was designed at the pilot scale. This process then was scaled to full production scale. A solution strategy consisting of three tiers then was adopted.
The tier one analysis was based on general guidelines and dimensional analysis for mixing equipment. Sizing and specification of equipment were carried out using this approach. General guidelines for mixing vessels were adopted to estimate tank size, baffle size, impeller type and size.
For liquid,"liquid mixing, a number of impellers ," including turbine, pitched blade, marine-propeller type or Lightnin A310 ," can be used. The overall stirred vessel configuration parameters at the pilot scale are outlined in Table 1.
The tier two analysis involved the application of empirical data, along with solutions of mass and momentum, on a global scale. In this approach, fundamental equations of fluid dynamics were simplified based on experimental results and solved for rapid analysis of stirred vessels. These tools are valuable when identifying good and bad blending practices and estimating average mixing characteristics.
For the given stirred vessel configuration, tier two methods were applied to estimate important mixing parameters such as tangential velocity distribution, power, axial velocity distribution, mixing time, minimum impeller revolutions per minute (rpm) for solid suspension, overall dissipation rate and turbulence. These parameters were computed at the lab scale, pilot scale or production scale for process scale-up or scale-down. The impact of configuration changes on stirred vessel performance then could be rapidly estimated.
This technique was applied to assess impeller performance and identify proper placement location of impellers in the vessel. Table 2 depicts the performance of stirred vessel for pitched-blade, Rushton-turbine and A310 impellers. The tank and baffle dimensions, impeller sizes and placement locations are the same for all impeller types.
The pitched-blade impeller provided the highest circulation flow rate, but also consumed the most power. The Rushton-turbine impeller provided a low mixing time and a low mean period of circulation for moderately-low power consumption. The A310 impeller provides low flow circulation at low power consumption.
The Rushton-turbine impeller, therefore, was selected for further investigation. The impeller location in the vessel was varied to further minimize mixing time and the circulation period. Table 3 depicts the stirred-vessel performance for the Rushton-turbine impeller placement parameters.
The stirred-vessel geometry at the pilot scale was scaled up to attain geometric parameters at the full scale. This scale-up process, to a large extent, depends on the mixing process parameters. It is essential to maintain the same mixing time at the pilot and full scale.
Appropriate scaling of the pilot-scale process was carried out to ensure the mixing time at the two scales was unchanged. The full-scale stirred-vessel configuration performance then was examined using tier two methods. Table 4 depicts the performance parameters at full scale.
The final configuration at the pilot scale and full scale was analyzed in detail using tier three methods. These methods are based on the solution of Navier-Stokes equations to predict stirred vessel behavior.
Computational fluid dynamics (CFD) is one such method. This method is useful in evaluating detailed flow patterns in complex geometries and in situations in which tier two methods are not applicable. CFD methods involve the resolution of conservation equations of mass, momentum and energy at thousands of locations within the flow domain. A CFD solution provides full-field data; flow variables at every location in the domain are available; and a graphical representation of the flow can be created.
For the stirred-vessel study, CFD was applied to identify regions of low or excessively high flows and attain information on localized flow behavior. Fig. 1 depicts the CFD computed velocity distribution in the stirred tank at the pilot-scale, and Fig. 2 depicts the CFD computed velocity distribution in the vessel at the full scale. The nondimensionalized velocity fields at the two scales, as depicted in Fig. 3 and Fig. 4, are similar, indicating the proper scaling from pilot scale to full scale was achieved.
The multi-tiered solution strategy provides information at various scales. It also can be applied to solve other problems associated with stirred vessels. This information can be assimilated for selected processes to generate specific guidelines for design and scale-up of stirred vessels. Scale-up or scale-down study for any stirred tank mixing process can be carried out using these techniques.
Pordal is a staff consultant and Matice is principal and team leader at SES-Process Technology Group, Mason, Ohio. Contact them at (513) 336-6701.
1. Harnby, N., M. F. Edwards and A. W. Nienow. Mixing in the Process Industries, Butterworth Heinemann, 1992.
2. Fluent-5 User's Guide, 1998.
3. Perry, R. H. and D. Green. Chemical Engineers Handbook, 1984.
4. Tatterson, G. B., Fluid Mixing and Gas Dispersion in Agitated Tanks, McGraw-Hill Inc., 1991.
5. Tatterson, G. B., Scaleup and Design of Industrial Mixing Processes, McGraw-Hill Inc., 1994.
6. Visimix. 2000 User's Guide, 1999.
7. Visimix. Math Model Guide, 1999.