Computational fluid particle dynamics modeling of tangential flow filtration in perfusion cell culture

Membrane fouling is a common and complex challenge with cell culture perfusion process in biopharmaceutical manufacturing that can have detrimental effects on the process performance. In this study, we evaluated a method to calculate the hollow fiber membrane resistance at different time points for water and supernatant. In addition, the number of subvisible particles of < 200 nm. diameter suspended in the supernatant were quantified using a nano-flow cytometry method. A computational fluid dynamics (CFD) model was developed to evaluate the impact of feed flow rate and particle count on the transmembrane pressure (TMP). Then a steady-state discrete phase model was applied to incorporate particles into the model and simulate the particles deposition over the membrane wall. The results showed an increase in the number of particles and the membrane resistance along the time course of the perfusion process. The CFD model illustrated that more particle deposition was observed at lower feed stream flow rates. The fraction of deposited particle was reduced by > 50% when the feed flow rate was increased from 35 ml/min to 300 ml/min. Our findings suggest that the total number of subvisible particles has a significant impact on TMP and membrane resistance and, thus, could play a major role in the mechanism of membrane fouling. CFD modeling can be a useful tool to predict the behavior of a process in a specific membrane. CFD simulations could also be used to optimize process parameters to improve membrane cleanability, reduce particle deposition, and reduce the risk of membrane fouling.