Beyond P/V: A Smarter Way to Scale Bioprocesses 

Scaling a bioprocess is not simply a matter of making the bioreactor bigger. As processes move from laboratory development towards industrial production, the engineering challenge becomes increasingly complex: How do we ensure that the cells continue to receive the conditions they need as the process grows? This question became particularly relevant during our work with Bluu Seafood, where we successfully scaled a cell cultivation process from 50 L to 200 L and ultimately to 1,000 L. 

A successful scale-up — and a new question 

The initial scale-up followed a conventional engineering strategy based on maintaining specific power input (P/V). The process showed robust performance across all three scales, demonstrating that a P/V-based approach can provide a practical route for transferring a cultivation process into pilot-scale operation. Reaching 1,000 L was an important milestone. But it also raised a question that becomes increasingly relevant as cell culture processes mature: 

What happens when we have more cells? 

A larger reactor does not necessarily solve the challenge. As cell density and biomass increase, oxygen demand increases as well. If we simply continue to maintain the same P/V, can we still provide enough oxygen to support the process? And if we cannot, how can we predict where the limits are before running another scale-up experiment? That question led us to look beyond conventional scale-up parameters. 

From P/V to oxygen transfer 

Instead of focusing solely on maintaining a particular agitation speed or specific power input, we focused on the parameter that directly describes the reactor's ability to supply oxygen to the culture: the volumetric oxygen transfer coefficient, kLa. We characterized our complete 50, 200 and 1,000 L bioreactor platform, correlating kLa with specific power input (P/V) and superficial gas velocity (Ug) using the empirical relationship: 

kLa = A(P/V)α × Ugβ 

By experimentally determining the reactor-specific parameters A, α and β for each scale, we transformed the experimental data into something much more useful than a one-time characterization. 

The result: a predictive scale-up tool 

We developed a predictive tool that allows us to estimate kLa at a given P/V and gas velocity across our 50, 200 and 1,000 L bioreactor systems. 

This means that instead of having to experimentally test every potential combination of operating parameters at every scale, we can use the established reactor-specific correlations to predict oxygen transfer performance before running the process. For process development, this can make a significant difference. 

When developing or transferring a process, we can use the tool to assess questions such as: 

• What kLa can we expect at a defined P/V? 

• What operating conditions are required to achieve a target oxygen transfer rate? 

• Can the oxygen demand of a higher-cell-density process be met at a given scale? 

• Where are the potential oxygen-transfer limitations when moving from 50 to 200 or 1,000 L? 

• What operating window should be investigated experimentally? 

This provides a quantitative basis for making scale-up decisions before committing time and resources to experimental runs. 

Why this matters for higher cell densities 

The importance of this approach grows as cultivation processes become more productive. Higher cell densities mean higher oxygen demand. At some point, simply maintaining a conventional scale-up parameter such as P/V may no longer be sufficient — or may require operating conditions that are impractical from an engineering or cell-stress perspective. A predictive understanding of kLa allows us to identify these constraints earlier. It also shifts the focus of scale-up from “How do we maintain the same P/V?” to “Can we provide the oxygen transfer capacity that the process requires — and at what operating conditions?” That is a much more process-relevant question. 

Making process development faster and more predictable 

The journey from 50 L to 1,000 L provided the practical foundation for this work. The next step was turning what we learned from those reactors into a tool that can support future development. The result is a predictive kLa framework for our 50, 200 and 1,000 L bioreactor systems, giving Cultivate at Scale the ability to estimate oxygen transfer performance across scales and operating conditions. For clients, this means the potential to shorten process-development timelines, reduce the number of experimental iterations and identify scale-up limitations earlier. For us, it represents another step towards our goal of making cell-culture scale-up more predictable, more data-driven and ultimately more efficient. The objective is not to replace experimental validation. Rather, it is to make experimentation more targeted and reduce time for process scale-up and development. Instead of testing a broad range of conditions simply to discover what works, we can use the model to narrow the operating space and focus experimental efforts where they provide the most value. 

The work was also supported by Jakob Bäum, a Biotechnology student at Hochschule Biberach, who contributed to the experimental work and reactor characterization underlying the predictive framework. 

Thank you for reading The Cultivate at Scale Chronicle!   

Until next time,   
The Cultivate at Scale Team