(372j) Digital Integration of Fermentation and Centrifugation Processes: Optimizing Pilot-Plant Biomanufacturing for Enhanced Efficiency

Adopting Industry 4.0 principles in the biomanufacturing industry marks a significant transition in manufacturing approaches. Digital transformation is the foundation of this evolution, holding immense potential for optimizing processes, improving product quality through enhanced control, and reducing time-to-market. Digitalization is crucial in enhancing biomanufacturing operations by bridging the gap between physical and digital entities. This convergence enables real-time monitoring, data analysis, and decision-making, paving the way for smart manufacturing. Biomanufacturing operations can streamline workflows, optimize resource utilization, and mitigate operational risks while attaining a competitive advantage in the global marketplace (Gargalo et al., 2021).
A challenge in bio-based manufacturing is the generally complex and non-linear nature of microbial and metabolic systems, making it difficult and computationally expensive to predict fermentation progression. Fermentation processes are influenced by many factors, including nutrient availability, pH levels, temperature, and microbial interactions, making them inherently dynamic and challenging to predict. Consequently, optimizing fermentation processes and product purification requires a deep understanding of microbial physiology, process kinetics, physics, and thermodynamics, coupled with advanced modeling and simulation techniques (Bähner et al., 2021).
To address these challenges, researchers increasingly turn to pilot-plant bioprocesses as valuable testbeds for innovation and optimization. Recent advancements in pilot-plant bioprocessing have yielded valuable insights into the digitalization of biomanufacturing operations. For instance, Stevnsborg et al. (2023) conducted research on modeling and optimizing fermentation processes using pilot-scale data for green fluorescent protein - GFPuv - production with aerobic fed-batch fermentation of E. coli BL21(DE3). Using process data collected at the Golden LEAF Biomanufacturing Training and Education Center (BTEC), the authors developed predictive models to estimate fermentation parameters, such as induction time, substrate utilization, and product yield. These models are valuable tools for optimizing fermentation conditions, improving process robustness, and accelerating process development timelines.
A similar methodology was developed for digitalizing a disk-stack centrifuge utilized to recover cells from the fermentation broth produced in the pilot fermenter employed in the studies conducted by Stevnsborg et al. (2023). This involved building a model to predict separation efficiency by manipulating key variables such as feed flow rate, bowl rotation speed, and solid discharge time. The objective was to optimize operational parameters while minimizing cell lysis and reducing operator workload, achieved through mechanistic models and parameter estimation techniques.
The ongoing study seeks to integrate digitalization models developed for fermentation and centrifugation processes, aiming to unify pilot plant unit operations within a single, cohesive digital framework. This integrated model holds promise for predicting the performance of individual unit operations by adjusting variables in interconnected processes. By integrating predictive models for upstream and downstream operations, comprehensive process models can be developed to capture the interdependencies between different unit operations. By employing advanced statistical analysis such as uncertainty and sensitivity analysis, we aim to unravel the impact of propagating the uncertainties from upstream into downstream processing. Ultimately, this methodology could be extrapolated for broader application in industrial – scale biomanufacturing plants, offering enhanced efficiency and performance optimization across the board.

Acknowledgments

This work is supported by: (i) the Accelerated Innovation in Manufacturing Biologics (AIM‐Bio) project funded by the Novo Nordisk Foundation (NNF19SA0035474) and (ii) Novo Nordisk Foundation-funded Sustain4.0: Real-time sustainability analysis for Industry 4.0 (NNF0080136).

References

Bähner F. D., Prado-Rubio O. A., and Huusom J. K., 2021. Challenges in Optimization and Control of Biobased Process Systems: An Industrial-Academic Perspective, Industrial and Engineering Chemistry Research, Vol. 60, 14985-15003.

Gargalo C. L., Caño de las Heras S., Jones M. N., Udugama I., Mansouri S. S., Krühne U. & Gernaey K. V., 2021. Towards the Development of Digital Twins for the Bio-manufacturing Industry, Advances in Biochemical Engineering/Biotechnology, Vol. 137, 1-35.

Muldbak M., Gargalo C. L., Krühne U., Udugama I, Gernaey K. V., Digital Twin of a pilot-scale bio-production setup, 2022. Computer Aided Chemical Engineering Vol. 49,1417-1422.

Stevnsborg M., Selle K., Barton R., Prado-Rubio O. A., Gargalo C., Gernaey K. V., Gilleskie G., and Huusom J. K., 2023. Digital Twins in Pilot Scale Fermentation: NonLinear State Estimation for Improving Induction Timing, Computer Aided Chemical Engineering, Vol. 52, 2637-2642.