Deeper process understanding and large cost savings through CFD

In 2017 Delvigne et al. [1] reported in their paper that the use of computational fluid dynamics (CFD) can reduce the number of scale-up trials by 80% with cost savings per trial from $40k to $80k. V. Atiemo-Obeng and S. Kresta, E. Paul reported in their Book “Handbook of Industrial Mixing: Science and Practice” [2], that “for each individual process test, the cost savings from using CFD were estimated to be between $500k –$1 million.”

These are only two examples that report the benefit of using CFD for scale-up and tech transfer in stirred reactors. The present use case shows one specific application of CFD and how it can help gaining deeper insights of the process and deriving conclusions for an informed decision making in the context of scale-up and tech transfer.

Problem statement

Computational Fluid Dynamics (CFD) modeling of stirred reactors can be approached at different levels of model fidelity. This is particularly true when simulating multiphase flows like liquid-air systems that are typically found in Bioreactors. The fidelity ranges from low for single phase simulations, to medium for Euler-Euler multiphase with single droplet diameter assumption, to high for population balance modeling with many droplet size classes. Since computational cost increases with higher model fidelity, it is usual to account only for the effects needed to solve the problem in scope.

For the optimization of a Bioreactor, where the positioning of the sparger and the aeration rate plays an important role, population balance modeling is the most suitable method. It allows to predict the temporal and spatial distribution of the air bubbles, as well as the size distributions of the bubbles.

Simulation insights and recommendations

Observations of transient effects in animation:

  • Left: As expected, air rises due to strong buoyancy effects and contributes significantly to the flow characteristics.
  • Left: The lower region of the reactor is not supplied with fresh air (blue).
  • Left: Coalescence occurs above the stirrer.
  • Right: At the current stirring rate, the peak of the bubble size distribution settles at around 3 mm.

Recommended actions:

  • To improve the air distribution and the gas holdup a circular sparger should be used.
  • The sparger should be positioned at a lower location.
  • Measure kLa and compare to kLa extracted from CFD for optimization.

Simulation metrics

The simulation is performed with OpenFOAM v10.

  • Solver: multiphaseEulerFoam
  • Impeller rotation model: MRF
  • Population balance model: bubbles
  • Population balance method: method of classes
  • Size groups: 22
  • Mesh: high quality cartesian mesh using cf-mesh+
  • Mesh size: ~ 300k

References

  1. Frank Delvigne, Ralf Takors, Rob Mudde, Walter van Gulik and Henk Noorman, TERRA Research Center, Microbial Processes and Interactions (MiPI), University of Liege, Li ege, Belgium. Institute of Biochemical Engineering, University of Stuttgart, Stuttgart, Germany 2017
  2. V. Atiemo-Obeng, S. Kresta, E. Paul. Handbook of Industrial Mixing Science and Practice (Wiley, Hoboken, NJ, 2004)
  3. Seidel, S.; Eibl, D. Influence of Interfacial Force Models and Population Balance Models on the kLa Value in Stirred Bioreactors. Processes 2021, 9, 1185. https://doi.org/10.3390/pr9071185 CFD-Model was newly built up with a different mesh and OpenFOAM v10 instead of OpenFOAM v7.
  4. J. Thomas, et al. “A Mechanistic Approach for Predicting Mass Transfer in Bioreactors,” Chemical Engineering Science online, DOI: 10.1016/j.ces.2021.116538 (Feb. 2021).
  5. https://www.pharmtech.com/view/computational-fluid-dynamics-in-upstream-biopharma-manufacturing-processes