Effective scale-up of biologic manufacturing processes requires balancing fluid dynamics, reaction kinetics, and process control strategies to ensure consistent product quality. While computational fluid dynamics (CFD) has been widely used to characterize transport phenomena at scale, traditional approaches often fail to incorporate the dynamic behavior of in-line proportional-integral-derivative (PID) controllers. Here, we present a physics-based digital twin framework that couples three-dimensional, time-accurate CFD simulations with embedded PID control logic to directly predict process performance across operating scales.
Our simulations, based on the CFD-PID modeling approach developed by Oliveira et al. (2023) and further extended in Dasgupta et al. (2024) , reveal that while single-probe PID controllers can effectively regulate bulk process variables, they are often insufficient for managing spatial heterogeneity in large-scale bioreactors. In particular, dissolved oxygen gradients—difficult to quantify in traditional experimental setups—emerge as critical bottlenecks, leading to local deviations in oxygen availability that impact cell growth and product yield. By integrating multi-probe feedback strategies within the CFD-PID framework, we demonstrate improved control performance, reducing dissolved oxygen variability and enhancing scale-up fidelity. Furthermore, optimization of controller tuning parameters alongside hydrodynamic and kinetic constraints enables a more predictive approach to process design.
This work highlights the importance of moving beyond single-point control assumptions in bioprocess scale-up. By leveraging CFD-PID models, we provide a systematic methodology for optimizing both transport and control strategies, bridging the gap between small-scale development and large-scale manufacturing .
Works Cited
- Oliveira, C. L., Pace, Z., Thomas, J. A., DeVincentis, B., Sirasitthichoke, C., Egan, S., & Lee, J. (2023). CFD‐based bioreactor model with proportional–integral–derivative controller functionality for dissolved oxygen and pH. Biotechnology & Bioengineering, 121(2), 655–669. DOI:10.1002/bit.28598.
- Dasgupta, A., Thomas, J., Anand, A., DeVincentis, B., McCahill, M., Sood, A., Kinross, J., & Rajendran, A. (2024). Integrative in-silico Models for Mammalian Cell Cultures in Single-Use Bioreactors: Bridging Hydrodynamics, Kinetics, and Process Control Across Scales. Pfizer Manufacturing Sciences and Technology (MSAT), Pfizer Bioprocess Research and Development (BRD), M-Star CFD. In Review at Journal of Biotechnology
