Quick Operating Model

A microbial fermentation does not run as four separate control loops. pH, dissolved oxygen (DO), aeration and agitation continuously influence one another because the culture changes its oxygen demand, carbon dioxide output, acid/base balance, broth viscosity, foam behavior and heat load as biomass develops.

The most useful way to operate a fermenter is therefore to read these variables as a system. A falling DO signal may indicate that oxygen demand has increased, but the correct response is not automatically “increase RPM.” The process may also need more airflow, oxygen enrichment, a different feed rate, better foam control, more cooling, or simply verification that the DO probe and gas path are functioning correctly. Likewise, a pH correction can change ionic load and feed volume, while a feed step can immediately increase oxygen demand and force the DO control cascade to react.

This article uses a control-room style approach rather than a component-by-component equipment guide. The goal is to show how the main variables interact over time, what common signal patterns mean, and how to build a practical control strategy for repeatable microbial fermentation.

1. Read the Fermentation as a Dynamic System

Figure 1. A systems view of the main fermentation control variables.

At inoculation, oxygen demand is often modest because biomass is still low. As the culture moves into active growth, oxygen uptake can rise rapidly. If the organism is aerobic, the bioreactor must transfer oxygen from the gas phase into the liquid fast enough to match that demand. This is where aeration and agitation begin to work together.

Aeration delivers gas to the vessel and creates the raw supply of oxygen. Agitation distributes bubbles, renews the liquid around them, keeps cells and nutrients mixed, and helps reduce concentration gradients. If either side is insufficient, the culture can experience local oxygen limitation even when the other setting looks high.

The DO probe is therefore best understood as a process-response signal. It does not create oxygen. It reports the balance between oxygen entering the liquid and oxygen being consumed by the culture. When oxygen uptake rises faster than oxygen transfer, DO tends to fall. When oxygen transfer exceeds demand, DO tends to recover.

That simple relationship is the center of the control strategy:

DO trend = oxygen transfer capability minus biological oxygen demand.

A stable DO setpoint can be achieved through many combinations of airflow, agitation, gas composition and, in some systems, pressure. The engineering task is to choose a response sequence that protects the culture while using the available equipment efficiently.

OLLITAL's current laboratory glass bioreactor for bacteria and cell culture is described with associated control among DO, stirring speed, feeding, airflow and tank pressure, which is the kind of multi-variable relationship that matters during an actual fermentation run.

2. DO Is a Result, Not a Single Control Knob

One of the most common interpretation errors is to treat a low DO reading as if it points to only one cause. In practice, a DO decline can come from several different mechanisms:

biomass concentration has increased and oxygen uptake is higher;

a carbon or nutrient feed has accelerated metabolism;

airflow has dropped or a filter/gas path is restricted;

agitation is insufficient for the current broth properties;

broth viscosity has increased and mass transfer has become more difficult;

excessive foam or antifoam addition has changed gas-liquid behavior;

the DO probe has drifted, become fouled or been calibrated incorrectly.

This is why trend context matters. A DO value that falls at the same time as a scheduled feed begins tells a different story from a DO value that suddenly collapses while agitation, airflow and feed are unchanged.

For aerobic microbial fermentation, the control system often uses a cascade: maintain a DO target by progressively changing one or more engineering variables. A typical sequence may begin with agitation, then increase airflow, then introduce oxygen enrichment if that capability is available. The exact order should be chosen for the organism, vessel geometry, impeller system and site utilities rather than copied as a universal recipe.

The OLLITAL laboratory fermentation bioreactor range lists temperature, DO, agitation speed, pH-related dosing, rotameter-based gas control and peristaltic pumps as part of the control architecture. That combination is useful because operators can observe how biological demand and mechanical response move together rather than viewing DO in isolation.

3. Agitation: More Than “Higher RPM Means More Oxygen”

Increasing agitation can improve mixing and oxygen transfer, but the operating effect is broader than the RPM number alone.

First, agitation changes how quickly nutrients, acid/base additions and dissolved gases are distributed through the broth. A vessel can show acceptable bulk pH while still developing local concentration zones if additions are poorly mixed. Faster and more effective mixing shortens the time required for an addition to become representative of the full vessel.

Second, agitation influences bubble break-up and gas dispersion. Smaller and more evenly distributed bubbles can increase gas-liquid contact area, helping oxygen transfer. But the result depends on impeller type, sparger design, liquid viscosity, fill level and gas flow. Two reactors at the same RPM may not deliver the same mixing intensity or oxygen-transfer performance.

Third, agitation creates shear. Many microbial cultures tolerate higher shear than fragile animal or plant cells, but mechanical stress still matters for morphology-sensitive organisms, filamentous cultures and broths whose rheology changes during growth. Increasing RPM should therefore be treated as an engineering intervention with both benefits and costs.

A useful operating question is not “What is the maximum RPM?” but “How much agitation is required to maintain mixing and oxygen-transfer performance at this stage of the batch without creating unnecessary shear, heat or foam?”

This is particularly important when a fermentation moves from a low-viscosity early phase to a more concentrated or non-Newtonian broth. The same agitation setting can behave very differently later in the run.

4. Aeration: Oxygen Supply, Gas Distribution and Foam Are Linked

Aeration seems straightforward because it is often shown as a flow value on a rotameter or flow controller. In reality, gas flow changes several things at once.

More airflow can increase oxygen availability and gas-liquid contact, but it can also increase foam formation, strip volatile components, alter carbon dioxide removal and change the effective mixing pattern. If the broth already foams heavily, increasing gas flow may solve one limitation while creating another.

The gas path should also be treated as a system. Supply pressure, sterile filters, tubing, valves, sparger condition and exhaust resistance all influence what actually reaches the culture. An operator who only reads the rotameter may miss a partially blocked filter or a downstream restriction.

In a simple laboratory setup, air may be the primary gas. More advanced configurations can add oxygen, carbon dioxide or nitrogen depending on process strategy. OLLITAL's current stainless steel bioreactor system lists four gas lines for Air/O2/CO2/N2 on the reviewed configuration and also lists optional gas-mixing and thermal mass flow control functions. Those capabilities become relevant when a process requires more flexible control than a single air line can provide.

The important point is that “more aeration” is not always the best first response. If airflow is already high and foam is difficult to control, oxygen enrichment or an agitation adjustment may be a more efficient intervention. Conversely, increasing agitation when the gas supply itself is inadequate can waste motor power without fully solving oxygen limitation.

5. pH Control Changes More Than pH

Microbial metabolism changes pH as substrates are consumed and acids, bases or other metabolites accumulate. Automated pH control usually corrects the deviation by adding acid or alkali through peristaltic pumps, but the correction itself becomes part of the process.

Every dosing event changes liquid volume slightly. Repeated additions can change ionic strength, osmolality and broth composition. Poor mixing can create a temporary high-concentration zone near the dosing point, especially if the corrective solution is strong. In a fed-batch process, pH control and nutrient feeding can therefore interact even when the software treats them as separate loops.

pH can also be connected indirectly to oxygen demand. A culture that receives more substrate may grow faster, consume more oxygen and generate more acidic or basic metabolites. The operator may first observe a feed event, then a DO decline, then increased agitation or airflow, followed by a pH correction. Those are not independent events; they are a chain.

For this reason, evaluate the timing of pH pump activity alongside DO, airflow, RPM and feed trends. If base consumption suddenly increases, ask what changed biologically before simply increasing the concentration or dosing rate.

The reviewed OLLITAL glass-system information describes PID pH control through built-in peristaltic pumps and also supports process data recording. That pairing is valuable because the quantity and timing of acid/base additions can be interpreted against the rest of the batch history.

6. The Fermentation Timeline: Five Operating Phases

Figure 2. Illustrative batch timeline showing why control signals must be read in context.

Instead of using one control strategy for the entire batch, it is often more useful to think in phases.

Phase A: Inoculation and Stabilization

Immediately after inoculation, the priorities are stable temperature, verified pH/DO measurement, gentle but sufficient mixing and confirmed sterile gas flow. Oxygen demand may still be relatively low. Large control changes during this phase can make it harder to distinguish normal adaptation from an instrumentation problem.

Phase B: Oxygen Demand Begins to Rise

As biomass increases, DO may start trending downward. This is the first point where the cascade logic becomes visible. If agitation responds, observe whether DO recovers smoothly or continues to decline. If the system needs progressively more RPM for the same DO target, biological demand is probably increasing.

Phase C: High-Growth or High-Demand Window

This is often the most demanding part of an aerobic microbial run. Airflow, agitation, heat removal and antifoam control may all be more active. The key is to avoid “control fighting,” where several loops make large corrections at the same time and create oscillation.

Phase D: Feeding and Process Intensification

In fed-batch operation, a feed event can create a rapid shift in metabolism. DO may fall, carbon dioxide generation may increase and pH demand may change. If the feed rate is increased faster than the oxygen-transfer system can support, oxygen limitation can become a process-design issue rather than a controller-tuning issue.

Phase E: Late Fermentation and Harvest Decision

As substrate becomes limiting, metabolism changes or product formation moves into a different phase, DO may rise even though agitation and airflow are unchanged. A rising DO late in the run can therefore indicate lower oxygen uptake rather than “better aeration.” Interpret the direction of the DO trend together with feed status, pH behavior and process objectives.

7. Designing a Practical DO Cascade

A DO cascade is simply an ordered response plan for maintaining the DO target. The most important design choice is the sequence.

One common approach is:

operate at a defined baseline agitation and airflow;

increase agitation within the validated operating range as DO begins to fall;

increase airflow when additional gas transfer is required;

use oxygen enrichment if the equipment and process strategy support it;

reconsider feed intensity if oxygen demand still exceeds available transfer capacity.

Another process may prefer airflow before agitation because shear must be minimized. A highly foaming broth may do the opposite and limit airflow earlier. There is no single cascade order that is correct for every organism.

The cascade should also avoid abrupt steps. Large sudden changes in RPM or gas flow can disturb foam, pressure and mixing. Smooth ramps or PID-controlled responses are often easier to interpret and reproduce.

When selecting equipment, ask whether the controller can associate DO with agitation, airflow, feeding and other variables in the intended order. OLLITAL's current glass bioreactor page specifically describes DO association with stirring motor speed, feeding, airflow and tank pressure, while its stainless-steel system provides additional gas-line flexibility. These features can support a staged control strategy, but the final sequence still needs to come from the process-development plan.

8. A Troubleshooting Matrix: Start With the Pattern, Not the Alarm

Figure 3. Diagnostic patterns for interpreting DO and pH behavior before changing the process.

The quickest way to misdiagnose a fermentation is to react to a single alarm without looking at the trend before and after it.

This matrix is intentionally diagnostic rather than prescriptive. The purpose is to narrow the cause before changing several variables at once.

9. Feed Rate Is Often the Hidden Fifth Control Variable

In a fed-batch fermentation, feed rate can be the strongest driver of oxygen demand. It is therefore difficult to optimize DO control without considering feeding.

If feed is increased, the culture may metabolize faster and consume oxygen more rapidly. The DO cascade reacts by increasing agitation, airflow or oxygen enrichment. If the feed continues to rise until the cascade reaches its engineering limits, DO falls below target. At that point the process is not simply “short of RPM”; the biological demand has exceeded the available oxygen-transfer capacity.

A practical control philosophy is to define feed and oxygen transfer together. If the goal is high cell density, the feed strategy should respect the available gas, agitation and cooling capacity. This helps prevent a situation where the fermenter is continuously operating at the edge of its mechanical limits.

OLLITAL's current control descriptions include feeding functions and the ability to associate feeding with DO and pH-related parameters on selected systems. That type of linkage is useful for process development because the controller can support coordinated actions rather than isolated pump operation.

10. Foam and Antifoam Can Distort the Oxygen Story

Foam is more than a housekeeping issue. It affects gas disengagement, exhaust filtration, level interpretation and sometimes oxygen transfer. A heavy foam layer can create false assumptions about liquid level and increase the risk of contamination if it reaches the exhaust filter.

Antifoam can solve the immediate foam problem but can also change gas-liquid transfer behavior. Large or repeated antifoam additions may alter bubble characteristics and make oxygen transfer less efficient. That means an operator can see a sequence such as:

If only the final DO value is reviewed, the reason for the change may be missed.

For repeatable work, record antifoam additions as part of the batch history and compare them with aeration, agitation and DO trends. If every batch begins requiring more antifoam at the same growth stage, the process may need a different aeration or foam-management strategy rather than simply a larger antifoam reserve.

11. What to Log for Batch-to-Batch Comparison

A strong control strategy depends on trend history. The minimum useful batch record should connect biological events to engineering response.

Capture at least:

pH setpoint and measured pH;

DO setpoint and measured DO;

agitation command and actual speed;

airflow and any oxygen-enrichment action;

temperature and cooling/heating activity;

acid/base pump activity;

nutrient feed rate and cumulative feed;

antifoam additions;

key sampling times and analytical results;

alarms, manual interventions and operator notes.

The goal is not to create more data for its own sake. The goal is to reconstruct cause and effect. If one batch reaches the same final result with lower agitation and less antifoam, the trend record can help explain why.

The reviewed OLLITAL systems list process data functions, touch-screen control and USB data export on selected configurations. For laboratories moving from trial-and-error operation toward formal process development, that history is often as important as the final setpoint values.

12. How to Tune the Process Without Changing Everything at Once

When a batch underperforms, changing pH, DO setpoint, airflow, agitation and feed simultaneously makes the next result difficult to interpret. A better approach is controlled iteration.

Start with a baseline recipe and identify the main limitation. If oxygen transfer is the problem, evaluate the DO cascade first while holding feed and pH strategy as constant as practical. If foam is the dominant issue, investigate airflow distribution and antifoam response before changing unrelated parameters.

Useful experiments include:

comparing two agitation ranges at the same airflow strategy;

comparing two airflow strategies at the same feed profile;

testing oxygen enrichment only after air/RPM behavior is understood;

examining pH dosing trends before and after a feed change;

testing whether a slower feed ramp prevents DO collapse;

comparing antifoam demand under different aeration patterns.

This stepwise approach creates process knowledge that can later support scale-up. It also helps separate a biological limitation from an equipment limitation.

13. What Changes When the Same Process Moves to Stainless Steel

The control relationships remain, but the engineering implementation changes. A larger stainless-steel vessel has different mixing geometry, gas distribution, heat-transfer area, pressure behavior and utility demands. The RPM number from a glass laboratory vessel cannot simply be copied to a larger tank.

What should transfer is the control objective: maintain adequate mixing, oxygen supply, pH stability and temperature while protecting the organism and preserving reproducible process behavior.

The OLLITAL stainless steel bioreactor adds a broader gas-control configuration and uses a recirculating pump and heat exchanger for heating/cooling on the reviewed model. Those differences illustrate why the same biological target can require a different engineering implementation at larger scale.

When planning scale-up, record which variable became limiting in the laboratory. If the lab process already needs maximum agitation and airflow to maintain DO, the scale-up project should address oxygen-transfer capacity directly rather than assuming a larger vessel will automatically provide more headroom.

14. A Simple Control-Room Checklist Before the Next Batch

Before starting a new microbial fermentation, confirm the following:

pH and DO probes are calibrated and installed correctly;

baseline airflow is stable and the gas path is unobstructed;

agitation responds smoothly throughout the intended operating range;

the DO cascade order and limits are documented;

acid, base, feed and antifoam pumps have the correct tubing and sufficient volume;

the feed strategy does not demand more oxygen than the system can provide;

foam response is defined before high aeration is required;

data logging is active before inoculation;

manual intervention rules are clear so operators do not change multiple variables simultaneously;

the batch review will compare trends, not only final yield or endpoint values.

FAQ

Is DO controlled directly by the DO probe?

No. The probe measures dissolved oxygen, but the controller maintains the target by changing engineering variables such as agitation, airflow, oxygen enrichment or other linked actions. The probe is the feedback signal, not the oxygen source.

Should agitation or airflow increase first when DO falls?

It depends on the process. Agitation-first cascades can improve mixing and bubble dispersion, while airflow-first strategies may reduce mechanical shear. Foaming tendency, gas capacity, impeller design and organism sensitivity should determine the order.

Why can DO fall immediately after increasing feed?

More available substrate can increase microbial metabolic activity and oxygen uptake. If oxygen demand rises faster than the bioreactor can transfer oxygen into the broth, the DO signal falls and the cascade must respond.

Can higher airflow always fix oxygen limitation?

No. Gas flow has practical limits and can increase foam, exhaust load and gas stripping. If oxygen transfer is constrained by mixing or broth properties, simply increasing airflow may provide diminishing returns.

Why does pH control matter when troubleshooting DO?

pH correction reflects metabolism and introduces additional liquid and ions into the broth. A change in pH dosing can indicate a metabolic shift that also changes oxygen demand. Reviewing pH pump activity beside DO and feed trends can reveal these interactions.

What data should be compared between successful and unsuccessful batches?

Compare DO, pH, agitation, airflow, feed, antifoam, temperature and operator interventions over time. The shape and timing of the trends usually reveal more than comparing only the final setpoints or endpoint result.

Conclusion

The most important lesson in microbial fermentation control is that pH, DO, aeration and agitation are coupled variables. DO reflects the balance between oxygen transfer and biological demand. Agitation changes mixing, gas dispersion and shear. Aeration changes oxygen supply, foam and gas removal. pH correction reflects metabolism and can interact with feeding and oxygen demand. Feed rate can push the entire system toward or beyond its transfer capacity.

A robust fermentation strategy therefore focuses on relationships, response order and trend interpretation rather than isolated setpoints. When the control system, gas hardware, agitation, pumps and sensors are configured around the process, the operator can see not only what changed, but why it changed.

Need help translating a fermentation recipe into a practical control strategy?

Send OLLITAL your microorganism, target working volume, feed plan, pH/DO targets, gas requirements and current operating challenge. The engineering team can review a suitable control configuration around the actual process rather than quoting a generic fermenter.

Contact OLLITAL