A CFD-Based Hybrid Modeling Approach For pCO2 Prediction During Bioreactor Scale-up

By SeungsunLee, DowonKang, JunseongLee, HeekookYang, Sihyeon Park, SeungeunOh, JiinYang, HeeyoonYang, YunjungChoi, MSAT Team, LOTTE Biologics Co., Ltd., Incheon 22032, Republic of Korea

Scaling up mammalian cell culture bioreactor operations presents significant hurdles, particularly because altering vessel dimensions disrupts fluid dynamics and mass-transfer characteristics. Standard matching metrics, such as power-to-volume ratios, fail to capture critical spatial variations or guarantee identical cellular environments. Predicting dissolved carbon dioxide levels remains especially complex, often stalling the development of reliable scale-down models.

A computational fluid dynamics (CFD) approach combined with a thermodynamic kinetic model solves this challenge. By integrating physical transport phenomena with dynamic cellular metabolic data—specifically balancing carbon dioxide evolution against gas stripping efficiency—bioprocess engineers can directly model target gas dynamics. This mechanistic framework bypasses empirical guesswork, predicting large-scale dissolved carbon dioxide levels within 9% accuracy using only standard process data.

Review the poster to optimize bioreactor scaling strategies and refine gas control parameters.