Engineering & Technology · Biomedical, Bioprocess & Biological Engineering

Bioreactors, biological production and process scaling

A small biological production system performs consistently, but a larger vessel gives a different result. Multiplying the recipe quantities preserves some ratios while changing transport distances, mixing and heat-removal demands. Scale is an engineering change that needs evidence, not just more volume.

Use Process Balances & Design, Material Structure & Properties and Instrumentation & Embedded Integration for balances, mechanical models and measurement chains. Biology’s chemistry of life and Chemistry’s enzymes and metabolism provide further preparation. These lessons teach engineering principles, not cultivation protocols, patient-device settings or permission for human research.

Back to the chapter

After reading, you can

Connect biological production to material and transport balances, and explain why a larger vessel does not reproduce every small-scale condition.

The vessel is one part of production

A bioreactor provides an engineered environment in which biological agents or components perform a process. Production can involve cells, enzymes or other biological systems. Inputs, products, biomass, by-products and waste need a common accounting boundary. A larger amount of biomass does not by itself establish a larger amount of the desired product or the same product quality.

Batch, fed-batch and continuous operation differ in when principal feeds and harvest streams cross the selected boundary. A fed-batch process adds a feed during production; a continuous process has ongoing feed and withdrawal. Actual production also requires suitable containment, contamination control, monitoring, downstream recovery and evidence that the recovered material has its intended identity and function. Here we study balances rather than cultivation instructions.

A constructed oxygen-transfer comparison

For a hypothetical well-mixed liquid, let C be dissolved-oxygen concentration and C* a supplied interface-equilibrium value. A simple volumetric transfer model is kLa(C*−C). The combined coefficient kLa has units of inverse time: it packages a liquid-side transfer coefficient and gas–liquid interfacial area per liquid volume. It is not simply the outer wall area divided by vessel volume.

Supply C*=0.25 mol/m³, C=0.10 mol/m³ and an invented kLa=0.04 per second. Transfer is 0.006 mol/(m³·s). If the hypothetical biological demand is 0.004 in the same units, the instantaneous difference is +0.002. Halving kLa gives transfer 0.003 and difference −0.001 at that same C. In a dynamic liquid balance, concentration would then start falling; it would not remain fixed without another contribution.

Solid transfer curve: kLa 0.04 per second. Long-dashed transfer curve: kLa 0.02 per second. Horizontal short-dashed line: fixed demand. Transfer kLa(0.25−C) uses kLa 0.04 and 0.02 per second. At C 0.10, transfer 0.006 and 0.003 mol/(m³·s). Fixed demand 0.004 meets the curves at C 0.15 and 0.05 respectively. Common x-axis 0–0.25 mol/m³; rate axis 0–0.010. No biological demand dependence or spatial gradients included.

Geometry changes different quantities differently

For geometrically similar vessels, multiplying every length by four multiplies a corresponding wall area by 16 and volume by 64. Wall area per volume becomes one quarter of its previous value. If a model heat-generation rate per volume stayed constant and all heat left through that wall, the required heat removal per wall area would rise fourfold.

That wall argument is distinct from gas-bubble transfer area. Mixing patterns, gas dispersion, fluid properties and biological sensitivity can also change with scale. Holding one selected ratio constant does not hold all of them constant. State the similarity target, the quantities allowed to change and how relevant gradients and product outcomes will be checked.

Find the model’s steady concentration

With fixed demand 0.004 mol/(m³·s), the simplified no-flow steady balance requires kLa(C*−C)=0.004. For kLa=0.04 per second, C=0.25−0.004/0.04=0.15 mol/m³. With kLa=0.02, C=0.05 mol/m³. The two intersections in the figure are conditional mathematical results, not recommended oxygen targets.

Real demand can depend on concentration, biomass, time and other conditions, and transfer coefficients require evidence for the actual system. The model excludes spatial gradients and any liquid-feed terms. It identifies a supply–demand question to investigate, while downstream recovery and biological product quality remain separate parts of the production system.

Try the reasoning

Why is kLa measured in inverse time?
Show the answer and reasoning
A transfer coefficient in length/time multiplied by interface area per volume in inverse length gives inverse time.
Does a fourfold length increase give fourfold volume?
Show the answer and reasoning
No. Geometric similarity gives 4³=64 times the volume and 4²=16 times a corresponding area.
Why can a transfer calculation not establish product quality?
Show the answer and reasoning
It describes one transport balance; identity, biological function, contamination and recovery require their own relevant evidence.

What to carry forward

Scale biological production through defined balances and competing transport effects, then check the resulting biological and product performance.

References

Teaching resources and research supporting this lesson’s explanations and examples.