Chap 8 Bioreactors

Introduction

 Applications of bioreactors: for production of vaccines, proteins, organic acids, amino

acids and antibiotics; enzymatic or microbial biotransformations; bioremediation, etc.
 A production facility usually has a train of bioreactors ranging from 20 L to 250,000 L.
The bioreactors are arranged in the series of increasing sizes, starting from small cultures
to the final production culture.

I. Bioreactor configurations
Stirred tank reactors

Features:
 Microbial reactors have impellers to provide D
agitation and generally have 4 baffles from the
walls to prevent vortexing of the fluid, the baffle d
width is 1/10 or 1/12 of the tank diameter.
 The vortex and circular flow result in little mixing 1-2 d

between fluids at different heights. At high speeds

the vortex may reach down to the impeller so that
gas from the surrounding is drawn into the 1/3 D
liquidhigh mechanical stress in the stirrer shaft,
bearings and seal.
 Bioreactors for animal cell cultures usually do not
have baffles (especially for small scale reactors) to
reduce turbulence.
 The aspect ratio (height-to-diameter ratio) of the vessel is 3-5 for microbial cultures but
is normally less than 2 for animal cell culture.
 Sparger: gas is sparged at the bottom using a perforated pipe ring sparger.

1
 Number of impellers depends on the aspect ratio. The bottom impeller is located at a
distance about 1/3 of the tank diameter above the bottom of the tank. Additional
impellers are spaced approximately 1 to 2 impeller diameter (d) distances apart.
 The superficial aeration velocity (the volume flow rates of gas divided by the cross-
sectional area of the vessel) in stirred vessel must be lower than that can flood the
impeller (an impeller is flooded when it receives more gas than it can effectively disperse) otherwise the mixing is
poor. Superficial aeration velocities generally do not exceed 0.05 m/s.
 Impellers: choice often depends on the viscosity of the liquid and sensitivity of the cells
to mechanical shear.
 Rushton (6-flat-blade) disc turbine (a) and
concave bladed impeller (b): impeller
diameter is about 1/3 of the vessel diameter
and is often used for bacterial cultures.
Rushton turbine is most commonly used in
fermentation technology.
 Hydrofoil impeller (c): diameter is about 0.5
to 0.6 times the tank diameter and is an
effective mixer for highly viscous mycelial
broths.
 Marine impeller (d): usually single, large diameter, low shear, used for animal cell
culture.

Flow pattern of Rushton Flow pattern of marine impeller

turbine (promotes axial flow)

2
 Impeller speed:
 Usually <120 rpm for animal cell cultures even for vessels >50 liters. Higher
stirring speeds can be used for microbial cultures.
 The impeller tip speed (3.14 × impeller diameter × speed of rotation) is usually less
than 7.6 m/s for filamentous fungi.

 Power requirement:
 critical for large-scale bioreactors
 Smaller for sparged reactors because (1) gas bubble decreases the liquid
density; (2) gas-filled cavities develop behind the stirred blades, which reduce
the resistance to fluid flow and decrease the drag coefficient of the impeller.
P0  N p  l N i Di5
3
for ungassed Newtonian fluid

P0: power input (W)

l: liquid density; Ni: rotational speed (1/s); Di: diameter of impeller
Np: power number (related to Reynolds number)
Pg Fg 0.25 N i2 Di4 0.20
 0.10  ( ) ( ) for gassed Newtonian fluid
P0 NiV gWiV 2 / 3
Pg: power consumption with sparging; Fg: volumetric gas flow rate;
V: liquid volume; g: gravitational acceleration; Wi: impeller blade width

3
Bubble column reactor

 Usually the height-to-diameter ratio is 4-6.

 Gas is sparged at the base through perforated pipes or plates or
metal porous spargers.
 O2 transfer, mixing and other performance factors are influenced
mainly by gas flow rate and rheological properties of the fluid.
 Mixing and mass transfer can be improved by placing perforated
plates or vertical baffles in the vessel.

Airlift bioreactor Air

 Separated as two zones: the sparged

zone is called the riser, and the zone
that receives no gas is the downcomer.
The bulk density in the riser region is
lower than that in the downcomer
region, causing the circulation (so
circulation is enhanced if there is little or no
gas in the downcomer).
 For optimal mass transfer, the riser to
downcomer cross-sectional area ratio
should be between 1.8 and 4.3.
 Highly energy efficient and
productivities are comparable to those of stirred tank bioreactors.
 The rate of liquid circulation increases with the square root of the height of the airlift
device. Consequently, the reactors are designed with high aspect ratios.
 A gas-liquid separator in the head-zone can reduce the gas carry-over to the downcomer
and hence increase the liquid circulation.

Fluidized bed reactor

 Suitable for reactions involving a fluid-suspended particulate biocatalyst such as

immobilized enzyme and cell particles

4
 Similar to the bubble column reactor except that the top
section is expanded to reduce the superficial velocity of the
fluidizing liquid to a level below that needed to keep the
solids in suspension. Consequently, the solids sediment in
the expanded zone and drop back, hence the solids are
retained in the reactor whereas the liquid flows out.

Packed bed bioreactor

 A bed of particles are confined in the reactor. The

biocatalyst (or cell) is immobilized on the solids which may
be rigid or macroporous particles.
 A fluid containing nutrients flows through the bed to provide the needs
of the immobilized biocatalyst. Metabolites and products are released
into the fluid and removed in the outflow.
 The flow can be upward or downward. If upward fluid is used, the
velocity can not exceed the minimum fluidization velocity.

II. Bioreactor design features

 Medium or feed nozzle (19).

 Vertical sight glass (15) and

ports for pH, temperature
and DO sensors (6).
 Connections for inoculum,
acid and alkali (for pH
control) and antifoam
agents are located above
the liquid level in the
reactor vessel (16).
 O2 and other gases (CO2 or
NH3 for pH control; N2 for
O2 control) can be

5
 introduced through a sparger at the bottom (8).
 Foam breaker (22) is used when antifoam is ineffective or the antifoam interferes with
downstream processing (antifoam tends to foul the membrane during filtration).
 Can be sterilized in-place using saturated steam (14) at a minimum absolute pressure of
212 kPa. Over–pressure protection is provided by a rupture disc (24) on the top of the
reactor, which cracks to relieve the pressure to avoid explosion.
 Maximum allowable working pressure is 377-412 kPa (absolute)1, allowable
temperature is usually 150-180C (>121C for sterilization). The vessel should
withstand full vacuum or it could collapse while cooling after sterilization.

BioFlo 6000® Sterilizable-In-Place

Fermentor-Bioreactor (50 - 130 L) BioFlo 110, 1.3 to 14 liters

 Usually made in Type 316L stainless steel, while the less expensive Type 304 (or 304L)
is used for the jacket. The L grades contain less than 0.03% carbon, which reduces
chromium carbide formation during welding and lowers the potential corrosion at the
welds.
 The vessel should have as few internals as possible and should be free of stagnant areas
where pockets of solids or liquids may accumulate.

III.Design for sterile operation

1 1 atm=1.01105 Pa=101 kPa

6
Sterilization-in-place

 A bioreactor must be
sterilized before inoculation
because contamination is a
common cause of process
failure.
 For large bioreactors, in situ
sterilization is common.
 The components should be
able to be sterilized
independently during
fermentation if required.
 The aeration and exhaust groups must also be sterilized. The filters are rated for
removing particles down to 0.22 m or even 0.1 m. Often the gas streams require two
filter cartridges in series, with the first serving to protect the second filter.

Clean-in-place (CIP) considerations

 Industrial bioreactors, including all pipings, should be cleaned in-place using automated
methods, so as to ensure consistency and reduce down-time.
 To remove solid particles and avoid sedimentation, a flow velocity of 2 m/s is preferred.
Also, the piping should be free of dead space as much as possible. For thorough cleaning,
the CIP solutions are sprayed through a spray ball. For cleaning with jet spray, pressures
of 308 to 377 kPa (absolute) are optimal.

7
 Procedures:
1. Pre-rinse for 5 min with deionized water (sufficient for bacteria, yeast and animal
cell cultures).
2. Circulate 1% (w/v) NaOH at 75-80C through all product contact surfaces for 15-20
min. Discard the solution afterwards.
3. Rinse at 25-35C with deionized water to remove all alkali.
4. Final wash with hot “water-for-injection” grade water.
 For stirred tank bioreactors, it is recommended to fill the vessel and agitate at Reynolds
numbers of 108-108.5 during pre-rinse, alkali recirculation and the final rinse for 2-3 min
(should be sufficient to dislodge adhering dirt or soil).
Note:
 Disposable bioreactors (e.g. Wave bioreactor, BelloCell) are gaining increasing interest
due to smaller capital investment, easier operation and elimination of the CIP process.

BelloCell 500 AP
(Cesco Bioengineering)
Wave Bioreactor (100-500 L, GE
http://www.cescobio.com.tw/products.php?tid=1
Healthcare, http://www.wavebiotech.com/)

8
 Water for Injection (WFI): high quality water subjected to the following treatment:

membrane filter (5-20 m)

Activated carbon (organic and other contaminants are adsorbed)

Ion exchange (cation & anion)

dI water (purified water)

Reverse osmosis distillation

WFI WFI

Note:
1. Ion exchanger need to be regenerated regularly by HCl and by NaOH
2. Reverse osmosis removes viruses, microorganisms, pyrogen and virtually all inorganic impurities.

9
Appendix

IV. Mass transfer steps

 In bioreaction processes, substrates are consumed for the conversion. Typical substrates
include carbon sources (e.g. sugar and oil), nitrogen sources (e.g. ammonia and amino
acids) and electron acceptors (e.g. O2).

Effects of transfer limitations

Two effects if one step is slower than the key kinetic reaction step:
1. The overall reaction rate is below the theoretical maximum, and the process output is
slower than desired.
 Reversible effect: for the production of gluconic acid (葡萄糖酸) from glucose by
Gluconobacter oxydans, the O2 transfer is limiting. Once O2 limitation is relieved,
there is no irreversible effect on this microorganism.
 Irreversible effect: for the production of penicillin, O2 limitation imposes an
irreversible damage to the biosynthetic capacity of the cell.
2. The selectivity of the reaction is altered.
 Ex: O2 serves as an electron acceptor in the formation of baker’s yeast from glucose.
In the absence of O2 the e- will be directed to pyruvate resulting in the formation of
ethanol and CO2.

10
 Transfer of oxygen involves a chain of mass transfer steps from a gas bubble. The
slowest one is the rate-limiting step and determine whether the mass transfer rate would
affect the overall process performance. When cells are well dispersed in the liquid and
the bulk liquid is well mixed, step (iii) is the limiting step.

Transfer across the cell envelope

 Transport across the cell envelope (may include cell wall and cytoplasmic membrane)
can be limited.
 Three typical mechanisms:
1. Free diffusion: passive transport down a concentration gradient
2. Facilitated diffusion: as above but speeded up by a carrier protein
3. Active transport: transport by a carrier protein with input of free energy

 The diameter of microbial cell itself is small (usually 1-5 m) so diffusion inside the
cell is more rapid and not a limiting factor.
 The transport barrier imposed by the membranes of intracellular organelles in eucaryotic
cells usually does not limit the overall transport rate.

V. Mass Transfer Equations

Fundamentals:

Fick’s Eqn.
J=-DdC/dx (at steady state)
J: molar mass flux (mol/m2s)
D: diffusion coefficient (m2/s)
C: concentration of the substance to be transported
x: distance

11
 Considering transport in solid phase, D is the effective diffusion coefficient which lumps
the diffusion coefficient, the porosity of the solid, and the shape of the channels.
 For a flat plate with thickness d in a stationary fluid
J=DC/d
Where D/d represents the mass transfer coefficient, while d/D can be interpreted as the
resistance against transport.
 For unsteady state:

 2 C C
D 
x 2 t
 These equations consider the diffusion process only but not convection. In reality,
convection is often encountered and flow pattern is not known, thus the mass transport
often relies on empirical approach.

Mass transfer between l-s phase or l-g phase (two film

theory)

For gas film transport:

J g  k g ( P  Pi ) (mass transfer rate, not flux)

Liquid film transport:

J l  k L a(Ci  C ) (the main resistance to oxygen transfer)

kL: liquid phase mass transfer coefficient

(m/h)
Pi Ci but can be correlated by Henry coefficient H (barm3/mol):
Pi=HCi
From here, the volumetric mass transfer rate (J,
mg/h/l), can be derived
J  k L a(C *  C )
 a (m2/m3): the gas-liquid
interfacial area per unit liquid
volume, or area per unit gross
vessel volume in the bioreactor.

12
 When dealing with the transfer of O2 from gas to liquid, J is called the OTR
(oxygen transfer rate)
 kLa (1/h): volumetric transfer coefficient, because a and kL are difficult to
evaluate separately, kLa is often expressed together. Various expressions of kLa
can be found in literature. The value is typically 0.020.25 s-1
 C*: saturated DO (in the case of oxygen transfer) concentration (g/l), i.e. the
solubility of O2
 C: actual DO concentration in the broth
 Note: mole fraction of O2 in air is 0.2099, so the partial pressure of O2 at 1 atm
air pressure is 0.2099 atm. Based on Henry’s law, the solubility of oxygen in
water under 1 atm air pressure is 0.2099 times that under 1 atm pure O2.
 Oxygen uptake rate:
QO  qO  x (g/lh) qo: specific oxygen uptake rate (g/gs), x: cell concentration (g/l)

 qo varies with cell species and nutritional environment

 For C> Ccrit (critical oxygen concentration), qo is a constant maximum
 For C< Ccrit , qo (usually 5-10% air saturation) is approximately linearly
dependent on C.
At steady state,
k L a(C *  C )  qo  x
This equation can be used to predict the response of fermenter to changes in mass transfer
operating conditions, for example, if kLa is raised by increasing stirrer speedC must rise.

Measurement of kLa

1. Oxygen-balance method: (based on steady state measurement)

at s.s.
1
J (( Fg C ) i  ( Fg C ) o )
VL
where l.h.s: rate of oxygen transfer from gas to liquid
r.h.s.: difference in oxygen flow between inlet (subscript i) and outlet (subscript o).
VL, volume of liquid; Fg, volumetric gas flow rate; C, gas phase concentration of O2

13
  Because gas phase concentrations are usually measured as partial pressures, ideal
gas law can be incorporated:
1 Fg p Fg p
J (( )i  ( )o )
RV L T T
where p is the partial pressure at the inlet and outlet. The difference between pi and po is
usually small thus pi and po need to be measured very accurately (e.g. by mass
spectrometry).
 For known p, Fg, VL, R and T, J is obtained, and kLa can be calculated from
J  k L a(C *  C ) when C* and C are measured.

2. Dynamic method: (based on unsteady state measurement)

 At time t0, the broth is de-oxygenated by
sparging N2 or by stopping the air flow if
the culture is consuming O2 C drops
 Air is then pumped into the broth at a
constant flow rate C increases and

reaches a s.s. value, C . C1 and C2 are two
oxygen concentrations measured at t1 and t2.
 During the re-oxygenation step (unsteady
state)
dC
 k L a(C *  C )  qo  x
dt

 At s.s., k L a(C *  C )  qo  x is substituted into the above equation to yield

dC 
 k L a(C  C )  qo  x
dt
Assume kLa is a constant, integrating the above
equation from t1 to t2 yields

14
 
C  C1
ln( 
)
C  C2
kLa 
t 2  t1

Gas-Liquid mass transfer in real systems (O2 transfer)

 Bubble size is critical in determining oxygen transfer, smaller bubbles leads to:
 higher a
 slower bubble rise velocity higher gas hold-up  (volumetric fraction of gas in the liquid).
VG
 VG: volume of gas bubbles in the reactor, VL: volume of liquid
V LVG
 Two extreme cases:
 Coalescing (融合) liquid (a liquid which greatly stimulates bubble
coalescence): mass transfer is poorest
 Non-coalescence fluid: highest mass transfer.
Stirred tank reactor:
 Sparged gas is usually rapidly collected in the gas cavities behind the rotating impeller
blades. The cavities flow in a highly turbulent vortex and the gas is dispersed into
smaller bubbles. These follow the liquid flow, but will also rise to the surface. They
will coalesce in areas that are relatively calm and re-disperse in places where the shear
stress is high. A part of the bubbles is re-circulated into the cavities and the rest escapes
at the surface.
 Empirical values for gas hold-up
  0.13  ( P / V ) 0.33 (v g p0 / p) 0.67 (for coalescing fluid)
where (P/V)=power input per unit volume, p=pressure in the system
 Empirical equations of kLa: (coarse equations with accuracy30%, valid for P/V=0.510
kW/m3).
k L a  0.026  ( P / V ) 0.4 (v g p0 / p) 0.5 for coalescing fluid (e.g. clean air-water system)

15
 k L a  0.002  ( P / V ) 0.7 (v g p0 / p) 0.2 for non-coalescing fluid (e.g. fermentation broth)

vg, superficial gas velocity ; po, reference pressure of 1 bar.

Bubble column:
 When air flow rate is high enough,
  0.6(v g po / p) 0.7

k L a  0.32  (v g po / p) 0.7

 In non-coalescing liquids (e.g. some fermentation broths), the bubble rises and does not
mix with other bubbles, provided that the size is smaller than the equilibrium average
diameter 6 mm.
 In a large bubble column (>50 m3), the bubbles will significantly expand as they rise
because of the decreasing hydrostatic pressure, which influence the mass transfer.
 kL: usually 3-410-4 m/s for bubbles>2-3 mm diameter, could be down to 110-4 m/s
for smaller bubbles, depending on bubble rigidity
Air-lift reactor:
 Although the riser resembles a bubble column, the gas-hold up is lower than predicted by
the above equation due to the interaction with the liquid flow.
 Correspondingly, kLa will be lower, up to a third of the bubble column. A precise
quantification, however, cannot be easily made.

16