NovaSolver›Biopharma Stirred Bioreactor Mixing Time Simulator Back
Biopharma Reactor
Biopharma Stirred Bioreactor Mixing Time Simulator
Design the operating point of a stirred-tank bioreactor used in monoclonal-antibody (mAb) production. As you change the scale, impeller diameter, RPM, impeller type, viscosity and aeration, the mixing time, kLa, volumetric power and tip-speed shear update in real time, so you can find a regime that keeps pH and DO uniform without killing the cells.
Bioreactor Operating Conditions
Reactor scale
Preset (initial volume & impeller diameter)
Working volume V
L
Impeller diameter D
mm
D/T ≈ 1/3 (Rushton) to 1/2 (axial-flow) is typical
Agitator speed N
RPM
Impeller type
Changes power number Np and mixing pattern
Liquid viscosity μ
cP
CHO culture ≈ 1.0–1.5 cP; high-density ferment > 10 cP
Liquid density ρ
kg/m³
Aeration Q_g
vvm
vvm = tank volumes/min. Mammalian cells: 0.05–0.5 vvm
Side view of a baffled stirred tank. The impeller circulates the liquid while bubbles rising from the sparger are dispersed across the volume. Colour intensity shows local shear (green → orange → red).
Average shear rate γ̇. Above 2000 1/s, animal cells suffer damage even with Pluronic F-68.
Biopharma Stirred Bioreactor — Mixing Time & kLa Design for mAb
🙋
"Biopharma" usually means antibody drugs, right? How are they actually made, and what's different from a normal small-molecule drug?
🎓
The big difference is that biopharma is not chemically synthesised — you "let cells make it". For a monoclonal antibody (mAb), CHO cells (from hamster ovary) are grown in a big stirred tank filled with culture medium for hundreds to thousands of litres, and you collect the antibody they secrete. So the reactor is not a "chemical reactor", it is a "fish tank where you keep cells happy". That changes what the agitator is for: it has to deliver oxygen and nutrients uniformly without injuring the cells.
🙋
So stronger stirring sounds better, but when I crank up the RPM on the left the tip-speed cell turns red. What is wrong with that?
🎓
That is the biggest contrast with a chemical reactor. Animal cells have no cell wall — just a membrane — so they are extremely shear-sensitive. Empirically, when the Metzner–Otto average shear γ̇ = 11.5·N exceeds 1500–2000 1/s, cell death starts to climb even with the surfactant Pluronic F-68 added as protection. In tip-speed terms the rule of thumb is below 2.5 m/s. A chemical plant might happily run at 5–10 m/s, but you cannot spin a mAb reactor that hard. So the design challenge is "gentle stirring that still keeps things uniform".
🙋
"Gentle and uniform" sounds contradictory — how do you actually achieve it?
🎓
The trick is to make the impeller bigger. Power P scales as N³·D⁵, so a 1.5× larger D needs only 1/2.5 of the speed for the same power input. Going to a large-diameter, low-speed impeller keeps the same P/V while dropping tip speed and shear. Modern mammalian-cell bioreactors typically use Elephant Ear or axial-flow impellers at D/T ≈ 0.5. Microbial fermentation goes the other way: D/T ≈ 1/3 Rushton turbines spinning fast to push kLa over 100 1/h. Same "stirred tank", opposite philosophy because the contents differ.
🙋
On the right, Rushton and Marine impellers give quite different mixing times. What about kLa?
🎓
kLa is set by van't Riet kLa ≈ 0.026·(P/V)^0.4·Us^0.5, so it mostly comes down to P/V and aeration. If you match P/V the impeller choice does not change kLa much. That said, Rushton is excellent at breaking up bubbles and dispersing them, so measured kLa often beats the correlation at high Us. Axial-flow impellers, on the other hand, create strong top-to-bottom circulation, giving short mixing times and stable pH control. That is why large tanks often combine them in a multi-impeller stack — a Rushton at the bottom for dispersion plus an axial impeller above for blending.
🙋
Last question — scale-up from bench to production is famously tricky. What is the most important thing?
🎓
The honest answer is "you cannot keep everything constant" — that is the heart of scale-up. Constant P/V, constant t_m, constant tip speed, constant Reynolds — pick one and the others shift. For mAb the de-facto standard is constant P/V (20–100 W/m³), which keeps kLa and shear reasonably stable, but mixing time still grows 2–3× when you scale 10×. So at 2000 L there is a real risk of a local pH spike right after you inject base. Toggle bench-2L → pilot-200L → production-2000L in this tool while holding P/V the same and you will feel that "something always breaks" reality.
Frequently Asked Questions
Mixing time t_m is the time required for a tracer injected somewhere in the tank to reach near-uniform concentration (typically within ±5%). It is shorter with larger impellers and higher RPM. This tool uses the Ruszkowski-type empirical relation t_m ≈ 5.4·(T/D)²/N (Rushton) and 4.8·(T/D)²/N (axial-flow impellers). In animal-cell culture, slow homogenisation of pH-control liquid or fresh medium can cause local pH spikes or substrate gradients that stress cells; under 30 seconds is a common target.
In aerobic culture, oxygen is poorly soluble in water, so the rate of transfer from bubbles to the liquid often limits cell growth. The available oxygen transfer rate is OTR = kLa·(C* − C), and a small kLa caps the achievable cell density. The van't Riet correlation gives kLa ≈ 0.026·(P/V)^0.4·Us^0.5, dominated by volumetric power P/V and superficial gas velocity Us (proportional to vvm). 10–30 1/h is usually enough for mammalian cells; microbial fermentation or high-density perfusion may need over 100 1/h.
The Metzner–Otto average shear rate γ̇ = 11.5·N (1/s, N in rps) is the common metric. For CHO cells, exceeding γ̇ ≈ 1500–2000 1/s increases cell death even with Pluronic F-68 added as a shear protectant. Tip speed πDN [m/s] is also used, with 2.5 m/s being a rule of thumb for mammalian cells. This tool warns when shearRate > 2000 1/s. To lower shear, use a larger impeller at lower RPM (D ≈ T/2) while keeping the same P/V.
The three classic criteria are "constant volumetric power P/V", "constant tip speed πDN" and "constant mixing time t_m". For animal-cell culture, constant P/V (typically 20–100 W/m³) is the most common choice because it keeps kLa and shear from drifting much. Mixing time, however, always grows on scale-up (D goes up, N goes down), so pH spikes must be confirmed acceptable separately. Switch between bench-2L, pilot-200L and production-2000L in this tool to see how P/V and t_m trade off.
Real-World Applications
Monoclonal antibody (mAb) manufacturing: Antibody drugs for oncology and inflammation support a market well over USD 100 billion per year. mAbs are produced by culturing CHO cells in 2,000–25,000 L stirred tanks for 12–18 days. The typical operating window targeted here is production-2000L, P/V ≈ 30–50 W/m³, kLa ≈ 10–20 1/h, v_tip < 2.5 m/s and mixing time of 30–60 s. Reaching titres above 1 g/L generally requires perfusion (continuous glucose / amino-acid feed plus efficient CO₂ stripping), with low VVM (0.05–0.1) using oxygen-enriched gas.
Single-use bioreactors (SUB): Gamma-sterilised plastic bags combined with rocking or bottom-mounted magnetic agitation have exploded in CMO and Phase I/II clinical manufacturing, where small lots of many products dominate. Single-Use 1000 L sits near the top of the SUB scale. Impellers are typically single-stage Elephant Ear or axial (multi-stage is hard to sew into a bag). Because the film material can leach into the culture (extractables & leachables, E&L), moving from stainless to SUB needs not just kLa / t_m parity but a full comparability study to show the cells behave the same.
Microbial fermentation (insulin, vaccine antigens): E. coli and yeast products grow fast with very high oxygen demand, so they need three-stage Rushton impellers at P/V = 1,000–5,000 W/m³, kLa = 200–500 1/h and v_tip = 5–10 m/s — orders of magnitude harsher than mammalian cell culture. The defaults of this tool target mAb, but setting viscosity to 5–20 cP and RPM around 800 brings the operating point into the microbial regime. Watch how kLa, not t_m, becomes the limiting factor as P/V climbs.
Cell & gene therapy (CGT): Expanding a patient's own T-cells for CAR-T therapies typically uses 20–200 L rocking or stirred SUBs. Total batch is small, but lentiviral vector production needs high-density HEK293 culture, which fits the bench-2L to Pilot 200L range of this tool. kLa requirements are moderate, but balancing CO₂ stripping (aeration) with pH control is hard, and shorter mixing time helps stability.
Common Misconceptions and Pitfalls
First, kLa correlations are not always reliable. The van't Riet correlation used here was derived for pure water + air. Real medium components — antifoam, salts, amino acids, Pluronic F-68 — change interfacial properties enough that measured kLa can be 0.3–2× the prediction. In practice you measure your own kLa = a·(P/V)^α·Us^β coefficients in the actual medium. Treat this tool's kLa as a starting point for design and relative impeller comparison, then recalibrate against the real reactor before locking down operating conditions for production.
Second, do not assume "shorter mixing time = better culture". Lowering t_m by raising RPM increases shear γ̇ = 11.5·N and quickly damages cells. Conversely, even at t_m ≈ 60 s, splitting base addition into "small, continuous, multiple injection ports" can avoid local pH spikes. In real plants the success or failure of a culture often hinges on the base-addition strategy and downstream responses (CO₂ stripping, DO controller tuning) far more than the bare mixing time. Do not over-optimise t_m in isolation.
Third, conditions that worked on the bench do not transfer directly to production. If you tried to keep P/V, kLa and t_m all constant going from 2 L to 2000 L, you would simply have too few free knobs — physics will not allow it. The realistic recipe is "primary + secondary + constraint": hold P/V constant, adjust kLa with aeration, and verify that t_m stays within the acceptable band. Additional issues — headspace CO₂ accumulation, jacket-wall temperature gradients, sampling-port uniformity — only show up at scale. Use this tool as a stirring-physics aid, not as a guarantee that the culture will succeed; the broader system integration decides the outcome.
How to Use
Enter reactor volume (1–2000 L for mAb production scales) and impeller diameter (0.1–0.8 m typical for Rushton or Lightnin turbines).
Set impeller speed (50–500 rpm) and liquid viscosity (0.5–10 cP, accounting for media, cells, and recombinant protein concentration during fed-batch culture).
Read outputs: tank diameter, power input, volumetric power (kW/m³), mixing time (seconds to homogeneity), kLa oxygen transfer coefficient, and impeller tip speed (m/s) for aeration sparger design.
Worked Example
A 200 L mAb bioreactor with Rushton turbine (impeller diameter 0.12 m) operating at 180 rpm in CHO cell medium (viscosity 1.2 cP) yields: tank diameter ≈ 0.68 m, impeller power ≈ 180 W, volumetric power ≈ 0.9 kW/m³, mixing time ≈ 45 s, kLa ≈ 28 h⁻¹, tip speed ≈ 1.13 m/s. This mixing time ensures nutrient and dissolved-oxygen homogeneity before oxygen gradients develop across the culture, critical for cell growth phase (days 2–6 post-inoculation).
Practical Notes
High viscosity feeds (glycerol, peptone stock) during perfusion increase mixing time; reduce speed or add co-solvents to maintain mixing below 60 s and prevent local osmotic stress.
Volumetric power 0.5–1.5 kW/m³ balances oxygen transfer with shear stress; exceed 2 kW/m³ only if cell line tolerates >95% viability loss acceptable (e.g., PER.C6 engineering strains).
Tip speed below 2 m/s reduces impeller cavitation and foaming; sparger ring placement near impeller suction ensures DO readings reflect bulk conditions, not local anaerobic zones.