Ask Greg McMillan - What are the opportunities for dynamic modeling of bioreactors?

Greg's Response:

Bioreactors (fermenters) are the key units of operation in biopharmaceutical, brewing, biochemical, and biofuel processes. Each bioreactor relies on the performance of billions of individual cells acting as bioreactors. Process control influences the sophisticated metabolic reactions inside the cell by controlling the environment immediately outside of the cell.

Doing modeling in the process development and early commercialization stages is advantageous because it increases process efficiency and provides ongoing opportunities for improving process control. When benchtop and pilot plant systems use the same industrial control systems and configuration expertise that are employed in manufacturing, applications of modeling and control can be developed as an integral part of the process definition and ported for industrial production via the control definition,

This synergistic discovery of knowledge is consistent with the intent behind the Process Analyzer and Process Control Tools sections of the Food and Drug Administration (FDA) Guidance for Industry: PAT-A Framework for Innovative Pharmaceutical Development, Manufacturing, and Quality Assurance.

Models offer benefits before they are put online. Significant improvements can result from the process knowledge and insight that are gained when building the experimental and first-principle models for process monitoring and control.

Because each bioreactor batch may be worth hundreds of thousands to millions of dollars (for high value-added products), opportunities for introducing perturbations are severely restricted. Because each batch takes days to weeks to complete, even if perturbations could be made within the specification limits set by the plant’s procedures for validating and managing change procedures, it might take months to years before enough plant data is available to develop an MPC, PLS, or ANN model.

There is a multiplicative effect for biological process kinetics that creates restrictions on experimental methods to analyze or predict cell growth or product formation. While the incentive is greater for high-value biologic products, there are challenges with models of biological processes due to multiplicative effects (neural networks and data analytic models assume additive effects).

First-principle dynamic models running at 200 times real-time in a digital twin can rapidly identify the golden batch and deal with the multiplicative effect plus the issues of not having sufficient data due to long batch times, concerns about creating a bad batch, and not knowing acceptable range of process parameters.

For much more knowledge see ISA Book New Directions in Bioprocess Modeling and Control – Maximizing Process Analytical Technology Benefits Second Edition.