(528d) Efficient Modeling of Critical Process Parameters in Bioreactors – a Case-Study across Scales

Critical process parameters during operation of industrial biofermenters such as volumetric mass transfer coefficient, gas-holdup, dissolved oxygen or sufficient supply of nutrients, are highly dependent on the fermenter geometry and scale and cannot be easily predicted by empirical correlations alone, which is apparent by the vast number of existing correlations and variations of up to orders of magnitude [1]. Ensuring optimal environmental conditions for microorganisms in large-scale production fermenters in the biopharmaceutical industry is of vital importance for the safety, stability, and optimized productivity of the fermentation process.

In this work, we present an in-silico approach to predict these critical parameters based on first principles using the commercial software SIMVANTAGE (simvantage.com), that utilizes the Lattice Boltzmann Method (LBM) for transient simulations of the liquid phase. It is based on an in-silico scale-independent simulation method to predict key parameters based on first principles, even suitable for large-scale fermenters. The implementation is optimized for state-of-the-art Graphics Processing Units (GPUs) and includes the motion, breakup, and coalescence of the dispersed gas phase via the Euler-Lagrange method, transport phenomena between phases and throughout the fermentation broth, as well as the movement and distribution of microorganisms. The inclusion of the microorganism movement allows for a detailed analysis of the inhomogeneities within the fermentation broth by tracking the individual organism lifelines and the environmental conditions of the whole population over time [2].

The first part of the presentation focusses on the characterization of a lab-scale reactor to show the applicability of the solver for capturing a multitude of operational regimes from the vortex Cavity regime to the loaded and flooded regime. In the second stage, results are compared to geometrically similar large-scale fermenters to assess the scale-independency of the simulation framework without re-tuning of modeling parameters.

Then, a use-case is shown on how to transfer process conditions from the lab scale to the production scale based on the similarity of the environmental conditions of the organism population. Possible applications include reducing experimental load during scale-up, as well as the construction of representative scale-down models of large-scale fermenters.

[1] Yawalkar et al. “Gas-Liquid Mass Transfer Coefficient in Stirred Tank Reactos”, Canadian Journal of Chemical Engineering, Volume 80, 2002, 840-848

[2] Haringa et al. “Euler-Lagrange computational fluid dynamics for (bio)reactor scale down: An analysis of organism lifelines” Engineering in Life Sciences, Volume 16 (7), 2016, 652-663