Reactive extraction columns - particle population balances and scale up

• Applicant: Professor Dr.-Ing. Hans-Jörg Bart
• Subject Area: Chemical and Thermal Process Engineering
• Term: from 2014 to 2018
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 257964779

Particle population balances form nowadays the cornerstone for modeling poly-dispersed systems arising in many engineering applications such as crystallization, granulation, emulsion polymerization, fluidized beds and microbial systems. The phenomenon of liquid flows with high particle densities is notorious complex, difficult to measure and predict and numerically more complex when chemical reactions are involved. Reactive extraction is in comparison to physical extraction the world-wide the dominant application (see Int. Solv. Extraction Conferences http://www.solventextract.org/index.html). However, mixer-settlers dominate the field. The economically more attractive use of columns is hindered by their more sophisticated design. In an early design phase (time to market) it is essential to have reliable models and simulation tools. Therefore, the consideration of the particulate nature via population balances is a central point. In that respect we want to validate our column scale-up and design methodology at a high generic level, so it should be approved on one certain type of a stirred (Kühni) resp. pulsed column (sieve tray), which are quite complementary in their internal flow structure. The working program shall validate our working hypothesis, that a scale up from bench and miniplant data is possible for any column type and even for complex reactive systems (e.g. with the chemically reactive system ZnSO4/D2EHPA/n-heptane, recommended by the European Federation of Chemical Engineering (EFCE) as reactive test system). As advantage, we have from previous works all equipment at hand (needs some adaptions) and have a highly automated data acquisition, which allows to undertake the proposed tight scheduled program.The flow behavior for columns smaller than DN200 is properly characterized with correlations from literature. However, with bigger columns the geometry ratios are no longer idem, the correlations no longer valid and scale up is thus uncertain. In that respect with the help of CFD methods (coupling of CFD & PPB) spatial resolved information should be derived as basis for a profound understanding (PC-based derived backmixing coefficients etc.) and resolution of local phenomena (dead zones, stirrer discharge flow) in any new compartment geometry.The outcome of this project is of utmost importance for the design and scale-up of industrial equipment involving dispersed reactive fluids in any geometry. The developed methodology serves as basis for gaining more insight and understanding in an empirically dominated field. Starting with standardized bench scale and miniplant experiments a computer aided scale up can reduce time and cost expensive experimentation and give results in domains otherwise difficult to explore.

• DFG Programme: Research Grants