Modeling and Validating of Oxygen Transport in Wave Bioreactors: Optimized Experimental Mass Transfer method and novel Lattice-Boltzmann CFD Approach

Sina Piontek¹Jürgen Fitschen¹Christian Weiland²Michael Schlüter²Thomas Wucherpfennig^1*

• ¹Boehringer Ingelheim Pharmaceutical Research and Development Centre Biberach, Biberach, Germany
• ²Technische Universitat Hamburg, Hamburg, Germany

Wave bioreactors are commonly used in biopharmaceutical upstream processes as an intermediate stage between shake flasks and stirred tanks within the seed train. They offer a controlled environment for cell cultivation while minimizing shear stress (Malla et al., 2021). Accurate characterization of these systems is essential for optimizing cell culture performance, particularly as state of the art cell lines require higher volumetric mass transfer coefficients kLa. This study aims to determine the volumetric mass transfer coefficient through experiments and computational fluid dynamics (CFD) simulations. An improved experimental method for the measurements of the volumetric mass transfer is presented, with results correlated to key process parameters: rocking angle, rocking rate, and filling volume. In addition, CFD simulations were caried out using M-Star CFD by means of a Lattice-Boltzmann Method-based solver. The mass transfer was calculated using Higbie's penetration theory, incorporating the Kolmogorov scale to define contact time (Higbie, 1935; Maltby et al., 2016). The analysis also integrates concepts from Friedl (Friedl, 2013) and the surface renewal model (Kermani et al., 2011), introducing the surface normal velocity as an additional parameter in the mass transfer coefficient kL calculation. Analyzes were carried out for 10 and 50 L wave bioreactors, with one degree of freedom movement. Optimized process parameters were identified and validated in biological cultivations, resulting in increased dissolved oxygen levels in the medium. These findings contribute to improved characterization and control of wave bioreactors, enabling more accurate prediction of process parameter effects.