Bioreactor scale-up: the key parameters from lab to pilot
A successful scale-up requires preserving three engineering parameters — volumetric oxygen transfer coefficient (kLa), power-to-volume ratio (P/V) and mixing time — across orders of magnitude in volume. Geometric similarity alone guarantees none of them.
Why geometric similarity fails
When you scale a stirred-tank bioreactor by multiplying dimensions by a factor k, volume grows by k³, surface area by k² and liquid depth by k. These mismatched exponents drive diverging values for every transport parameter.
This is the scale-up trilemma: you cannot simultaneously preserve P/V, kLa, tip speed and mixing time with a single criterion. The art is knowing which parameter your culture cannot afford to lose.
The three non-negotiable parameters
| Culture type | Target kLa | Typical P/V | Tip speed |
|---|---|---|---|
| Mammalian (CHO, HEK293) | 1–10 h⁻¹ | 10–100 mW/L | ≤ 1.2–1.5 m/s |
| Yeast | 50–200 h⁻¹ | 0.5–5 W/L | ≤ 3 m/s |
| Bacteria (E. coli) | 100–500 h⁻¹ | 1–10 W/L | ≤ 5 m/s |
- kLa: governs dissolved-oxygen availability. Always measure it in the real medium: antifoam reduces it by 20–50%.
- P/V: sets mixing intensity and macro-scale shear.
- Mixing time: in fed-batch processes, localised pH gradients can suppress productivity by 15–40% if θ exceeds 60 seconds.
Choosing the strategy
- Constant P/V: the most common for microbial fermentation.
- Constant kLa: correct when oxygen is the limiting substrate (high-density E. coli).
- Constant tip speed: the safest for fragile mammalian cultures.
In practice, most programmes use a hybrid approach validated by a bench-scale DOE before committing to pilot hardware.
Conclusion
A unified bioreactor and SCADA platform from bench to pilot removes control-system migration as a scale-up variable. Vinci Biotechnologies systems — Biobook, Biosip, Bioferm — share geometry, materials and Bioflex™ software from 0.5 L to 350 L+, reducing variables and accelerating process transfer.