Process installation with tall stainless steel vessels, pipework and access platforms.
Stainless steel vessels, pipework and access platforms in a real process installation. Vessels and facility utilities are designed together.Photo: bbi-biotech

Three questions prevent misleading comparisons

A stirred tank can use glass, stainless steel or a disposable container and operate in batch, fed-batch or continuous mode. Single-use primarily describes how product-contact components are used, rather than a separate mixing principle. Similarly, perfusion is an operating mode involving medium exchange and cell retention, rather than one vessel design.

A sound comparison starts with three questions: how is movement introduced into the culture, are the cells freely suspended, attached or entrapped, and which substances should pass through the system or remain inside it? Mammalian cells on carriers might grow in a stirred tank or a fixed bed, for example, but their transport pathways differ from those in a free bacterial suspension in either case. An aseptic strategy must be developed separately for each selected combination.

Selection adds a fourth question: where must a substance overcome resistance on its way to a cell? In a free suspension, oxygen may travel from gas across the bubble interface into the liquid and onwards to the cell. In a colonised fixed bed, distribution through the bed and transport into the cell layer add further steps. A hollow fibre can introduce the membrane wall as another resistance. A large exchange area at one point consequently cannot compensate without limit for a restriction elsewhere.

Describe the architecture without ranking it first. Identify moving phases, energy input, gas route, cell location, product route and retention. Then assess whether this arrangement fits the particular task. Two devices with the same name may have different internals, gas paths or measurement options. Conversely, different architectures can fulfil the same biological purpose in different ways. Comparing functional routes is therefore more useful than ranking equipment solely by vessel names.

Operating mode changes the requirements for a given architecture. A stirred tank in batch operation does not necessarily need continuous cell retention. Perfusion additionally requires a selective outlet that retains its performance throughout the intended run. A fixed bed retains cells spatially, but genuine medium exchange still needs a fresh-medium inlet and an outlet. Recirculation through a larger reservoir does not renew the medium: it distributes the existing nutrients and waste substances through a larger total volume.

Every comparison must also define success. If a secreted product is to be removed regularly, passage through retention components matters. If living cells are the harvest, their recovery from the system is central. If an immobilised catalyst is intended to operate for a long period, sustained activity and pressure drop may matter more than recoverable cell number. The architectures below are therefore considered through their transport routes and limits that can actually be investigated.

Stirred tanks: many adjustments, careful coordination

In a stirred tank, a driven impeller creates liquid motion. Impeller design, diameter, position and speed combine with vessel geometry to determine the flow. Baffles can limit rotation of the contents as a body. A sparger distributes gas below or near a suitable impeller; some systems also provide headspace gassing.

The principle allows mechanical power input and gas supply to be adjusted separately within equipment limits. This helps accommodate changing cell numbers and process phases. Nevertheless, the variables remain coupled: gas changes the loading on an impeller, and the impeller affects the bubbles. Shaft arrangements, bearings and seals or a magnetic coupling must suit the aseptic concept. High local stress, foam and heat generation prevent unrestricted increases in power. Properly designed stirred tanks support both microbial processes and cell cultures.

A principal degree of freedom is the ability to drive liquid circulation mechanically even at low gas flow. This can help when a process temporarily needs little gas but still requires rapid distribution of feed or correction liquids. A carrier culture adds the task of maintaining particles in the intended state. An impeller that disperses gas effectively need not produce uniform carrier distribution at the same setting. Assess impellers, gas supply and particle behaviour together.

Average power input provides only limited information about local mechanical exposure. Motion near an impeller, around a bubble and in the main circulation differs. For a sensitive culture, the useful question is therefore not simply maximum rotational speed, but which combination provides adequate transfer and suitable biological results. Data should connect geometry, fill, fluid properties, aeration, agitation state and the actual measured response of the culture.

Hypothetical geometry example: Two similar stirred tanks differ by a factor of two in every length. The larger vessel then has eight times the volume but only four times the geometrically corresponding jacket area. At equal heat load per litre, each unit of that area would need to transfer twice as much heat. An unchanged temperature difference and heat transfer coefficient would not achieve this. Additional area, changed utilities or a different process load would need examination; similar shape alone does not solve the cooling question.

Nor can all agitation parameters be preserved together. Under simplified assumptions of equal power number, equal density and similar geometry, P/V is proportional to N³ · d², where N is rotational frequency in s−1 and d is impeller diameter in m. Doubling d while maintaining P/V requires a speed ratio of 2−2/3 = 0.630. Equal tip speed instead requires a ratio of 0.500. Starting from 300 min−1 in the smaller system gives approximately 189 and 150 min−1, respectively. These are calculated consequences of different objectives, not recommended operating speeds.

A plausible selection case is a free microbial suspension with changing oxygen demand and several feeding phases. The stirred tank offers useful control options if gas transfer, cooling and exhaust handling cover the combined load case. A particular configuration could be excluded because essential internals do not operate at the required initial volume, or because transfer has not been demonstrated in the later viscous medium. That exclusion concerns the configuration and process, rather than all stirred tanks.

Bubble columns and airlift reactors: gas provides the drive

In a bubble column, supplied gas rises through the culture liquid. The bubbles transfer momentum, creating liquid motion and gas transfer without a mechanical impeller. The resulting flow depends on gas rate, geometry and fluid properties. Without additional guidance, circulation is not necessarily uniform throughout the vessel.

An airlift reactor instead provides a defined circulation path with a riser and downcomer. Differences in gas content produce different mean densities and drive circulation. A draft tube can separate the regions within one vessel; alternatively, separate connected limbs provide an external loop. Some gas separates at the top.

Comparison of a bubble column and airlift reactor: the airlift draft tube guides the rising gas-rich liquid and the descending return flow with less gas.
Pneumatic mixing compared. One possible airlift arrangement is shown, with gas supplied in the inner riser. The downcomer need not be completely gas-free. Scroll sideways on small screens or open to enlarge.

Removing the impeller does not remove stress associated with bubbles. Gas rate also couples mixing, mass transfer and gas removal more strongly. For viscous liquids or demanding oxygen requirements, the actually achievable performance needs to be demonstrated.

In an airlift, the driving pressure difference results from the density difference between the two legs over an effective height. It must overcome resistance in turns, the draft tube and connecting paths. Additional gas can change this relationship but cannot improve it indefinitely. If considerable gas also enters the downcomer, the density difference may decrease. A functioning gas disengagement region therefore belongs to the circulation concept. A visually large bubble population in the riser does not establish adequate exchange of liquid.

The central coupling is that gas flow contributes simultaneously to motion, oxygen supply and removal of dissolved gases. When the culture needs less oxygen, total gas flow may not be reducible in the same proportion without changing circulation. Altering gas composition can provide some flexibility, but still affects gas consumption and stripping conditions. A suitable solution follows from the complete gas and flow operating region, rather than from a single dissolved oxygen measurement.

An assumed comparison case involves a low-viscosity suspension with moderate oxygen demand. A pneumatic design merits investigation if it provides sufficiently uniform supply at gas exposure compatible with the culture. If the same culture later develops large aggregates or substantially different flow properties, those final states belong in the investigation. A cell-free demonstration in a thin liquid does not answer the later circulation question. For a process requiring strong mixing at very low total gas input, the close coupling can become a concrete exclusion criterion.

Increasing size changes hydrostatic pressure, gas residence and the paths leading to disengagement. Equal gas volume flow per liquid volume therefore does not guarantee identical circulation. Required information includes riser and downcomer cross-sections, their connections, effective height, gas reference conditions, gas hold-up and suitable circulation or tracer measurements. Foam, liquid carryover and exhaust pressure drop also help define the operating limits. Together, these data show whether the architecture can support the process throughout its development.

A useful technical case analysis considers reducing gas flow while liquid composition remains unchanged. Even if the oxygen reading initially remains sufficient, circulation may slow. A correction liquid dosed later may then distribute differently from the preceding condition. Preliminary selection therefore examines low gas input combined with a continuing need for homogenisation, as well as the most gas-intensive state. Appropriately defined tracer trials can support the assessment. Comparing different addition positions helps establish whether a favourable result comes from an especially accessible location or whether the intended process connections actually receive adequate distribution.

Fixed beds: cells stay in place

A fixed bed holds carriers or a structured growth matrix in a defined position. Medium flows through the void space between carriers. Adherent cells grow on accessible surfaces; immobilised enzymes and entrapped microorganisms provide other possible applications. A comparatively small vessel can contain a substantial attachment area.

Keeping cells in place facilitates retention during medium exchange. However, it also makes supply spatially dependent: downstream regions receive medium whose composition has already changed. Cell growth can narrow flow paths or increase maldistribution. An outlet sample does not automatically describe every part of the bed. Inoculum distribution, pressure drop and harvestability are therefore central questions. If intact living cells are the product, their eventual detachment from the carrier requires particular attention.

Fixed-bed selection requires separate consideration of local and overall supply. The complete loop may contain enough oxygen while individual bed regions receive too little because flow is uneven. Within a cell layer, transport from the flowing liquid to deeper cells adds another path. A higher recirculation flow does not necessarily improve this internal transport proportionally. Available surface, actually colonised surface, spatial distribution and the cell population needing supply must therefore be assessed together.

Hypothetical example of external supply: Medium flows through a bed at 12 L/h with an inlet dissolved oxygen concentration of 0.20 mmol/L. A minimum outlet concentration of 0.12 mmol/L is adopted as a provisional criterion requiring process-specific assessment. With no oxygen addition inside the bed and stationary inventories, the stream could deliver at most 12 · (0.20 − 0.12) = 0.96 mmol/h to the cells. An assumed total demand of 1.50 mmol/h exceeds this capacity. Bed surface area does not change the molar balance.

At unchanged concentration conditions, an external flow of at least 1.50/(0.20 − 0.12) = 18.75 L/h would be needed. Alternatively, inlet conditions or cell load could be changed if the process concept allowed it. None of these alternatives has already been fully calculated or biologically accepted. It would additionally be necessary to show that the higher flow is distributed properly, remains within pressure limits and supplies the cells adequately. The calculation is a necessary preliminary check, not proof of uniform conditions throughout the bed.

Oxygenating the recirculation loop is also distinct from adding fresh medium. A rapid return flow can repeatedly move the same limited nutrient inventory through the bed. Fresh medium and corresponding removal are what change the external nutrient and by-product balance. Documentation should therefore distinguish recirculation flow, fresh-medium flow, harvest flow and external reservoir volumes. The term perfusion rate on its own can otherwise obscure which flow is actually being described.

A concrete hypothetical case is recovery of a secreted product from adherent cells, where complete cell detachment is not a process objective. A fixed bed may fit if inoculation, supply and product removal are demonstrated. If intact cells themselves are the product, insufficient evidence of cell recovery becomes a possible exclusion criterion. Compare harvest yield, cell condition and cells remaining in the bed with a carrier suspension or another appropriate architecture.

The measurement strategy therefore has a concrete task: a reading from the external reservoir describes that reservoir and cannot automatically be assigned to every cell in the bed. Inlet and outlet analyses can estimate overall consumption across the bed if flow, timing and changes in inventory are known. They do not reveal whether some liquid preferentially follows a lightly colonised channel. Additional distribution studies must answer the relevant spatial question. Indirect biomass estimates also require a justified calibration for the colonised state. The selection decision should explicitly identify which cellular and product quantities remain accessible during operation and which can only be checked after disassembly or harvest.

Fluidised beds: flow supports the carriers

In a liquid-fluidised bed, an upward medium stream lifts suitable particles. Fluid forces oppose the particles' weight after allowing for buoyancy. The bed expands and the carriers can move relative to one another. The flow required depends on particle size, density and shape as well as liquid properties.

Movement can improve external mass transfer and reduce persistent preferential flow paths. It also introduces different requirements: carriers must not escape uncontrollably, while collisions and attrition can damage cells or carrier material. Growing biomass changes particle properties as the process proceeds. A pump setting established in a cell-free test is therefore insufficient evidence of sustained performance. Fluidised beds, freely moving biofilm carriers in wastewater treatment and microcarriers in stirred tanks are related but distinct concepts.

A useful quantity is superficial liquid velocity, u = Q/A. Q is liquid volume flow in m³/s, A is the empty reactor cross-section in m², and u is in m/s. This definition does not use only the space remaining between particles. Actual local velocities also depend on void fraction and flow distribution. A change in vessel diameter therefore requires, at minimum, consideration of the changing cross-sectional area.

Hypothetical teaching example: A cylindrical bed cross-section with a diameter of 0.10 m has A = π · 0.10²/4 = 0.007854 m². A recirculation flow of 1.20 L/min equals 0.0000200 m³/s. The result is u = 0.00255 m/s, or 2.55 mm/s. Doubling diameter to 0.20 m multiplies A by four. At unchanged flow, u falls to approximately 0.637 mm/s. Merely preserving u would require 4.80 L/min.

This calculation does not establish minimum fluidisation velocity. Missing information includes carrier density and size distribution, viscosity, shape, void fraction and properties of the colonised bed. It only explains why the same pump flow should not be expected to provide a comparable mechanical effect at a larger cross-section. Equally, the same superficial velocity can produce a different state with different particles. The calculation checks geometry and must then be complemented by appropriate observations.

The operating region includes a sufficiently expanded bed, limited particle loss and acceptable exposure of the culture. Starting, growing and harvesting can each involve different particle properties. A developing cell layer may change diameter and effective density, while dissolved materials can change liquid properties. Observe bed height, pressure drop, particle motion, carryover and biological results. A steady pump indication does not prove that the same bed state is being maintained.

An assumed selection case involves suitable colonised carriers for which external supply and flow through a fixed bed appear inadequate. A fluidised bed then merits investigation as an alternative. If particle contacts adversely affect the culture or carriers cannot be retained reliably throughout the run, this alternative may be excluded. The testable question is whether motion, retention and cell quality can be achieved together, rather than which architecture sounds gentler in an abstract description.

Membrane and hollow-fibre systems: selective separation

A membrane can retain cells while allowing dissolved substances to pass. In hollow-fibre systems, the porous fibre wall separates two spaces. Depending on the configuration, cells are cultured inside or outside the fibres while a medium stream supports their supply. Other arrangements combine a stirred tank with an external filtration loop for cell retention.

The label membrane reactor does not establish which components can pass: pore structure, membrane material, operating conditions and solute properties determine separation. Deposits and cell material can increase resistance. A high flow rate also cannot resolve diffusion limitations within a dense cell aggregate by itself. Pressure conditions, flux, mass transfer, product passage and stress in the circulation loop all need assessment. A wastewater membrane bioreactor uses a related separation principle but has a fundamentally different microbiological operating concept.

External retention separates two tasks: supplying the reactor with medium and maintaining the intended separation. Recirculation through a module may be much greater than the filtrate flow removed. Pump and tubing exposure then depend on circulation, while fresh-medium consumption and the liquid balance depend on exchange and other outlets. Volume outside the main vessel also remains relevant to residence, cellular exposure and inventory calculations.

Hypothetical teaching example: A 2.0 L system receives 4.0 L of fresh medium per day. With equal external outflow and constant volume, the hydraulic exchange rate is D = 4.0/2.0 = 2.0 d−1; the corresponding ideal hydraulic residence time is 0.50 d, or 12 h. Effective assumed retention makes cellular residence different from that of a freely passing liquid inventory. The 12 h must therefore not be described as cell age or mean cell residence time.

If 3.8 L/d leaves as essentially cell-free filtrate and 0.2 L/d as a representative cell bleed, the proportional cell removal caused by that bleed alone is 0.2/2.0 = 0.10 d−1. Under ideal complete retention, its reciprocal of 10 d is a removal timescale. It does not describe growth, death or additional cell losses. The contrast between a 12 h hydraulic time and a 10 d removal timescale demonstrates why retention and medium exchange need separate balances.

Product may also be retained. A simplified sieving coefficient SP = Cfiltrate/Creactor compares simultaneous product concentrations on an appropriate analytical basis. At an assumed reactor concentration of 1.0 g/L and SP = 0.80, the filtrate contains 0.80 g/L. A filtrate flow of 3.8 L/d therefore removes 3.04 g/d; the cell bleed removes a further 0.20 g/d at the same product concentration as the reactor. Total removal is 3.24 g/d. This calculation assumes constant conditions and establishes neither formation rate nor complete product recovery.

A decreasing sieving coefficient can impair the intended harvest even while liquid exchange still works. Adsorption, fouling, changing material properties and analytical performance are possible lines of investigation. Module configuration, operating duration, pressure trends, filtrate flow, retention efficiency and product passage therefore belong together. A short initial test alone is inadequate for a long-running process. A concrete exclusion criterion would be insufficiently demonstrated passage of the target product or an unmanageable change in the retention path.

For hollow fibres, identify which membrane side contains the cells and where measurements are taken. A well-controlled concentration in the flowing fibre lumen initially describes that compartment. Transport resistance can give the cell compartment a different state. The product balance additionally includes samples from both spaces, residual liquid and possible binding to contact surfaces. If the selected analysis covers only the accessible recirculation stream, document that limitation explicitly. Architecture selection therefore also depends on whether the quality attributes required for the objective can be adequately observed under actual operating conditions or supported by appropriate additional investigations.

Rocking and other mixing principles

In a rocking system, a platform tilts a partly filled culture container. Movement of the liquid surface supports gas exchange and mixing. Rocking angle, motion profile, frequency and fill volume need to be considered together; the same movement at a different fill level does not automatically provide the same supply. These systems are used for applications including suspension cell cultures and seed expansion.

Other designs include rotating systems, modules providing additional growth surfaces and solid-state fermentation reactors. The latter cultivate organisms on moist solid material with little free liquid. Moisture distribution, aeration and removal of locally generated heat become particularly relevant. A solid-state reactor is therefore not simply a poorly mixed liquid reactor. Measurement and cleaning strategies must suit the physical phase actually present.

With rocking motion, surface area, liquid depth and the motion regime are closely connected. Adding more liquid to a partly filled container changes the free surface, possible wave forms and liquid pathways. Treating initial and final fill as separate test conditions is consequently useful. Video of the visible wave can document motion but does not by itself demonstrate homogeneity or oxygen availability everywhere in the culture.

Appropriate characterisation connects tracer tests, gas transfer and biological outcomes. Add tracer at a location relevant to the actual dosing route and define the evaluation criterion beforehand. Sampling only near the addition point can make mixing appear more favourable than it is in distant regions. Model and experimental results therefore need explicit spatial and temporal definitions. A published mixing-time number without those conditions has limited value as an acceptance criterion.

A hypothetical selection case is a suspension seed culture growing from a small initial amount to a larger transfer quantity. For a rocking configuration, assess adequate measurement at small fill, supply at maximum cell load and the intended withdrawal. If later operation requires carriers to remain suspended, an additional task appears. Previous suitability for free cells does not demonstrate this new function. An exclusion criterion could be a required particle distribution that has not been achieved within the permitted motion range.

Other moving architectures deserve the same reasoning. Rotating vessels and expanded-surface modules alter contact between cells, liquid and gas in their own ways. For moist solids, the balance may need water content and gas distribution; a mean liquid temperature may not even exist. Comparison only becomes meaningful when the measured quantity and transporting phase fit the architecture. The common basis is the actual supply task, rather than the external movement of the equipment.

Photobioreactors add light delivery

Photobioreactors cannot be described by mixing principle alone. An illuminated stirred tank, transparent bubble column, tubular arrangement or flat panel can each supply phototrophic cultures. The spatial distribution of light adds another requirement to gas and nutrient transfer. As cell density increases, less light reaches the deeper layers.

A large transparent surface can shorten the light path while increasing the surface area that requires cleaning. Long tubing runs create further demands on circulation and gas removal. Closed construction reduces certain entry routes but does not establish an axenic culture. The page on photobioreactors explores light, carbon dioxide supply, oxygen removal and the limits of sterile operation.

The additional geometric question is how much illuminated area serves a given culture volume. For a simplified flat plate illuminated on both sides, neglecting edge areas, Alight/V = 2/b, where b is liquid gap in m. A hypothetical 0.04 m gap gives 50 m−1; a 0.08 m gap gives only 25 m−1. Doubling thickness therefore does not increase illuminated area in proportion to volume. This is a geometric relationship, not a productivity prediction.

Actual light availability also depends on incident spectrum, intensity, surface condition, scattering and absorption within the culture. Cells travel between regions receiving different illumination. The suitability of that time-dependent exposure must be established for the organism and process question. More incident power may also create greater heat load. A thin architecture therefore does not automatically resolve gas removal, temperature control or cleanability.

A hypothetical selection case compares a tubular system and a flat plate for the same required culture quantity. The tubular system is assessed for circulation path, pressure drop, gas exchange locations and cleaning. The plate is assessed for distribution across its width, optical path and heat removal. The appropriate concept follows from the requirements achieved together. A dense, strongly light-limited culture may exclude a long optical path; a location with difficult access may instead restrict assembly and cleaning of large surfaces.

A useful comparison consequently records light conditions at the reactor, geometric area and liquid gap, cell concentration, medium properties, temperature and gas data. Electrical lamp power alone does not measure photon supply usable by the culture. Productivity per illuminated area must likewise not be compared without qualification to productivity per liquid volume. Numerator, denominator and operating duration need identical definitions before architectures can be evaluated fairly.

Compare designs through their likely bottlenecks

Guidance for shortlisting; suitability still requires process-specific assessment
DesignUseful characteristicCheck early
Stirred tankWidely adjustable mixingGas transfer, local stress and heat
Bubble column / airliftNo mechanical impeller requiredCirculation with actual fluid properties
Fixed bedSubstantial attachment areaDistribution and harvest within the bed
Fluidised bedMoving, supplied carriersStable expansion and carrier retention
Membrane systemSelective exchange and retentionFouling and product passage
Rocking systemMixing through container movementGas transfer across the fill range

A useful shortlist starts with the critical task: high oxygen uptake, required attachment area, a sensitive product or light delivery. Sterilisability, sampling and transfer to the intended scale follow. Experiments should deliberately investigate the suspected limits. The comparison of glass, stainless steel and single-use then addresses materials, operation and purchasing.

A first hypothetical comparison concerns a secreted protein from freely suspended cells. The task requires reliable supply, controlled removal and a traceable harvest balance. Both a suitable stirred tank with retention and another appropriately mixed perfusion architecture could enter the assessment. A fixed bed requiring attachment would not become equivalent solely because it offers a high nominal surface area: the culture would first need biological suitability for that arrangement. A technical change can therefore create an additional process-development task.

A second assumed case concerns living adherent cells as the product. Available attachment area matters, but the eventual condition of recovered cells becomes a decisive comparison criterion. A densely packed bed may look useful during cultivation and become the limitation during harvest. A carrier suspension offers different possibilities but introduces particle motion and separation. Selection requires a connected trial from inoculation to harvest; comparing growth alone does not answer the question.

A third case is a microbial culture that becomes viscous. Comparison using water-based values may misrepresent both alternatives. The configuration must provide adequate transport at the later fluid properties while remaining within permissible mechanical and thermal conditions. If no representative liquid or justified transfer method is available for an architecture, its suitability remains unresolved. Missing evidence is recorded as an open question rather than replaced by a particularly favourable individual result.

These cases explain why exclusion criteria precede weighted scoring. Biologically required attachment, necessary product passage, unattainable minimum fill and inadequate harvestability can each be decisive. Only among the remaining technically suitable candidates is comparison of adaptability, operation and workload useful. Keep assumptions explicit: a provisional preference may change when new evidence disproves the limitation initially considered decisive.

Add four practical fields to a comparison table for each alternative: mandatory criteria met, missing evidence, next decisive experiment and consequence of a negative result. This turns an overview into a working tool. For the fixed-bed case, the next trial might examine cell recovery; for the pneumatic culture, mixing at low gas input; for perfusion, product passage over time. If a negative finding would exclude the option, that trial may be particularly valuable before detailed planning. A result affecting only an easily adjusted secondary feature has a different priority. This keeps the sequence of investigations aligned with the actual decision.

A workflow for checking the operating envelope

An operating envelope comprises conditions that can be achieved together. Maximum agitation, highest gas flow and largest volume on separate specification lines do not yet describe a jointly usable operating point. For each shortlisted architecture, describe at least the initial state, greatest expected biological load and final state as connected cases. Add planned transitions, such as starting circulation, connecting a retention module or changing gas composition. The critical condition may lie in a transition rather than at a stationary endpoint.

The workflow starts with a process profile and a drawing of all transport paths. Identify the resistances most likely to matter: gas–liquid transfer, liquid circulation, bed distribution, membrane transport, heat exchange or optical path. For each resistance, define a measurement that could test its importance. Then select appropriate experiments. The purpose is to obtain information that changes the architecture decision, rather than measuring every available characteristic simply because it can be measured.

Preliminary work with substitute liquids addresses geometric and engineering questions. A suitable biological investigation then examines effects the substitute cannot reproduce: colonisation, changing rheology, cell quality, product passage and harvest. Link the results to the intended operating conditions. Successful investigation of an intermediate fill does not demonstrate the complete volume range. Equally, a short trial must not silently be extrapolated to a long operating duration.

The data handover includes geometry and internals, medium or substitute description, temperature, fill, all relevant flows, measurement methods and their time basis. Record biological load, retention condition, cumulative operating duration and deviations as well. The result is a clear statement of which combination works, the evidence behind it and the limit still unresolved. That statement provides the engineering foundation for measurable requirements under choosing a bioreactor.

Frequently asked questions

Is airlift always gentler than agitation?

Airlift does not require a mechanical impeller. However, bubbles, local flows and gas disengagement can still stress cells. Suitability must be checked with the culture and required operating conditions.

Is a perfusion reactor always a membrane reactor?

No. A fixed bed or another separation method can provide cell retention. A membrane is one possible engineering solution.

Can the same reactor design be used for sterile and non-sterile processes?

Yes. A similar mixing principle can appear in both. Construction, connections, medium treatment and working practices together establish the microbiological operating concept.

Terms in this chapter

69 terms explained in the glossary