Bioreactor for the Cultivation of Mammalian Cells

Abstract

The present invention relates to large-scale bioreactors having at least two impellers, large-scale bioreactor systems and methods for the large scale cultivation and propagation of mammalian cells using these bioreactors.

Claims

1-18. (canceled)

19. A bioreactor for the cultivation of mammalian cells, wherein the bioreactor has a volume of at least 4000 l and at least one top impeller and at least one bottom impeller, wherein the top impeller is a hydrofoil impeller.

20. The bioreactor according to claim 19, wherein the top impeller is a three bladed hydrofoil design impeller, and the bottom impeller is a four pitched-bladed high solidity impeller.

21. The bioreactor according to claim 19, wherein the impeller to tank diameter ratio is at least 0.35 and at most 0.55.

22. The bioreactor according to claim 19, wherein the at least one top impeller has a power number (N.sub.p) between 0.1 and 0.9.

23. The bioreactor according claim 19, wherein the at least one top impeller has a flow number (N.sub.q) between 0.4 and 0.9.

24. The bioreactor according to claim 19, wherein the at least one bottom impeller has a power number (N.sub.p) between 0.5 and 0.9.

25. The bioreactor according to claim 19, wherein the at least one bottom impeller has a flow number (N.sub.q) between 0.50 and 0.85.

26. The bioreactor according to claim 19, wherein the bioreactor has at least one sparger.

27. The bioreactor according to claim 19, wherein the bioreactor has at least one baffle.

28. The bioreactor according to claim 19, wherein the bioreactor has at least two ports for alkali addition.

29. The bioreactor according to claim 19, wherein the bioreactor has a volume of at least 10000 l.

30. The bioreactor according to claim 21, wherein the impeller to tank diameter ratio is at least 0.40 and at most 0.48.

31. The bioreactor according to claim 21, wherein the impeller to tank diameter ratio is at least 0.44 and at most 0.46.

32. The bioreactor according to claim 28, wherein said at least two ports are spatially separated from one another.

33. The bioreactor according to claim 19, wherein the bioreactor has a volume of at least 20000 l.

Description

[0091] The present invention is illustrated in more detail in the following examples and the accompanying figures.

[0092] FIG. 1 shows a bioreactor according to the invention. 1 is the bioreactor. 10 is the diameter of the tank (T). 20 is the total straight height of the bioreactor (H). 30 is the base height of the bioreactor (H.sub.b). 40 is the head height of the bioreactor (H.sub.h). 50 is the liquid height at the maximum operating volume (H.sub.L). 60 is the top impeller diameter (D.sub.top). 68 is the top impeller. 70 is the bottom impeller diameter (D.sub.bottom). 78 is the bottom impeller. 80 is the clearance between tank bottom and centre line of the bottom impeller (D.sub.c). 90 is the impeller separation (D.sub.s). 100 is the clearance of the top impeller below the liquid surface (D.sub.o). 108 is a sparger. 110 is the sparger to tank bottom clearance (S.sub.c). 120 is the sparger to bottom impeller clearance (D.sub.cS.sub.c). 128 is a baffle. 138 is a port located at the lower ring. 148 is a port located at the centre-line of the top impeller 68.

[0093] FIG. 2 shows a bioreactor system of the present invention. 111 is a bioreactor with a volume of 1000 l. 11 is a bioreactor with a volume of 4000 l. 1 is a bioreactor according to the invention with a volume of 20 000 l.

EXAMPLE 1: 20 000 L BIOREACTOR

[0094] The 20 000 l bioreactor is operated in batch and fed batch mode for to 15 days for the cultivation of mammalian cells. The mammalian cells are kept in a homogeneous suspension by agitation via an impeller system.

Vessel Geometry

[0095] The vessel geometry for the 20 000 litre bioreactor was determined by an iterative design basis in which the maximum working volume, freeboard straight side distance, aspect ratio H.sub.L/T and impeller to tank diameter, D/T ratio are altered until an acceptable aspect ratio is achieved.

Bioreactor Aspect Ratio H.SUB.L./T

[0096] This critical design parameter allows characterisation of bioreactor geometry. Tanks with higher aspect ratio offer longer gas residence time allowing greater K.sub.La. However increased head pressure can cause build up of soluble gases. Smaller aspect ratio H.sub.L/T in tanks can lead to shorter gas residence time requiring greater gas flow for aeration resulting in greater foam build up. Impeller driven agitation to increase K.sub.La is also limited by H.sub.L/T as surface breakage and vortex creation will occur at lower impeller revolutions in a low aspect ratio. Thus choice of aspect ratio is largely experience based with some thought on issues highlighted in table 1.

TABLE-US-00001 TABLE 1 Summary of effect of varying aspect ratio Process factor High aspect ratio Low aspect ratio Radial mixing More effective Less effective Mixing time Higher Lower Oxygen transfer rate Determined by Determined by dissolved oxygen dissolved oxygen control control Gas flow rate Lower Higher Cell damage Less More Carbon dioxide stripping Less effective More effective Pressure variations Higher Lower Ease of scale More difficult away More difficult away up/scale down from currently used from currently used (access to scale data) aspect ratios aspect ratios Cleanability Not affected directly Not affected directly by aspect ratio by aspect ratio Volume flexibility Less More

[0097] Table 2 describes the aspect ratios in the 20 000 litre bioreactor at various operating volumes during normal processing. The aspect ratios have been tested at 500 litre scale and provided the superficial gas velocity and power per unit volume are kept constant the K.sub.La remains constant.

TABLE-US-00002 TABLE 2 Key operating volumes and aspect ratios in the 20000 litre bioreactor Volume, Liquid head, Aspect ratio, L mm H.sub.L/T Pre-Inoculation 13913-17096 2458-2977 0.88-1.07 Post Inoculation 17391-19231 3025-3325 1.08-1.19 Harvest 20000-21739 3451-3734 1.23-1.34

Tank Diameter

[0098] The tank diameter is altered to obtain the optimal aspect ratio H.sub.L/T. Changes to tank internal diameter (ID) are limited by acceptable aspect ratio and plant footprint. The ID is 2.794 m.

Tank Height

[0099] Tank height is determined from the maximum operating volume, aspect ratio H.sub.L/T, freeboard straight side length, base and top plate design. The final tank height is a compromise value determined from volumetric contingency for foam, plant height and impeller shaft length. The tank height from base to head tan line is 4.933 m.

Freeboard Height

[0100] The freeboard height is defined as the length of straight side above the liquid head when the bioreactor is filled to it's maximum operating volume. This is determined by taking into account the extent of: [0101] Foam build up during operation. [0102] Gas hold up at maximum allowed agitation and aeration. [0103] Errors in metering liquid.

[0104] In absence of knowing the exact contribution of each with piloting the process at full scale an estimate is usually made. The amount of freeboard height is balanced with the desire to reduce the impeller shaft length for a top-driven system, where extra length can complicate the design and selection of available mechanical seals, the requirement for steady bearing or stabilising impeller rings. A minimum freeboard height of 1000 mm (or 6100 litre volumetric capacity or 28% v/v of the maximum operating volume) is therefore used.

Head and Base Plate

[0105] The selection of head and base plate design was made with a consideration for desired mechanical strength, free draining clean design and fluid flow. Maintaining consistent plate design between scale down and full scale will contribute towards maintaining geometric similarity. The base plate is of American Society of Mechanical Engineers Flanged and Dished (ASME F&D) design. The head-plate design accommodates a manway or a flanged head plate to allow access/removal of the impellers.

Bioreactor Agitation Requirement

[0106] The agitation of the bioreactor is to achieve rapid mixing, maintain homogeneity, maintain mammalian cells in suspension and gas bubble dispersion. The underlying issue with achieving the above objectives is minimising cell damage through shear forces originating from impeller geometry and eddies or vortices created behind the impeller blades. A compromise of the above objectives can be achieved by selection of an appropriate impeller type.

Bottom Versus Top Driven Impeller Shaft

[0107] The decision to drive the agitator shaft from the top or the bottom of the bioreactor is important and is determined following a review of a number of issues highlighted in table 3.

TABLE-US-00003 TABLE 3 Key design issues for selection of top versus bottom entry of impeller shaft Top entry Bottom entry Shaft Length Long Short Shaft Weight High Low Shaft Diameter Larger Smaller Impeller shaft on-site Greater plant height Less plant height installation and removal for servicing and repair Exposure of cell culture to No exposure Exposure moving and stationary seal faces .sup.1Pressurization between Lower Higher due to seal and vessel the liquid head Seal Lubricant leakage rate Lower Higher Base plate Design Simple Complex Sparger to tank bottom Unrestricted Restricted positioning CIP validation Simple Complicated by submerged mechanical seal Scale up and scale down Consistent Inconsistent consistency with lab and with lab and pilot scale pilot scale .sup.1Pressure differential between seal and bioreactor critical for lubrication and cooling.

[0108] Top-entry impeller shafts tend to be longer than bottom-entry, which results in the shaft being heavier and larger diameter. Additionally the shaft length together with the inherent clearance between the two faces of the mechanical seal may dictate the requirement for steady bearings or stabilising ring to prevent excessive shaft wobble. Service and maintenance are affected by the available space around the agitator, gearbox and seal assembly, and on-site shaft installation and removal is limited by plant height.

[0109] The protrusion of the seal and impeller shaft at tank bottom restricts the placement of the sparger near the tank bottom. This dimension affects the tank hydrodynamics and therefore its amenability to change is important in specifying an optimal design.

[0110] The downwards load of down pumping impellers together with the liquid head have an accumulative greater load (compared to up pumping or top-entry shaft) between the moving and stationary faces of the seal resulting in greater wear of the seal faces. Furthermore loss of over pressure in the condensate line supplying the seal can result in the culture seeping into the seal. This makes the subsurface seal a less sanitary design.

[0111] The submerged seal complicates the design of a free draining bioreactor by compromising the position of the harvest drain valve. Secondly the diameter of the harvest nozzle may be restricted thus restricting the flow rate of harvest stream. Therefore a top entry impeller shaft is used in the 20 000 litre bioreactor.

Baffles

[0112] The baffle requirement for centre mounted impeller is critical to prevent vortex formation. The critical issues related to baffles are baffle number, baffle width (W), baffle length (H.sub.baffle) and baffle to tank wall clearance (W.sub.c).

[0113] The recommendation for four equally spaced baffles that are 0.1T or 279 mm wide 1.1HH.sub.h or 3882 mm tall and have a baffle to tank wall clearance, W.sub.c of 0.01T or 28 mm.

[0114] The thickness of baffle is not specified but the thickness needs to ensure rigidity to the radial component of the fluid flow. Additionally thickness needs to ensure the baffle plates are not warped during SIP thereby affecting the baffle to tank wall clearance.

Impeller Type

[0115] High shear, such as Rushton (or Rushton-type), impellers offer high power dissipation for gas dispersion but lack in axial flow necessary for mixing and homogeneity. Additionally, agitation from high shear impellers suffers from dangers of excessive cell damage.

[0116] Table 4 shows the impellers tested at lab scale (12.2 litre) that gave equivalent hydrodynamic and cell growth performance. The hydrofoil is mounted above the high solidity pitched blade impeller.

[0117] The Lightnin A310 and A315 at the D/T ratio described in table 4 are used in the bioreactor.

TABLE-US-00004 TABLE 4 Impeller types short-listed for scale down study Impellers D/T ratio .sup.1N.sub.p/.sup.2N.sub.q Vendor Description A310 0.44 0.30/0.56 Lightnin Three bladed hydrofoil design A315 0.46 0.75/0.73 Lightnin Four pitched- bladed high solidity impeller SC-3 0.40 0.90/0.90 Chemineer Three bladed hydrofoil design 3HS39 0.46 0.53/0.58 Philadelphia Four pitched- Mixers bladed high solidity impeller .sup.1N.sub.p is characteristic impeller power number. It is a measure of an impeller efficiency to impart the kinetic energy of the rotating impeller blades to the fluid. It is important in quantifying the gas dispersion .sup.2N.sub.q is characteristic impeller flow number. It is a measure of pumping ability of the impeller and is important in quantifying fluid bulk movement.

Impeller to Tank Diameter, D/T Ratio

[0118] The diameter for axial flow impellers is recommended to be less than 0.5T. A diameter greater than this results in disruption in axial flow, hence poor agitation and aeration.

[0119] Power dissipation into the bioreactor and Reynold's number also need to be sufficiently high to maintain a turbulent (loaded) regime. Therefore the selection of impeller diameter is a compromise between choosing large enough diameter to ensure adequate homogeneous mixing without exceeding the hydrodynamic characteristics of the bioreactor. These include throttling axial flow, insufficient power dissipation, exceeding upper limits of impeller tip speed and creation of poorly mixed laminar zone.

[0120] Once a diameter is selected, than maintaining constant D/T ratio is critical between scale down pilot vessels in order to maintain the central assumption of scale studiesthat of maintaining geometric similarity.

[0121] The K.sub.La scale up correlation at 12.2 litre has been determined for the four impellers at the D/T ratios shown in table 4. From a geometric similarity standpoint A310 diameter of 1.229 m (D/T of 0.44) and A315 diameter of 1.285 m (D/T of 0.46) is recommended. However a manway diameter can restrict the largest impeller diameter that can be installed and removed to 1.219 m. Therefore A310 and A315 to be 1.219 m diameter are used thereby keeping with ease of impeller installation and removal and maintaining close to the geometric similarity proposed in scale down study.

The Impeller Clearance, D.sub.c and Spacing, D.sub.s

[0122] The spacing between impellers in a bioreactor with multiple impellers is an important dimension to consider. For a bioreactor with dual Rushton turbine (radial flow) the ungassed power consumption is equivalent to a single impeller when the dual impeller are spaced less then 0.5D along the shaft. At a spacing of 2D the power consumption becomes adductive. Thus efficiency of the impeller is reduced when the impeller spacing becomes less then 0.5D and the requirement for multiple impellers becomes unnecessary. It is important to note that impeller spacing also impacts on the potential of creating dead zones (poorly mixed zones) within the bioreactor. An additional constraint on the choice of impeller spacing is discrete working volumes required within the bioreactor.

[0123] The impeller spacing, D.sub.s, of 1.229D.sub.bottom (1498 mm) allows both impellers to remain submerged at the lowest post-inoculation volume of 17392 litres with liquid head above the upper impeller, D.sub.o, of 0.5D.sub.top (615 mm) and off bottom clearance, D.sub.c, of 0.75D.sub.bottom (913 mm).

[0124] Table 5 highlights volumes that will form liquid surfaces or lower liquid cover, above the impellers. Agitation needs to be modified to avoid foaming at these critical volumes.

TABLE-US-00005 TABLE 5 Key operating volumes that cause interaction with impellers and liquid surface Interaction Volume, L Potential Operation Submerge top impeller with 17399 Minimum post inoculation volume 17391 L 0.5D.sub.A310 liquid cover Liquid surface touching top 13973 Pre-inoculation volume of 13913 L edge of top impeller liquid surface breakage Liquid surface touching 13283 Bolus addition of pre-inoculation bottom edge of top impeller medium will pass through this liquid head Submerge bottom impeller 8381 Bolus addition of pre-inoculation with 0.5D.sub.A315 liquid cover medium will pass through this liquid head Liquid surface touching top 5592 Bolus addition of pre-inoculation edge of bottom impeller medium will pass through this liquid head Liquid surface touching 3291 Bolus addition of pre-inoculation bottom edge of bottom medium will pass through this liquid head impeller .sup.(1) Minimum operating volume with lower impeller submerged is 8379 litres and minimum operating volume with both impellers submerged is 17399 litres .sup.(2) The operating volume range is 13913 to 21739 litres.

Clearance of Top Impeller Below Liquid Surface, do.

[0125] The breakage of the impeller blade above liquid surface is undesirable as this will make the flow and power dissipation of the impeller ineffective. In addition it will create unknown K.sub.La values due to significant surface entrapment of headspace gas into the fluid and excessive foam. D.sub.o is 0.3D for radial flow impellers and 0.5D for axial flow impellers such as A310. However as D.sub.o approaches 2D the impeller provides gentle blending duty. This is acceptable for the production bioreactor application as K.sub.La study has shown that bioreactor K.sub.La is influenced mostly by the bottom A315 impeller and the top A310 impeller contributes to bulk mixing.

[0126] As a result of setting D.sub.c and D.sub.s at values D.sub.o is maintained at an optimal range for the duration of operation of the production bioreactor. During the course of a batch the liquid cover above the top impeller will change from 0.5D.sub.A310 and 1.08D.sub.A310. The liquid cover above the top impeller will increase as the bioreactor is fed nutrient feeds and alkali to maintain constant pH. Table 6 shows a range of liquid cover above the top impeller for a range of operating volumes.

TABLE-US-00006 TABLE 6 Key operating volumes and the liquid cover above top impeller, Do Cylinderical Do as height, H mm Do, mm ratio of Operating volume, L (inches) (inches) D.sub.A310 Pre-Harvest, 21739 L 3252 (128) 1324 (52) 1.08D.sub.A310 Pre-Harvest, 20000 L 2968 (117) 1040 (41) 0.85D.sub.A310 Post-Inoculation, 19231 L 2843 (112) 915 (36) 0.74D.sub.A310 Post-Inoculation, 17391 L 2543 (100) 615 (24) 0.5D.sub.A310 Pre-Inoculation, 15385 L 2215 (87) 287 (11) 0.23D.sub.A310 Pre-Inoculation, 13913 L 1973 (78) 45 (2) 0.04D.sub.A310 .sup.(1) Off bottom impeller clearance, Dc = 913 mm (0.75D.sub.A315), Impeller separation, Ds = 1498 mm (1.229D.sub.A315), tank ID of 2794 mm and height of ASME F&D base plate, H.sub.h = 483 mm .sup.(2) D.sub.o = H D.sub.s (D.sub.c H.sub.h)

Agitation RateRpm, P/V and Tip Speed

[0127] Table 7 below specifies the agitation rate for the 20 000 litre bioreactor. The bioreactor is agitated typically at 20-260 W/m.sup.3, preferably at 55-85 100 W/m.sup.3. The agitation strategy is being developed during the 500 litre pilot fermentations. The agitation rate of 0 to 801 rpm is therefore used as an operational range.

TABLE-US-00007 TABLE 7 Agitation rate for the 20000 L bioreactor Agitation rate, Power per unit volume, Tip Speed, rpm W .Math. m.sup.3 m/s Pre- Typically 28-30 can Typically 20 can 1.8-1.9 inoculation be higher be higher Post- Typically 56 can be Typically 103, can be 3.6 can be inoculation up to 80 up to 260 up to 5.1 until harvest

Mechanical Seals Specification

[0128] For bioreactor all seals are to be double mechanical seals with a maximum run out or wobble tolerance of 0.2 mm. Three types were considered; these include: [0129] Wet seal lubricated with sterile condensate. [0130] Dry seal lubricated with sterile gas such as N.sub.2 or CA. [0131] Non lubricated or floating seal that are uni-rotational.

[0132] All mechanical seals are recommended to be serviced on an annual basis. This requires the removal of seal from the bioreactor and sending the seal assembly to the vendor. Therefore the design must consider ease of routine maintenance.

[0133] The dry type seal (John Crane5280D type) will produce 3 g per year of shedding (seal face and seal seat material) composed of resin impregnated carbon. This is based on continuous 24 hour operation over a year. The amount of shedding for the wet seal is significantly less. Therefore a wet condensate-lubricated seal is adopted for all bioreactor double seals.

Bioreactor Aeration Requirement and Gassing Strategy

[0134] The aeration duty of the 20 000 litre bioreactor is governed by: [0135] K.sub.La requirement. [0136] DOT control strategy. [0137] pCO.sub.2 control/stripping strategy. [0138] Use of sintered or fluted spargers.

[0139] The 20 000 litre bioreactor is designed to provide K.sub.La values of up to 20 h.sup.1 for processes with oxygen uptake rates of 5 mmolL.sup.1h.sup.1. The bioreactor design needs to be flexible enough to allow cultivation of processes reaching 2010.sup.6 cellsmL.sup.1.

[0140] The aeration requirement can be achieved by a number of different approaches. However the use of a fluted sparger with air and oxygen enrichment to make up any deficit in oxygen transfer rate (OTR) during peak oxygen demand was used. The advantages of this approach are: [0141] Easier CIP and SIP validation of fluted sparge design. [0142] Larger air throughput to aid dissolved CO.sub.2 stripping. [0143] Reduced operating cost through the avoidance of purchase of single use sintered elements.

[0144] The disadvantages of the approach selected above also need to be considered. These include: [0145] Inherent lower K.sub.La for the low power number impellers selected.

[0146] Therefore the bioreactor aeration design must have the flexibility to be modified to meet the desired K.sub.La.

[0147] Table 8 describes the gassing requirements for the 20 000 litre bioreactor. The gas flow rates were scaled up on constant superficial gas velocity.

[0148] Two spargers are used. The main or DOT control sparger supplied by dual range clean air, mass flow controller (MFC) and oxygen MFC with gas flow metered via a DOT control loop and a CO.sub.2 MFC metering gas via the acid pH control loop. The dual range MFC's are used to achieve precise flow control at the extreme ends of the desired operating ranges.

[0149] The second or ballast sparger is supplied by a CA MFC to which nitrogen is also supplied. It was measured that early DOT control requires small nitrogen ballast to assist in early DOT demand and lower the DOT to set point. The ballast sparger also meters ballast air to facilitate stripping out excess pCO.sub.2.

[0150] The headspace purge is used to allow removal of CO.sub.2 and oxygen from the headspace. This is to facilitate better pH and pCO.sub.2 control and dilution of high oxygen blend prior to exhausting to environment. The ability to vary headspace flow rate allows design of gassing strategy for various processes requiring different blends of oxygen enrichment and control point pCO.sub.2.

TABLE-US-00008 TABLE 8 Gas flow rate and MFC operating ranges for the 20000 litre bioreactor Gas Operating range Comments Head Space.sup.1 1.) Clean air 1.) 0-1000 SLPM 1.) Head space purging of CO.sub.2 and O.sub.2 2.) Nitrogen 2.) Utility rated 2.) For rapid DOT probe zeroing 3.) Helium 3.) Utility rated 3.) Tank integrity testing DOT control Sparger 1.) Clean air.sup.2 1.) 10-500 SLPM 1.) Gas flow under DOT control 2.) Oxygen 2.) 10-100 SLPM 2.) Gas flow under DOT control Carbon dioxide.sup.3 3.) 2-150 SLPM 3.) Gas flow under pH control Ballast Sparger 1.) Clean air 1.) 20-500 SLPM 1.) Variable ballast for dCO.sub.2 stripping 2.) Nitrogen.sup.4 2.) 20-500 SLPM 2.) Early DOT control by variable flow .sup.1The air and nitrogen gas flow into headspace enters via a bypass for post SIP tank pressurisation. .sup.2Clean air gas flow operating range achieved by a dual CA MFC at 5-50SLPM and 50-500SLPM respectively. .sup.3CO.sub.2 gas flow operating range achieved by a dual CO.sub.2 MFC at 2-30SLPM and 30-150SLPM respectively. .sup.4Both air and nitrogen gas flow metered from a common CA MFC.

[0151] The bioreactor ports for sparger installation are designed to fit pipe design of diameter of 51 mm. The position of port should allow the placement of control sparger (D.sub.cS.sub.c) at a distance of 320 mm below the bottom edge of the lower impeller and no greater 593 mm from tank bottom (S.sub.c).

[0152] This results in a S.sub.c value of 593 mm or (0.65D.sub.c) and this falls outside the acceptable range of 0.2Dc to 0.6Dc. However hydrodynamic trials in 500 l suggest S.sub.c clearance of 0.41 to 0.71D.sub.o has no impact on measured K.sub.La.

[0153] A separate port for the installation of the ballast sparger was also built. The position of this port allows the placement of ballast sparger at a distance of 320 mm, (D.sub.cS.sub.c) below the bottom edge of the lower impeller and no greater then 593 mm from tank bottom (S.sub.c). The requirement to add ballast from a separate sparger is due to three reasons: [0154] Firstly, it prevents dilution of oxygen or oxygen enriched DOT demand gas with the ballast gas. This ensures the best OTR, as the oxygen concentration gradient of the bubbles emerging from the sparger is greatest. [0155] Secondly, it allows ballast sparger to be located at a different position from DOT control sparger to avoid impacting DOT control on delivering desired ballast for pCO.sub.2 control. [0156] Thirdly, the ballast sparger can be independently designed from the DOT control sparger.

[0157] The calculation of hole size and number of holes is iterated until the target Reynold's number, Re of gas emerging from holes is <2000 and the Sauter mean diameter for a bubble is 10-20 mm during chain bubble regime. Table 9 shows the key specifications for the control and ballast sparger for the 20 000 litre bioreactor.

TABLE-US-00009 TABLE 9 Design specification for the 20 000 litre bioreactor spargers DOT control Ballast Parameter sparger sparger Gas flow, SLPM 850 500 Number of sparge holes 250 100 Orifice diameter, d.sub.o, m 0.004 0.006 Gas flow, m.sup.3 .Math. s.sup.1 1.42E02 8.33E03 Orifice area, m.sup.2 1.26E05 2.83E05 Total orifice area, m.sup.2 3.14E03 2.83E03 Density of air, Kg .Math. m.sup.3 1.166 1.166 Viscosity, Nm .Math. s.sup.2 1.85E05 1.85E05 Sauter mean diameter, d.sub.vs, mm (d.sub.vs = 16.34 19.06 1.17 V.sub.o.sup.0.4d.sub.o.sup.0.8g.sup.0.2) Gravitational acceleration, g m .Math. s.sup.2 9.807 9.807 Density difference, Kg .Math. m.sup.3 1048.834 1048.834 Reynold's number, >2000 jetting regime 1139 1117 Gas velocity at sparger, V.sub.o, m/s 4.51 2.95 Sparger length, S.sub.L, m 1.077 1.077 Combined length to drill required holes, m 1.000 0.6 Number of rows to fit required holes in length S.sub.L 2 1 Sparger to tank bottom clearance, Sc, m 0.593 0.593 Sparger to bottom impeller clearance, Dc-Sc, m 0.320 0.320

[0158] A ring sparger of 0.8D.sub.bottom (80% diameter of bottom A-315 impeller diameter) is used to distribute the holes under the blades and not the impeller hub. However the CIP and installation of this configuration is difficult. Therefore selection of sparger geometry that permits distribution of the desired number of holes in a manner that is consistent with best to distribute the holes and sanitary design can be used also.

[0159] As an option a crescent rather then straight pipe is explored. The curvature of the crescent is 0.8D.sub.bottom. In order to aid installation and removal from side ports of the bioreactor the crescent circumference is 240 of the complete circumference of 0.8D.sub.bottom ring, this is 1077 mm.

[0160] The DOT control sparger is 1077 mm long and has a 51 mm diameter. The holes have a 4 mm diameter. A total of 250 holes divided into 2 rows (2125) at 45 from the dorsal (vertical) are used. Drain holes of 4 mm diameter on both ends of the sparger are drilled on the ventral side of the sparger to aid free CIP drainage of the sparger.

[0161] The ballast sparger is 1077 mm long and of 51 mm diameter and has a total of 100 6 mm diameter holes in a single dorsal row. Drain holes of 4 mm diameter on both ends of the sparger are drilled on the ventral side of the sparger to aid free CIP drainage of the sparger.

Position of Probes, Addition and Sample Ports

[0162] The probe ring position must be placed in a well-mixed representative region of the bioreactor. Additional considerations included working volume range and ergonomic operations. The location of probe ports, sample valve and addition points were considered together to avoid transitory spikes. Furthermore the position of the sample valve with respect to controlling probes needs to permit accurate estimation of off-line verification of the measured process parameter. This is shown in table 10.

TABLE-US-00010 TABLE 10 Probe, addition and sampling port specification for the 20000 litre bioreactor .sup.2Diameter, .sup.1Position, Probe/Port Location mm (inches) mm (inches) Rational Temperature Lower 38.1 (1.5) 1.) 913 (36) In the plane of (main) ring 2.) 30 centre-line of bottom impeller Temperature Lower 38.1 (1.5) 1.) 913 (36) In the plane of (backup) ring 2.) 170 centre-line of bottom impeller pH Lower 38.1 (1.5) 1.) 913 (36) In the plane of (main) ring 2.) 10 centre-line of bottom impeller pH Lower 38.1 (1.5) 1.) 913 (36 In the plane of (backup) ring 2.) 20 centre-line of bottom impeller DOT Lower 25.0 (0.98) 1.) 913 (36) In the plane of (main) ring 2.) 150 centre-line of bottom impeller DOT Lower 25.0 (0.98) 1.) 913 (36) In the plane of (backup) ring 2.) 160 centre-line of bottom impeller pCO.sub.2 Lower 50.8 (2)tbd 1.) 913 (36) In the plane of (spare) ring 2.) 20 centre-line of bottom impeller Biomass Lower 50.8 (2) 1.) 913 (36) In the plane of (spare) ring 2.) 160 centre-line of bottom impeller Spare probe port Lower 25.0 (0.98) 1.) 913 (36) In the plane of (DOT-type) ring 2.) 150 centre-line of bottom impeller Spare probe port Lower 38.1 (1.5) 1.) 913 (36) In the plane of (pH-type) ring 2.) 10 centre-line of bottom impeller Sample valve Lower 12.7 (0.5) 1.) 913 (36) NovAseptic (main) ring 2.) 40 type Sample valve Lower 12.7 (0.5) 1.) 913 (36) NovAseptic (backup) ring 2.) 50 type Alkali addition 1 - Lower 50.8 (2) 1.) 913 (36) Diametrically Tank 1 ring 2.) 190 opposite pH probes Alkali addition 2 - Centre-line 50.8 (2) 1.) 2411 (95) Diametrically Tank 1 of upper 2.) 190 opposite pH impeller probes Continuous feed 1 - Lower 50.8 (2) 1.) 913 (36) Diametrically Tank 2 ring 2.) 200 opposite pH probes Continuous feed 2 - Lower 50.8 (2) 1.) 913 (36) Diametrically Tank 3 ring 2.) 210 opposite pH probes DOT control sparger N/A 101.6 (4) 1.) 593 (23) Diametrically orifice 2.) 0 opposite ballast sparger Ballast sparger N/A 101.6 (4) 1.) 593 (23) Diametrically orifice 2.) 180 opposite control sparger Overlay gas Head 101.6 (4) 1.) N/A Diametrically plate 2.) 135 opposite vent out Exhaust Impeller 50.8 (2) 1.) N/A Diametrically vent out flange 2.) 315 opposite over- plate lay gas in Harvest Base 76.2 (3.0) 1.) N/A NovAseptic valve plate 2.) Centre type to allow free draining Antifoam Head 50.8 (2) 1.) N/A Liquid addition plate 2.) 170 surface/0.25T from tank centre Shot feed 1 - Head 50.8 (2) 1.) N/A Liquid surface LS1 plate 2.) 190 Shot feed 2 - Glu- Head 50.8 (2) 1.) N/A Liquid surface cose shot plate 2.) 180 Small add - Head 50.8 (2) 1.) N/A Liquid surface - Spare plate 2.) 200 directed into vessel wall Media Head 101.6 (4) 1.) N/A Nozzle directed inlet plate 2.) 310 onto vessel wall Inoculum transfer Head 101.6 (4) 3.) N/A Nozzle directed from 4000 L to plate 4.) 320 onto vessel wall 20000 L CIP - Impeller 76.2 (3) 1.) N/A CIP'ing of Spray ball flange 2.) 270 highest point plate CIP - Head 76.2 (3) 5.) N/A As per CIP Spray ball plate 6.) 60 design CIP - Head 76.2 (3) 1.) N/A As per CIP Spray ball plate 2.) 180 design CIP - Head 76.2 (3) 1.) N/A As per CIP Spray ball plate 2.) 300 design Pressure indicating Head 38.1 (1.5) 1.) N/A As per vessel transmitter (PIT) plate 2.) 60 vendor design Pressure Head 38.1 (1.5) 1.) N/A As per vessel gauge plate 2.) 50 vendor design Rupture Head 101.6 (4) 1.) N/A As per vessel disc plate 2.) 280 vendor design Spare Head 101.6 (4) 1.) N/A As per vessel nozzle plate 2.) 160 vendor design Sight Head 101.6 (4) 1.) N/A As per vessel glass plate 2.) 70 vendor design Light Head 76.2 (3) 1.) N/A As per vessel glass plate 2.) 75 vendor design Manway Head 457.2 (18) 1.) N/A Personnel plate 2.) 90 entry Agitator head/ Head 1320.8 (52) N/A Entry/removal flange plate Impeller shaft Agitator 304.8 (12) N/A Entry/removal and seal head/ manway flange .sup.1Measured from the tangential line of the base plate. Degrees pertain to plane of clockwise rotation. .sup.2Diameter of nozzle at bioreactor.

Addition Ports, Surface and Sub-Surface

[0163] The need to determine addition ports that terminate at liquid surface and those that are subsurface was determined by operational scenarios and the effects of feed strategy on process control.

[0164] Currently the protein free process has two continuous feeds that need to be discharged in well-mixed area of the bioreactor. Additional provision for glucose and an LS1-type shot addition is also integrated in the well mixed region. The foam is controlled by surface addition of 1 in 10 diluted C-emulsion. The inoculation of seed into the pre-inoculation bioreactor is served by avoiding build up of foam which will arise as the culture is dropped onto the surface of the medium. Following ports were designed: [0165] Six surface additions with media inlet, inoculum inlet, one small addition inlet directed into the wall of the vessel while the others dropped onto the liquid surface away from the tank wall. [0166] Four subsurface additions comprised of inlets from the two feed tanks and bi-level inlet from the alkali tank.

Sample Ports

[0167] The sample port design allows a representative sample to be taken from the bioreactor. Therefore any residual material must be as small as possible. The samples taken are used to determine off line checks for dissolved gases, pH, nutrients and biomass concentration. The orifice of the port opening is large enough to prevent sieving causing biomass aggregates to be retained. The 2 mm orifice NovaSeptum sampling device was used. However this has to be balanced with the desire to keep residual volume of the port low. The port needs to be positioned in a well-mixed zone adjacent to the probes that need to be verified by off-line checks and will be determined via nozzle position (see table 10).

Add Tanks

[0168] In order to reduce cost and time the add-tanks supplying the bioreactor are of modular design. The production bioreactor has three litre nominal volumes add tanks. The add tanks are filled at 25 I/min. The flow rate of feeds from the add tanks to the bioreactor is controlled at 0.2 to 1.0 millilitres of feed per litre of post-inoculation bioreactor volume per hour (ml/l/h). It is expected that feed rate is controlled at 5% of set point.

[0169] The production bioreactor is serviced by three 1372 mm ID by 1880 mm add tanks. These tanks have the capability to be cleaned and sterilised independently and together with the production bioreactor.

Manway

[0170] Access into the bioreactor is required for certain service operations. Access can be gained by considering a flanged head plate or incorporation of a manway into the head plate. The need for access into the bioreactor is for: [0171] Installation of impellers. [0172] Installation and replacing of impeller and impeller shaft. [0173] Installation and replacing of mechanical seal. [0174] Service of vessel furniture. [0175] Potential modification of sparger position to obtain desired hydrodynamic characteristics.

[0176] The size of the manway must be sufficient to allow access for the above objectives. The manway used was of sufficient diameter to allow the removal of two impellers of 1219 mm diameter.

Volume Measurement

[0177] The design ensures that any sensor gives sufficient precision in volume measurement around the operating range.

[0178] The volume measurement in bioreactor is able to measure a range of 13 000 to 25 000 litres. The sensor sensitivity needs to be at least 0.5% of full span.

[0179] Volume measurement in the feed add-tanks and alkali tank is able to measure 0 to 2200 and 2500 litres respectively. The sensor sensitivity needs to be at least 0.2% of full span. This will permit hourly verification of feed flow rate at the minimum flow rate of 0.2 ml/l per hour or 3.5 l per hour by measuring the volume decrease in the add tanks.

Bioreactor Temperature Control

[0180] The medium is brought to operating temperature and pH by process control. This is achieved by gentle heating of the jacket (avoid high temperature at vessel wall). The temperature control range during operation is 36 to 38 C. with an accuracy of 0.2 C. at set point.

Jacket

[0181] The bioreactor jacket area is specified with the following considerations in mind: [0182] Steam sterilisation at 121-125 C. [0183] Warming up of medium from 10 C. to 36.5 C. in <2 h. [0184] All points within the bioreactor must reach 0.2 C. of set point, typically 36.5 C., as measured by thermocouples. [0185] Chilling of medium from 36.5 C. to 10 C. in <2 h.

Bioreactor pH Control

[0186] The process pH is monitored and controlled with probes connected via a transmitter to a DCS based process controller. The process is be controlled by addition of CO.sub.2 to bring the pH down to set point and addition of alkali to bring pH up to set point. pH is controlled at 0.03 of set point.

[0187] Alkali is added through two addition points to distribute the alkali. This ensures quicker blending of alkali in the event of long recirculation time in the tank. The CO.sub.2 is added via the control sparger.

[0188] Control and back-up probes are located in the lower port ring at 913 mm (see table 10) from tank bottom. Additionally the pH probes are located diametrically opposite the alkali addition points into the bioreactor.

Bioreactor DOT Control

[0189] Dissolved oxygen is monitored and controlled with polarographic DOT probe. The DOT set point maintained by sparging: [0190] Initial N.sub.2 ballast and/or air on demand. [0191] Air ballast with air on demand. [0192] Air ballast with oxygen on demand. [0193] Reversing gas usage once oxygen demand decreases.

[0194] Cascade DOT control allows DOT set point to be maintained through changes in the ballast and demand gas in conjunction with ramping of agitator speed.

[0195] In order to control pCO.sub.2 the ballast required to strip out excess dCO.sub.2 impacts DOT control. Therefore the DOT control is considered together with pCO.sub.2 control for those processes where metabolic CO.sub.2 is liberated. DOT is controlled at 2% of set point. Control and back-up probes are located in the lower port ring at 913 mm from tank bottom.

Bioreactor Dissolved CO.SUB.2 .Control

[0196] The process dCO.sub.2 is monitored with an pCO.sub.2 probe and excess dCO.sub.2 is stripped by gassing CA through the ballast sparger. The optimal position for this probe is close to the pH probes.

Feed Addition Control

[0197] The feeds (SF22 and amino acid) are high in pH and osmolality. Therefore bolus additions need to be avoided to maintain good pH control. However the control of desired flow rates (5% of set point) is technically challenging. Therefore an addition strategy that encompasses point of addition with delivery mode avoids the circulation of feed bolus and potential variations of pH control.

[0198] Therefore the point of addition is in the plane of the centre-line of the bottom impeller that is 913 mm from tank bottom to assist in the rapid blending of feed bolus.

Antifoam Addition Control

[0199] Antifoam (C-emulsion) addition is added as required to maintain the bioreactor liquid surface free of foam. A working stock of 1 in 10 diluted C-emulsion can be dosed on the liquid surface. The antifoam suspension is continuously agitated in the storage container to prevent partitioning. It is important to dose the antifoam close to the centre of the tank to diminish the effects of the radial component of the fluid flow carrying the antifoam to the tank walls where it will adhere. Therefore the addition point is 0.25T toward the tank centre or 699 mm from tank centre.

EXAMPLE 2: 4000 LITRE BIOREACTOR

Vessel Geometry

[0200] The vessel geometry for the 4000 litre bioreactor was determined by an iterative design basis in which the maximum working volume, freeboard straight side distance aspect ratio (H.sub.L/T) and impeller to tank diameter ratio (D/T) are altered until an acceptable aspect ratio is achieved.

Bioreactor Aspect Ratio H.SUB.L./T

[0201] Table 11 describes the aspect ratios in the 4000 litre bioreactor at the various operating volumes during normal processing. These aspect ratios arise from the selection of tank ID and the operating volume required. From a processing perspective the mixing requirements at the three operating conditions are different. During pre-inoculation stage the bioreactor mixing is important to allow medium to equilibrate with minimal K.sub.La requirement. However for post-inoculation and pre-transfer stages both mixing and K.sub.La are important considerations. Therefore both these features were tested at the aspect ratio range.

TABLE-US-00011 TABLE 11 Key operating volumes and aspect ratios in the 4000 litre bioreactor Volume, Liquid head, Aspect ratio, L mm H.sub.L/T Pre-Inoculation 1.) 1914 1.) 1031 1.) 0.63 2.) 2782 2.) 1448 2.) 0.89 3.) 3077 3.) 1590 3.) 0.98 Post Inoculation & 1.) 2153 1.) 1146 1.) 0.70 Pre-transfer 2.) 3478 2.) 1783 2.) 1.10 3.) 3846 3.) 1960 3.) 1.21

Tank Diameter

[0202] The tank diameter was altered to obtain the optimal aspect ratio H.sub.L/T. Changes to tank internal diameter are limited by acceptable aspect ratio and plant footprint. The tank ID is 1626 mm.

Tank Height

[0203] Tank height is determined from the maximum operating volume, aspect ratio H.sub.L/T, freeboard straight side length, base and top plate design. The final tank height is a compromise value determined from volumetric contingency for foam, plant height and acceptable impeller shaft length. The head to base tan line height is 2817 mm.

Freeboard Height

[0204] The freeboard height of 500 mm (1039 litre or 27% v/v of the maximum operating volume) is used for this seed bioreactor.

Head and Base Plate

[0205] The base and head plate design is a ASME F&D design for this seed bioreactor.

Bioreactor Agitation Requirement

[0206] The agitation of the bioreactor is to achieve rapid mixing, maintain homogeneity, maintain mammalian cells in suspension and gas bubble dispersion. The underlying issue for achieving the above objectives is minimising cell damage through shear forces originating from impeller geometry and eddies or vortices created behind the impeller blades. A compromise of the above objectives was achieved by selection of an appropriate impeller type.

Bottom Versus Top Driven Impeller Shaft

[0207] The motor drive is top mounted for the benefits already highlighted.

Baffles

[0208] The baffle requirement for centre mounted impeller is critical to prevent vortex formation. The critical issues related to baffles are baffle number, baffle width (W), baffle length (H.sub.baffle) and baffle to tank wall clearance (W.sub.c).

[0209] Four equally spaced baffles that are 0.1T or 163 mm wide 1.1HH.sub.h or 2195 mm tall and have a baffle to tank wall clearance, W.sub.c of 0.01T or 16 mm were used.

[0210] The thickness of the baffles is not specified but the thickness needs to ensure rigidity to the radial component of the fluid flow. Additionally thickness needs to ensure the baffle plates are not warped during SIP thereby affecting the baffle to tank wall clearance.

Impeller Type, Size and Number

[0211] The impellers for this bioreactor are identically formed to the 20000 litre vessel and have a identical D/T ratio of 0.44. The bottom impeller is a Lightnin's A315 at 710 mm of diameter and the top impeller is a Lightnin's A310 at 710 mm of diameter.

The impeller spacing, D.sub.c, D.sub.s and D.sub.n

[0212] The impeller spacing, D.sub.s, between the centre-line of the top impeller and the centre-line of the lower impeller is 1.229D.sub.bottom or 872 mm. The off bottom impeller clearance, D.sub.c is 0.75D.sub.bottom or 531 mm. This allows the lower impeller to remain submerged at the lowest post-inoculation volume of 2153 litres and both impellers submerged at 3367 litres with liquid head above the upper impeller (D.sub.o) of 0.5D.sub.top or 358 mm.

[0213] Table 12 highlights the volumes that will form liquid surfaces or lower liquid cover above the impeller. Agitation can be modified to avoid foaming at these critical volumes.

TABLE-US-00012 TABLE 12 Key operating volumes that cause interaction with impellers and liquid surface Interaction Volume, L Potential Operation Submerge top impeller 3433 Volume seen during inoculation of 1 with 0.5D.sub.A310 liquid in 5 processes cover Liquid surface touching 2758 Volumes seen during pre-inoculation top edge of top impeller fill of 1 in 5 processes. Liquid surface touching 2621 Volumes seen during pre-inoculation bottom edge of top impeller fill of 1 in 5 processes. Submerge bottom impeller with 1654 Volumes seen during pre-inoculation 0.5D.sub.A315 liquid cover fill of 1 in 5 processes. Liquid surface touching 1104 Volumes seen during pre-inoculation top edge of bottom impeller fill of 1 in 5 and 1 in 9 processes. Liquid surface touching 650 Volumes seen during pre-inoculation bottom edge of bottom fill of 1 in 5 and 1 in 9 processes. impeller .sup.(1) Minimum operating volume with lower impeller submerged is 1654 litres and minimum operating volume with both impellers submerged is 3433 litres .sup.(2) The operating volume range is 1914 to 3846 litres.

[0214] The 4000 l bioreactor can operate at two discrete post-inoculation volumes with either the lower impeller submerged (during cultivation of 1 in 9 seeding process) or with both impellers submerged (during 1 in 5 seeding process), table 13 shows the liquid cover obtained for the upper and lower impeller during its operation.

[0215] A liquid cover of 0.67 to 0.82D.sub.bottom above the lower A315 impeller is observed during cultivation of the 1 in 9 seeded processes. This is within the recommendations of 0.5 to 1D.

[0216] A liquid cover of 0.06 to 0.78D.sub.top above the top A310 impeller is observed during cultivation of 1 in 5 seeded process. The lower liquid cover is outside the recommendation. However this liquid cover is observed during pre-inoculation when mixing and agitation are less critical.

TABLE-US-00013 TABLE 13 Key operating volumes and the liquid cover above top impeller, Do and bottom impeller, D.sub.Bo Cylinderical Do, D.sub.Bo, Do as ratio D.sub.Bo as ratio Operating volume, L height, H (mm) mm mm of D.sub.A310 of D.sub.A315 Post - Inoculation and 863 or 614 or 0.82D.sub.A315 Pre-Transfer, 2153 L 34 24 Post - Inoculation and 1501 or 380 or 0.53D.sub.A310 Pre-Transfer, 3478 L 59 15 Post - Inoculation and 1678 or 557 or 0.78D.sub.A310 Pre-Transfer, 3846 L 66 22 Pre-Inoculation, 1914 L 748 or 499 or 0.67D.sub.A315 29 20 Pre-Inoculation, 2782 L 1166 or 45 or 0.06D.sub.A310 46 2 Pre-Inoculation, 3077 L 1308 or 187 or 0.26D.sub.A310 52 7 .sup.(1) Off bottom impeller clearance, Dc = 531 mm (0.75D.sub.A315), Impeller separation, Ds = 872 mm (1.229D.sub.A315), tank ID of 1626 mm and Height of ASME F&D base plate, H.sub.h = 282 mm .sup.(2) Do = H Ds (Dc H.sub.h) and D.sub.Bo = H (Dc H.sub.h)
Agitation Raterpm, P/V and Tip Speed

[0217] Table 14 specifies the agitation rate for the 4000 litre bioreactor. The bioreactor will be agitated typically at 20-260 W/m.sup.3, preferably at 55-85 W/m.sup.3. The agitation strategy was developed during the 500 litre pilot fermentations. An agitation rate of 0 to 881 rpm is therefore used as an operational range.

TABLE-US-00014 TABLE 14 Agitation rate for the 4000 L bioreactor Agitation rate, Power per unit volume, Tip Speed, rpm W/m3 m/s .sup.10-88 0-150 0.0-3.3 .sup.20-86 0-150 0.0-3.2 .sup.1When both impellers submerged .sup.2When bottom impeller submerged

Mechanical Seals Specification

[0218] A double mechanical seal that is condensate lubricated is used as described.

Bioreactor Aeration Requirement

[0219] Table 15 shows the gas flow, based upon scale up of constant superficial gas velocity, for DOT and pH control during the inoculum expansion in the 4000 litre bioreactor. Oxygen is not required for DOT control. However oxygen enriched air can be used to facilitate lower gassing to prevent excess foaming. It is recommended that a smaller range N.sub.2 MFC should supply nitrogen for early DOT control and reducing deviant, high levels of DOT.

TABLE-US-00015 TABLE 15 Gas flow rate and MFC operating ranges for the 4000 litre bioreactor Gas Operating range Comments Head Space.sup.1 1. Clean air 1. 0-200SLPM 1. Head space purging of CO.sub.2 and O.sub.2 2. Nitrogen 2. Utility rated 2. For rapid DOT probe zeroing 3. Helium 3. Utility rated 3. Tank integrity testing Control Sparger 1. Clean air.sup.1 1. 10-60SLPM 1. Gas flow under DOT control 2. Oxygen 2. 1.0-10SLPM 2. Gas flow under DOT control 3. Carbon dioxide 3. 1.0-20SLPM 3. Gas flow under pH control 4. Nitrogen.sup.2 4. 2.0-15SLPM 4. Early DOT control by ballast 5. Helium 5. Utility rated 5. Tank integrity testing .sup.1The air and nitrogen gas flow into bioreactor via a bypass for post SIP tank pressurisation. .sup.2Nitrogen delivered via the 2 to 15SLPM N.sub.2 MFC and could be used during early DOT control

[0220] The calculation of hole size and number of holes, for the fluted sparger, is iterated until the target Reynolds number of gas emerging from holes is <2000 and the Sauter mean bubble diameter for a bubble chain regime is approximately 10 mm.

[0221] Table 16 show the key sparger design specification for the 4000 litre bioreactor. The sparger length, S.sub.L of 568 mm is determined for pipe geometry. The holes are distributed on either end of the sparger to prevent bubble liberating directly under the A315 hub. Alternatively a crescent geometry can be used. The pipe diameter is selected to aid spacing of the desired number of holes. The diameter is 38 mm. The mm holes are located on the dorsal surface of the sparger with a single 2 mm hole located on the ventral surface to aid free CIP drainage of the sparger.

[0222] The bioreactor port for sparger installation is designed to a fit pipe design of diameter of 38 mm. The position of the port allows the placement of a control sparger at a distance of 194 mm, D.sub.cS.sub.c below the bottom edge of the lower impeller and no greater 337 mm from tank bottom, S.sub.c.

TABLE-US-00016 TABLE 16 Design specification for the 4000 litre bioreactor sparger Control Parameter Sparger Gas flow, SLPM 105 Number of sparge holes 100 Orifice diameter, d.sub.o, m 0.002 Gas flow, m.sup.3 .Math. s.sup.1 1.75E03 Orifice area, m.sup.2 3.14E06 Total orifice area, m.sup.2 3.14E04 Density of air, Kg .Math. m.sup.3 1.166 Viscosity, Nm .Math. s.sup.2 1.85E05 Sauter mean diameter, mm (d.sub.vs = 1.17 V.sub.o.sup.0.4 d.sub.o.sup.0.8 g.sup.0.2) 10.21 Gravitional acceleration, g, m .Math. s.sup.2 9.807 Density difference, Kg .Math. m.sup.3 1048.834 Reynold's number, >2000 jetting regime 704 Gas velocity at sparger, V.sub.o, m/s 5.57 Sparger length, S.sub.L, m 0.568 Combined length to drill required holes, m 0.2 Number of rows to fit required holes in length S.sub.L 1 Sparger to tank bottom clearance, Sc, m 0.337 (13) Sparger to bottom impeller clearance, Dc-Sc, m 0.194 (8)

Position of Probes, Addition and Sample Ports

[0223] The design basis for positioning of probes, addition and sample ports has been covered in example 1 and are listed in table 17:

TABLE-US-00017 TABLE 17 Probe, addition and sampling port specification for the 4000 litre bioreactor .sup.2Diameter, .sup.1Position, Probe/Port Location mm (inches) mm (inches) Rational Temperature Lower 38.1 (1.5) 1.) 531 (21) In the plane of (main) ring 2.) 30 centre-line of bottom impeller Temperature Lower 38.1 (1.5) 1.) 531 (21) In the plane of (backup) ring 2.) 170 centre-line of bottom impeller pH Lower 38.1 (1.5) 1.) 531 (21) In the plane of (main) ring 38.1 (1.5) 2.) 10 centre-line of bottom impeller pH Lower 1.) 531 (21) In the plane of (backup) ring 2.) 20 centre-line of bottom impeller DOT Lower 25.0 (0.98) 1.) 531 (21) In the plane of (main) ring 2.) 150 centre-line of bottom impeller DOT Lower 25.0 (0.98) 1.) 531 (21) In the plane of (backup) ring 2.) 160 centre-line of bottom impeller Spare-1 Lower 25.0 (0.98) 1.) 531 (21) In the plane of (nutrient) ring 2.) 170 centre-line of bottom impeller Spare-2 Lower 38.1 (1.5) 1.) 531 (21) In the plane of (pCO.sub.2) ring 2.) 180 centre-line of bottom impeller Spare-3 Lower 50.8 (2) 1.) 531 (21) In the plane of (biomass) ring 2.) 190 centre-line of bottom impeller Sample valve Lower 12.7 (0.5) 1.) 531 (21) NovAseptic type (main) ring 2.) 40 Alkali Lower 50.8 (2) 1.) 531 (21) Diametrically addition ring 2.) 190 opposite pH probes Feed 1 Lower 50.8 (2) 1.) 531 (21) Diametrically ring 2.) 200 opposite pH probes Feed 2 Lower 50.8 (2) 1.) 531 (21) Diametrically ring 2.) 210 opposite pH probes Antifoam Head 50.8 (2) 1.) N/A Liquid surface/ addition plate 2.) 170 0.25T from tank centre Spare surface Head 50.8 (2) 3.) N/A Liquid surface addition plate 4.) 180 directed to vessel wall DOT control N/A 50.8 (2) 337 (13) sparger orifice 1.) 0 Overlay Head 101.6 (4) 1.) N/A Diametrically gas plate 2.) 135 opposite vent out Exhaust Head 50.8 (2) 1.) N/A Diametrically vent out plate 2.) 315 opposite overlay gas in Transfer Base 76.2 (3.0) 1.) N/A NovAseptic type valve plate 2.) Centre to allow free draining Inoculum transfer Head 101.6 (4) 1.) N/A Directed into from 1000 L to 4000 L plate 2.) 320 vessel wall Media Head 101.6 (4) 3.) N/A Directed into inlet plate 4.) 310 vessel wall CIP - Impeller 76.2 (3) 1.) N/A CIP'ing of Spray ball flange plate 2.) 270 highest point CIP - Head 76.2 (3) 1.) N/A Spray ball plate 2.) 60 CIP - Head 76.2 (3) 1.) N/A Spray ball plate 2.) 180 CIP - Head 76.2 (3) 1.) N/A Spray ball plate 2.) 300 Pressure indicating Head 38.1 (1.5) 1.) N/A transmitter (PIT) plate 2.) 60 Pressure Head 38.1 (1.5) 1.) N/A gauge plate 2.) 50 Rupture Head 101.6 (4) 1.) N/A disc plate 2.) 280 Spare Head 101.6 (4) 1.) N/A nozzle plate 2.) 160 Sight Head 101.6 (4) 1.) N/A glass plate 2.) 70 Light Head 76.2 (3) 1.) N/A glass plate 2.) 75 Agitator Head 813 (32) N/A Entry/removal head/flange plate Impeller shaft and Agitator 152 (6) N/A Entry/removal seal manway head/flange .sup.1Measured from the tangential line of the base plate. Degrees pertain to plane of clockwise rotation. .sup.2Diameter of nozzle at bioreactor.

Addition Ports, Surface and Sub-Surface

[0224] The need to categorise additions ports that terminate at liquid surface and those that are subsurface is determined by the operational scenarios and effects of feed strategy on process control.

[0225] The 4000 litre bioreactor has been designed to accept two subsurface feeds and alkali that need to be discharged in well-mixed area of the bioreactor. The foam is controlled by surface addition of 1 in diluted C-emulsion. A single spare above surface addition port directed to the vessel wall is also designed for future flexibility. The splashing of culture onto the surface of the medium during inoculation of the seed bioreactor can be avoided to prevent build up of foam. Therefore the inoculum addition port is above surface and directed to the vessel wall. The use of the harvest port in the base plate is the ideal port for removal of inoculum during transfer of inoculum. Additionally the medium addition port is directed to the vessel wall. In summary the total addition ports are: [0226] Four surface additions with medium inlet, inoculum inlet and a spare small addition directed to the vessel wall and addition port for antifoam dropped on to the liquid surface away from the vessel wall. [0227] Three subsurface additions for feeds and alkali.

Sample Ports

[0228] The sample port design is similar to that specified for the 20 000 litre bioreactor.

Volume Measurement

[0229] The level sensor is able to measure up to 4000 litres with an accuracy 0.5% of full span.

Bioreactor Temperature Control

[0230] The 1914 to 3077 litre of medium are brought to operating temperature, typically 36.5 C. by process control. This is achieved by gentle heating of the jacket and avoid high temperature at vessel wall.

Jacket

[0231] The bioreactor jacket area is specified with the following considerations in mind: [0232] Steam sterilisation at 121-125 C. [0233] Warming up of 1914-3077 litres of medium from 10 C. to 36.5 C. in <2 h. [0234] All points within the bioreactor must reach 0.2 C. of set point, typically 36.5 C. as measured by thermocouples. [0235] Chilling of 1914-3077 litres of medium from 362 C. to 10 C. in h.

Bioreactor pH Control

[0236] The process pH is monitored and controlled with probes connected via a transmitter to a DCS based process controller. The process pH is controlled by addition of CO.sub.2 through the control sparger to bring the pH down to set point and addition of alkali to bring pH up to set point.

[0237] Alkali is added through at least one subsurface port at centre-line of the bottom impeller. The CO.sub.2 will be added via the control sparger.

[0238] Control and backup probes are in the lower port ring at 531 mm from tank bottom as shown in table 17.

Bioreactor DOT Control

[0239] Dissolved oxygen is monitored and controlled with polarographic DOT probe. The DOT set point maintained by sparging: [0240] Initial N.sub.2 ballast and/or air on demand. [0241] Air ballast with air on demand. [0242] Air ballast with oxygen on demand.

[0243] DOT control allows DOT set point to be maintained through interchangeable use of oxygen or air as demand gas. It is not envisaged that pCO.sub.2 control is required in the inoculum bioreactor. Control and backup probes are in the lower port ring at 531 mm from tank bottom as shown in table 17.

Feed Addition Control

[0244] The point of addition is 531 mm from tank bottom, in the plane of the centre-line of the lower impeller to assist in the rapid dissipation of feed bolus.

Antifoam Addition Control

[0245] The addition point is at surface projecting 0.25T toward the tank centre or 407 mm from centre of tank.

EXAMPLE 3: 1000 LITRE BIOREACTOR SPECIFICATION

Vessel Geometry

[0246] The vessel geometry for the 1000 litre bioreactor was determined by an iterative design basis in which the maximum working volume, freeboard straight side distance, aspect ratio (H.sub.L/T) and impeller to tank diameter ratio (D/T) are altered until an acceptable aspect ratio is achieved.

Bioreactor Aspect Ratio H.SUB.L./T

[0247] Table 18 below describes the aspect ratios in the 1000 litre bioreactor at various operating volumes during normal processing. These aspect ratios arise from the selection of tank ID and the operating volume required. From a processing perspective the mixing requirements at the different operating conditions are different. During pre-inoculation stage the bioreactor mixing is important to allow medium to equilibrate with minimal K.sub.La requirement. However with post-inoculation and pre-transfer stages both mixing and K.sub.La are important considerations. Therefore both of these features were tested at the aspect ratio range.

TABLE-US-00018 TABLE 18 Key operating volumes and aspect ratios in the 1000 litre bioreactor Volume, Liquid head, Aspect ratio, L mm H.sub.L/T Stage N-3 Pre-Inoculation 250 484 0.56 Stage N-3 Post Inoculation & 300 570 0.66 Pre-transfer/Harvest Stage N-2 Pre-Inoculation, Post- 1.) 400.sup.1 1.) 740 1.) 0.86 drain Pre-refill 2.) 50-100.sup.1 2.) 143-228 2.) 0.17-0.26 3.) 192.sup.2 3.) 385 3.) 0.45 Stage N-2 Post Inoculation & 1.) 450 1.) 826 1.) 0.96 Pre-transfer/Harvest 2.) 450-900.sup.3 2.) 826-1594 2.) 0.96-1.84 3.) 960.sup.4 3.) 1696 3.) 1.96 .sup.1Pre-inoculation volume and rolling seed inoculation volume for the 1 in 9 sub-cultivation process. .sup.2Rolling seed inoculation volume for the 1in 5 sub-cultivation processes. .sup.3Rolling seed post inoculation & pre-transfer volume for the 1in 9 sub-cultivation processes. .sup.4Rolling seed post inoculation & pre-transfer volume for the 1in 5 sub-cultivation processes.

Tank Diameter

[0248] The tank diameter is altered to obtain the optimal aspect ratio H.sub.L/T. Changes to tank internal diameter are limited by acceptable aspect ratio and plant footprint. The tank ID is 0.864 m.

Tank Height

[0249] The tank height is determined from the maximum operating volume, aspect ratio H.sub.L/T, freeboard straight side length, base and top plate design. The final tank height is a compromise value determined from volumetric contingency for foam, plant height and acceptable impeller shaft length. The head to base tangent line height is 2.347 m.

Freeboard Height

[0250] The freeboard height of 500 mm (293 litres or 31% v/v of the maximum operating volume) is used for this seed bioreactor.

Head and Base Plate

[0251] The base and head plate design is ASME F&D for this seed bioreactor.

Bioreactor Agitation Requirement

[0252] Agitation of the bioreactor is to achieve rapid mixing, maintain homogeneity, maintain mammalian cells in suspension and gas bubble dispersion. The underlying issue with achieving the above objectives is to minimise cell damage through shear forces originating from impeller geometry and eddies or vortices created behind the impeller blades. A compromise of the above objectives was achieved by selection of an appropriate impeller type and gassing strategy.

Bottom Versus Top Driven Shaft

[0253] The motor drive is top mounted for the benefits as already highlighted.

Baffles

[0254] The baffle requirement for a centre mounted impeller is critical to prevent vortex formation. The critical issues related to baffles are baffle number, baffle width (W), baffle length (H.sub.baffle) and baffle to tank wall clearance (W.sub.c).

[0255] Four equally spaced baffles that are 0.1T or 86 mm wide 1.1HH.sub.h or 2099 mm tall and have a baffle to tank wall clearance, W.sub.c of 0.01T or 9 mm were used.

[0256] The thickness of baffle is not specified but the thickness needs to ensure rigidity to the radial component of the fluid flow. Additionally thickness needs to ensure the baffle plates are not warped during SIP thereby affecting the baffle to tank wall clearance.

Impeller Type, Size and Number

[0257] The impellers for the 1000 l bioreactor should be identical formed to the 20 000 litre vessel with an identical D/T ratio. Therefore the bottom impeller is a Lightnin's A315 at 381 mm diameter and the top impeller is a Lightnin's A310 at 381 mm diameter.

The impeller spacing, D.sub.c, D.sub.s and D.sub.o

[0258] The impeller spacing (D.sub.s) between the centre-line of the top impeller and the centre-line of the bottom impeller is 2D.sub.bottom (762 mm). The off bottom impeller clearance (D.sub.o) is 0.4D.sub.bottom(152 mm). This allows the bottom impeller to remain submerged with liquid cover (Do) of 0.5D.sub.bottom or 190 mm at the lowest post-inoculation volume of litres and both impeller submerged at 616 litres with liquid head above the upper impeller, D.sub.o, of 0.5D.sub.top (190 mm).

[0259] Table 19 highlights volumes that will form liquid surfaces or lower liquid cover above the impeller Agitation can be modified to avoid foaming at these critical volumes.

TABLE-US-00019 TABLE 19 Key operating volumes that cause interaction with impellers and liquid surface Interaction Volume, L Potential Operation Submerge top impeller with 616 Volume seen during inoculation of 1 in 0.5D.sub.A310 liquid cover 5 processes and rolling operation of the 1 in 9 process Liquid surface touching top 512 Volume seen during inoculation of 1 in edge of top impeller 5 processes and rolling operation of the 1 in 9 process Liquid surface touching bottom 492 Volume seen during inoculation of 1 in edge of top impeller 5 processes and rolling operation of the 1 in 9 process Submerge bottom impeller 167 Volume seen during inoculation of 1 in with 0.5D.sub.A315 liquid cover 5 processes and rolling operation of the 1 in 9 process Liquid surface touching top 90 Volume seen during rolling operation of the 1 in edge of lower impeller 9 process Liquid surface touching 21 Volumes seen during pre-inoculation fill of 1 in 5 bottom edge of lower impeller and 1 in 9 processes.

[0260] The 1000 l bioreactor operates at two discrete post-inoculation volumes with either the bottom impeller submerged during the 1 in 5 processes and 1 in 9 processes or with both impellers submerged during the N-2 phase of the 1 in 5 process and rolling seed operations for both 1 in 5 and 1 in 9 processes.

[0261] Table 20 shows the liquid cover above the upper and lower impeller during operation of the 1 in 5 and 1 in 9 sub-cultivation processes. During rolling operation of the 1 in 5 and 1 in 9 processes the liquid cover above the lower impeller falls below 0.5D. It is therefore important to reduce the agitation rate, to avoid surface gas entrainment, whilst operating at this low volume. At 960 litres a liquid cover, (D.sub.o) of 2.05Dtop is obtained. At this level K.sub.La has been shown not to be adversely affected and bulk blending is not an issue.

TABLE-US-00020 TABLE 20 Key operating volumes and the liquid cover above top impeller, Do and bottom impeller, D.sub.Bo Cylinderical Do, D.sub.Bo, Do as ratio D.sub.Bo as ratio Operating volume, L height, H (mm) mm mm of D.sub.A310 of D.sub.A315 Pre-Inoculation, 250 L 334 332 0.87D.sub.A315 Pre-Inoculation, 400 L 590 588 1.54D.sub.A315 Post - Inoculation and 419 417 1.10D.sub.A315 Pre-Transfer, 300 L Post - Inoculation and 675 673 1.77D.sub.A315 Pre-Transfer, 450 L Post drain, pre-bulk 235 233 0.61D.sub.A315 192 L Post drain, pre-bulk 50-100 L 0-78 76 0.2D.sub.A315 Post - Inoculation and 1443 679 1.78D.sub.A310 Pre-Transfer, 900 L Post - Inoculation and 1545 782 2.05D.sub.A310 Pre-Transfer, 960 L .sup.(1) Off bottom impeller clearance, Dc = 152 mm (0.4D.sub.A315), Impeller separation, Ds = 762 mm (2D.sub.A315), tank ID of 864 mm and Height of ASME F&D base plate, H.sub.h = 151 mm .sup.(2) Do = H Ds (Dc H.sub.h) and D.sub.Bo = H (Dc H.sub.h)
Agitation Raterpm, P/V and Tip Speed

[0262] Table 21 specifies the agitation rate for the 1000 litre bioreactor. The bioreactor is agitated at around 20-260 W/m.sup.3, preferably at 55-85 W/m.sup.3. The agitation strategy was developed during the 500 litre pilot fermentations. An agitation rate of up to 1551 rpm is used as an operational range.

TABLE-US-00021 TABLE 21 Agitation rate for the 1000 L bioreactor Agitation rate, Power per unit volume, Tip Speed, rpm W .Math. m.sup.3 m .Math. s.sup.1 .sup.10-155 0-150 3.1 .sup.20-145 0-145 2.9 .sup.1When both impellers submerged .sup.2When bottom impeller submerged

Mechanical Seals Specification

[0263] A double mechanical seal that is condensate lubricated as described was used.

Bioreactor Aeration Requirement

[0264] Table 22 shows the gas flows based upon scale up of constant superficial gas velocity, for DOT and pH control during the inoculum expansion in the 1000 litre bioreactor. Oxygen will not be required for DOT control. However oxygen enriched air may be used to facilitate lower gassing to prevent excess foaming. It is recommended that the smaller range CA MFC should be used to delivery nitrogen for early DOT control and reducing deviant, high levels of DOT.

TABLE-US-00022 TABLE 22 Gas flow rate and MFC operating ranges for the 1000 litre bioreactor Gas Operating range Comments Head Space.sup.1 1. Clean air 1. 0-50 SLPM 1. Head space purging of CO.sub.2 and O.sub.2 2. Nitrogen 2. Utility rated 2. For rapid DOT probe zeroing 3. Helium 3. Utility rated 3. Tank integrity testing Control Sparger 1. Clean air.sup.1 1. 2-20SLPM 1. Gas flow under DOT control 2. Oxygen 2. 0.2-5SLPM 2. Gas flow under DOT control 3. Carbon dioxide 3. 0.2-10SLPM 3. Gas flow under pH control 4. Nitrogen.sup.2 4. 0.2-5SLPM 4. Early DOT control by ballast 5. Helium 5. Utility rated 5. Tank integrity testing .sup.1The air and nitrogen gas flow into bioreactor via a bypass for post SIP tank pressurisation. .sup.2Nitrogen delivered via the 0 to 5SLPM CA MFC, could be used during early DOT control.

[0265] The calculation of hole size and number is iterated until the target Reynolds number of gas emerging from holes is <2000 and the Sauter mean bubble diameter for a bubble chain regime is approximately 10 mm.

[0266] Table 23 shows the key sparger design specification for the 1000 litre bioreactor. The sparger length, S.sub.L of 305 mm is determined for pipe geometry. The holes are distributed on either end of the sparger to prevent bubble liberating directly under the A315 hub. Alternatively a crescent geometry can be considered.

[0267] The pipe diameter is 25 mm. 30 2 mm holes are located on the dorsal surface of the sparger with a single 2 mm hole located on the ventral surface to aid free CIP drainage of the sparger.

[0268] The bioreactor port for sparger installation is designed to fit pipe design of diameter of 25 mm. The position of port allows the placement of control sparger at a distance of 88 mm (D.sub.cS.sub.c) below the bottom edge of the bottom impeller and no greater than 64 mm from tank bottom (S.sub.c).

TABLE-US-00023 TABLE 23 Design specification for 1000 litre bioreactor spargers Control Parameter Sparger Gas flow, SLPM 35 Number of sparge holes 30 Orifice diameter, d.sub.o, m 0.002 Gas flow, m.sup.3 .Math. s.sup.1 5.83E04 Orifice area, m.sup.2 3.14E06 Total orifice area, m.sup.2 9.42E05 Density of air, Kg .Math. m.sup.3 1.166 Viscosity, Nm .Math. s.sup.2 1.85E05 Sauter mean diameter, mm (d.sub.vs = 1.17 V.sub.o.sup.0.4 d.sub.o.sup.0.8 g.sup.0.2) 10.65 Gravitional acceleration, g, g m .Math. s.sup.2 9.807 Density difference, Kg .Math. m.sup.3 1048.834 Reynold's number, >2000 jetting regime 782 Gas velocity at sparger, V.sub.o, m/s 6.19 Sparger length, S.sub.L, m 0.305 Combined length to drill required holes, m 0.06 Number of rows to fit required holes in length S.sub.L, m 1 Sparger to tank bottom clearance, Sc, m 0.064 Sparger to bottom impeller clearance, Dc-Sc, m 0.088

Position of Probes, Addition and Sample Ports

[0269] The design basis for positioning of probes, addition and sample ports is the same as for the 20 000 l bioreactor.

TABLE-US-00024 TABLE 24 Probe, addition and sampling port specification for the 1000 litre bio-reactor Diameter, .sup.1Position, Probe/Port Location mm (inches) mm (inches) Rational Temperature Lower 38.1 (1.5) 1. 286 (11) Positioned to (main) ring 2. 30 minimise monitored volume Temperature Lower 38.1 (1.5) 1. 286 (11) Positioned to (backup) ring 2. 170 minimise monitored volume PH Lower 38.1 (1.5) 1. 286 (11) Positioned to (main) ring 2. 10 minimise monitored volume PH Lower 38.1 (1.5) 1. 286 (11) Positioned to (backup) ring 2. 20 minimise monitored volume DOT Lower 25.0 (0.98) 1. 286 (11) Positioned to (main) ring 2. 150 minimise monitored volume DOT Lower 25.0 (0.98) 1. 286 (11) Positioned to (backup) ring 2. 160 minimise monitored volume Spare-2 Lower 38.1 (1.5) 1. 286 (11) Positioned to (spare - ring 2. 180 minimise monitored pCO.sub.2) volume Spare-3 Lower 50.8 (2) 1. 286 (11) Positioned to (spare - ring 2. 190 minimise monitored Biomass) volume Sample valve Lower 38.1 (1.5) 1. 286 (11) NovAseptic (main) ring 2. 40 type Sample valve Lower 38.1 (1.5) 1. 286 (11) NovAseptic (back up) ring 2. 40 type Alkali Lower 12.7 (0.5) 1. 286 (11) Diametrically addition ring 2. 190 opposite pH probes Feed 1 Lower 12.7 (0.5) 1. 286 (11) Diametrically ring 2. 200 opposite pH probes Feed 2 Lower 12.7 (0.5) 1. 286 (11) Diametrically ring 2. 210 opposite pH probes Antifoam Head 50.8 (2) 1. N/A Liquid surface/0.25T addition plate 2. 170 from tank centre Spare surface Head 50.8 (2) 1. N/A Liquid surface addition plate 2. 180 directed to vessel wall DOT control - N/A 50.8 (2) 1. 64 (2.5) sparger orifice 2. 0 Overlay Head 38.1 (1.5) 1. N/A Diametrically gas plate 2. 135 opposite vent out Exhaust Head 38.1 (1.5) 1. N/A Diametrically vent out plate 2. 315 opposite overlay gas in Transfer Base 50.8 (2.0) 1. N/A NovAseptic type to valve plate 2. Centre allow free draining Media Head 76.2 (3) 1. N/A Directed into inlet plate 2. 310 vessel wall Inoculum transfer Head 50.8 (2.0) 1. N/A Directed into from S200 to plate 2. 320 vessel wall 1000 L CIP - Head 76.2 (3) 1. N/A CIP'ing of Spray ball plate 2. 270 highest point CIP - Head 76.2 (3) 1. N/A Spray ball plate 2. 60 Pressure Head 38.1 (1.5) 1.) N/A gauge plate 2.) 50 Rupture Head 50.8 (2) 1.) N/A disc plate 2.) 280 Spare Head 101.6 (4) 1.) N/A nozzle plate 2.) 160 1. Hand hole Head 1. 203.2 (8) 1.) N/A Single port 2. Sight glass plate 2. 101.6 (4) 2.) 70 permitting two functions Agitator shaft Head 152.4 (6) 1.) N/A Centre of opening plate 2.) 75 head plate .sup.1Measured from the tangential line of the base plate. Degree pertains to plane of clockwise rotation. .sup.(2) Diameter of nozzle at the bioreactor

[0270] In order to monitor, control and sample from a volume of 50 l, the probes and port ring needs to be 151 mm from tank bottom. However the probe/port ring cannot be located this low as it falls on the weld of the base plate and the straight cylindrical side of the bioreactor. The probe and port ring has been specified at 286 mm from tank bottom. This permits a volume of 134 litres to be monitored, controlled and sampled. The probes/port ring is located as close to the tank bottom as permitted to minimise the monitored/controlled volume.

Addition Ports, Surface and Sub-Surface

[0271] The 1000 litre bioreactor has been designed to accept two subsurface feeds and alkali to be discharged into a well-mixed area of the bioreactor. The foam is controlled by surface addition of 1 in 10 diluted C-emulsion. A single above surface spare addition port directed to the vessel wall was also integrated for future flexibility. The splashing of culture on to the surface of the medium during inoculation of seed bioreactor should be avoided to prevent build up of foam. Therefore the inoculum addition port is above surface and directed to the vessel wall. The use of the harvest port in the base plate is the ideal port for removal of inoculum during transfer of inoculum. Additionally the medium addition port is directed on to the vessel wall. In summary the total addition ports are: [0272] Four surface additions with medium inlet, inoculum inlet and a spare small addition directed to the vessel wall and addition port for antifoam dropped on to the liquid surface away from the vessel wall. [0273] Three subsurface additions for feeds and alkali.

Sample Ports

[0274] The sample port design is similar to that specified for the 20 000 litre bioreactor. The sample port is located 286 mm from tank bottom to minimise the volume that can be sampled.

Volume Measurement

[0275] The level sensor is able to measure up to 1000 litres. The level sensor sensitivity is at least 0.25% of full span.

Bioreactor Temperature Control

[0276] The 250 to 800 litres of medium is brought to operating temperature, typically 36.5 C. during initial inoculation and seed rolling operation by process control. This is achieved by gentle heating of the jacket and avoid high temperature at vessel wall.

Jacket

[0277] The bioreactor jacket area is specified with the following considerations in mind: [0278] Steam sterilisation at 121-125 C. [0279] Warming up of 250-800 litres of medium from 10 C. to 36.5 C. in <2 hrs. [0280] All points within the bioreactor must reach 0.2 C. of set point, typically 36.5 C. as measured by thermocouples. [0281] Chilling of 400 litres of medium from 362 C. to 10 C. in 2 hrs.

Bioreactor pH Control

[0282] The process pH is monitored and controlled with probes connected via a transmitter to a DCS based process controller. The process pH is controlled by addition of CO.sub.2 to bring the pH down to set point and addition of alkali to bring pH up to set point. Alkali is added through at least one subsurface port at centre-line of the bottom impeller. The CO.sub.2 is added via the control sparger.

[0283] The control and back up probes are in the lower port ring at 286 mm from tank bottom to minimise the volume that can be monitored as shown in Table 24.

Bioreactor DOT Control

[0284] Dissolved oxygen is monitored and controlled with polarographic DOT probe. The DOT set point maintained by sparging: [0285] Initial N.sub.2 ballast and/or air on demand [0286] Air ballast with air on demand [0287] Air ballast with oxygen on demand

[0288] DOT control allows DOT set point to be maintained through interchangeable use of oxygen or air as demand gas.

[0289] Control and back up probes are in the lower port ring at 286 mm from tank bottom minimise the volume that can be monitored, as shown in table 24.

Feed Addition Control

[0290] The point of addition is 286 mm from tank bottom, in the vicinity of the centre-line of the bottom impeller to assist in the rapid dissipation of feed bolus.

Antifoam Addition Control

[0291] The addition point is at surface projecting 0.25T toward the tank centre or 216 mm from tank centre.

EXAMPLE 4: BIOREACTOR TRAIN

[0292] The bioreactor design is based on the ability to perform both 1 in 5 (20% v/v) and 1 in 9 (11% v/v) subculture ratios. The bioreactor train consists of a 1000 litre (Stages N-3 and N-2) and 4000 litre (Stage N-1) seed bioreactors followed by a 20 000 litre production bioreactor (Stage N). The operating volumes for each bioreactor are defined in examples 1 to 3. The bioreactors are based on a stirred tank design and a top driven agitator system was used.

[0293] The design is based on the need to ensure a homogenous environment with respect to process parameters such as pH, dissolved oxygen tension (DOT) and temperature, maintaining a well mixed cell suspension and blending nutrient feeds within the bioreactor. This provides the necessary physicochemical environment for optimal cell growth, product accumulation and product quality. Key to the design philosophy is the need to maintain geometric similarity. This allows a scale down model to be developed at 12 litre laboratory and 500 litre pilot scales. The design of the seed and production bioreactors are based on the same principles although some departures are required to allow for flexibility in processing. The aspect ratios (H.sub.L/T) selected are typical of those used in mammalian cell culture and are in the range 0.17 to 1.96 post-inoculation.

TABLE-US-00025 TABLE 25 Key bioreactor design parameters 1000 litre 4000 litre 20000 litre Aspect ratio (H.sub.L/T) 0.17-1.96 0.63-1.21 0.83-1.34 Impeller to tank 0.44-0.46 0.44-0.46 0.44-0.46 diameter (D/T) Operating Volume (L) 50-960 1914-3846 13913-21739 Agitator speed (rpm) 0-155 0-88 0-80 Control sparger CA (SLPM) 2-20 0-60 0-1000 Ballast sparger No ballast No ballast 0-500 CA/N.sub.2 flow (SLPM) sparger sparger Cultivation .sup.2-5.sup.1 2-5 10-15 residence time (days) Feed additions 2 surface 2 surface 4 surface 3 sub-surface 3 sub-surface 4 sub-surface .sup.1The culture residence time in 1000 litre bioreactor may be higher depending on the length of time the bioreactor is repeatedly sub-cultivated or rolled.

[0294] The design constraint is based upon a seeding ratio of 11% v/v (1 in dilution) and 20% v/v (1 in 5 dilution), with feed application of 4% v/v to 25% v/v of the post-inoculation volume. The post-inoculation volume in the production bioreactor is adjusted for feed applications up to 15% such that after the addition of all the feeds the final volume at harvest ends up at 20 000 l. However for feed applications greater then 15% v/v the post-inoculation volume is adjusted for a 15% v/v feed but following the application of feeds the final pre-harvest volume will be a minimum of 20 000 and a maximum 22 000 litres. The production bioreactor is expected to hold a total of 20 000 to 22 000 litres at the end of a batch. Table 26 shows the pre-inoculation volume, inoculation volume and transfer or harvest volume for each of the three inoculum expansion stages and the production bioreactor.

[0295] The seed bioreactors (stage N-1 to N-3) are unlikely to be fed therefore the maximum operating volume will be at inoculation. The operating volume range for the 4000 litre seed bioreactor (stage N-1) is to 3846 litres. In order to design a bioreactor that can grow cells from 20% v/v seed split ratio, the 1000 litre seed bioreactor (stages N-2 and N-3) will operate at two operating ranges. For the 11% v/v seed split ratio the bioreactor train can produce sufficient culture to meet forward processing cell concentration criteria in a single expansion/sub-cultivation stage. However the bioreactor train requires two expansion/sub-cultivation stages to meet forward processing criteria for 20% v/v seed split ratio process. Thus for 11% v/v seed split ratio process an operating range of 400 to 450 litres is required and for the 20% v/v seed split ratio process an operating volume range of 250 to 960 litres is required.

TABLE-US-00026 TABLE 26 Vessel sizes for bioreactor train 1000 4000 20000 litre litre litre Stage N-3 N-2 N-1 N 11% v/v Seed with 4 to 25% v/v production feed Pre-inoculation Volume (L) 400 1914 15456-17096 Inoculation Volume (L) 450 2153 17391-19231 Transfer or Harvest Volume (L) 450 2153 20000-21739 20% v/v Seed with 4 to 25% v/v production feed Pre-inoculation Volume (L) 250 768 2782-3077 13913-15385 Inoculation Volume (L) 300 960 3478-3846 17391-19231 Transfer or Harvest Volume (L) 300 960 3478-3846 20000-21739 Assumed operating volume Minimum Volume (L) 250 1914 13913 Maximum Volume (L) 960 3846 21739 Ratio of Maximum 3.84 2.01 1.56 volume/Minimum volume

[0296] It is recommended that the 1000 litre seed bioreactor is inoculated from culture produced in an S200 Wave bioreactor.

[0297] 1000 l: This bioreactor is operated in batches of up to 5 days, with potential shot additions of feeds, for cultivation of mammalian cells. However due to repeated drain and refill operation at the end of each batch the total process residence time in this bioreactor can exceed 30 days. The mammalian cells are kept in a homogeneous suspension by agitation via an identical impeller system to the 20 litre bioreactor. Additionally other features will be kept geometrically similar to the 20 000 litre bioreactor, where possible.

[0298] Sparging air or oxygen and air or nitrogen respectively will control process DOT. Process pH is controlled by addition of alkali for base control and of sparged CO.sub.2 for acid control.

[0299] The process operating volume of the bioreactor changes at different phases of operation. Initially the bioreactor is aseptically filled with a bolus of medium at 250 to 400 litres in 0.5 h. The bioreactor is operated in a pre-inoculation phase to bring the process variables to predefined set points. 50 litre culture from a (N-4) S-200 seed wave bioreactor is inoculated, by pneumatic assisted flow, or pumped with a peristaltic pump in 25 to 30 minutes into the 1000 litre bioreactor at 1 in 5 or 1 in 9 dilutions. The post-inoculation operating volume is 300 and 450 litres for 1 in 5 and 1 in 9 seeded process respectively. The addition of alkali for base control and 1 in 10 antifoam suspension for suppression of foam contributes towards the final volume. The inoculum culture may be fed by a shot addition if the culture interval is longer then expected. As a result of mixing and gassing the liquid volumes described above will expand due to gas hold up. The extent of this rise is dependent on the sparger type used, power per unit volume imparted by impellers and superficial gas velocity of sparged gasses.

[0300] The N-3 stage ends when viable cell concentration reaches transfer criteria. The N-2 stage for 1 in 5 process begins with a bulk up in volume to 960 litre by draining of 192 litre excess culture and addition of 768 litre fresh medium in 1.5 h. 696 to 769 litre of culture are transferred at the end of N-2 stage to the 4000 litre bioreactor for the in 5 processes. For the 1 in 9 processes 239 litres are transferred to the 4000 litre bioreactor.

[0301] The 1000 l bioreactor is continuously drained and refilled with fresh medium or rolled to provide back up culture for the 4000 litre bioreactor. The duration of the rolling seed operation is dependent on the length of the production campaign and the permissible elapsed generations number of the seed culture. Typically it is assumed that rolling seed operation is in excess of 30 days. The rolling operation consists of retaining approximately 192 litres of the 960 litre culture and diluting with 768 litre fresh medium for the 1 in 5 processes. For the 1 in 9 processes the 1000 litre bioreactor is expected to be rolled by retaining 50 to 100 litre of the 450 to 900 litre culture and diluting with 400 to 800 litre fresh medium. Process control ranges are relaxed over this operation. The medium added to the bioreactor during rolling operation is warmed to 30 C.

[0302] 4000 l: This bioreactor is operated in batch of no more then 5 days, with potential shot additions of feeds, for cultivation of mammalian cells. The mammalian cells are kept in a homogeneous suspension by agitation via an identical impeller system described in example 1. Additionally this vessel is geometrically similar to the 20 000 litre bioreactor.

[0303] Sparging air or oxygen and air or nitrogen respectively controls process DOT. Process pH is controlled by addition of alkali for base control and of sparged CO.sub.2 for acid control.

[0304] The process operating volume of the bioreactor changes at different phases of operation. Initially the bioreactor is aseptically filled with a bolus of protein free medium at 1914 to 3077 litres in 1.5 h. The bioreactor operates in a pre-inoculation phase to bring the process variables to predefined set points. Culture from the 1000 litre (N-2) seed seed bioreactor is inoculated by pneumatic flow at a flowrate to allow transfer in one hour, at 1 in 5 or 1 in 9 dilutions. The post-inoculation operating volume is 2153 to 3846 litres. The addition of alkali for base control and 1 in 10 antifoam suspension for suppression of foam contributes towards the final volume. The inoculum culture may be fed by a shot addition if the culture interval is longer then expected. As a result of mixing and gassing the liquid volume expands due to gas hold up. The extent of this rise is dependent on the sparger type used, power per unit volume imparted by impellers and superficial gas velocity of sparged gasses.

[0305] 20 000 l: This bioreactor is operated in batch or fed batch mode for to 15 days for the cultivation of mammalian cells. The mammalian cells are kept in a homogeneous suspension by agitation via an impeller system.

[0306] The process operating volume of the bioreactor changes at different phases of operation. Initially the bioreactor is aseptically filled with cell culture medium at 13913 to 17096 litres in 1-2 h. The bioreactor is operated in a pre-inoculation phase to bring the process variables to predefined set points. Culture from the 4000 litre seed bioreactor (N-1) is inoculated by pneumatic flow at a flow rate range of <4000 l/h into the 20 000 litre bioreactor at 1 in 5 or 1 in 9 dilutions. The post-inoculation volume continuously increases following an application of sub-surface feeds to maximum of 20 000 litres (two feeds totalling 4 to 25% v/v). The addition of alkali for base control and 1 in 10 antifoam suspension for suppression of foam accounts for about 100 litres and 20 litres respectively. As a result of mixing and gassing the liquid volume expands due to gas hold up. The extent of this rise is depended on the sparger type used (fluted or sintered), power per unit volume imparted by impellers and superficial gas velocity of sparged gasses.

[0307] Table 27 describes the aspect ratios in the 20 000 litre bioreactor at various operating volumes during normal processing. The aspect ratios have been tested at 500 litre scale and provided the superficial gas velocity and power per unit volume are kept constant the K.sub.La remains constant.

TABLE-US-00027 TABLE 27 Key operating volumes and aspect ratios in the 20 000 litre bioreactor Volume, Liquid head, Aspect ratio, L mm H.sub.L/T Pre-Inoculation 13913-17096 2458-2977 0.88-1.07 Post Inoculation 17391-19231 3025-3325 1.08-1.19 Harvest 20000-21739 3451-3734 1.23-1.34