A novel model for the simulation of cell respiration in a mammalian cell culture process in an aerated stirred bioreactor

The cell respiration process of a Chinese Hamster Ovary (CHO) cell culture in a large-scale aerated and stirred bioreactor was modelled and simulated by using M-Star’s LES software. The simulation set-up was first validated with no cells present. The volume-averaged volumetric mass transfer coefficient ( k L a ) and dissolved O 2 (DO, %) plots are in good agreement with experimental data. Then, the cells were added to the simulation hypothetically by defining a reaction that couples the O 2 consumption rate to the CO 2 production rate. A controller was set on the pure O 2 sparger to control the DO in the tank. Some simplifying assumptions were made such that only the CO 2 variations as a result of the cell respiration were modelled. The simulations were run on GPUs for almost one month to find how the dissolved O 2 and CO 2 concentrations evolve with time. The average rates of O 2 transfer and consumption as well as of CO 2 production and absorption look qualitatively correct and are in good agreement with the limited validation data available. The effect of varying the Viable Cell Density (VCD) was investigated as well. Finally, the local dissolved CO 2 concentration data were used to calculate local and averaged partial pressures of CO 2 (pCO 2 ) which remained well within the suggested range of 20-80 mmHg reported by (Mostafa and Gu, 2003; Xing et al., 2017). Overall, the simulation approach showed great potential in capturing the local heterogeneities in dissolved O 2 and CO 2 concentrations. • A novel cell respiration model focuses on dissolved O 2 & CO 2 and on transfer rates • Our CFD results for K L a and dissolved O 2 agree well with experimental data • Spatial inhomogeneities in dissolved gases result in varying bubble compositions • Supplying additional pure O 2 has a pronounced impact on DO and pCO 2 • Upon adding pure O 2 , CO 2 removal lags production and raises dissolved CO 2 levels.