INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

CH-528
Biochemical Engineering
Scale up of bioreactors
By: Prof. Prasenjit Mondal
Department of Chemical Engineering
Contents

8.1 Scale-up and Dimensionless numbers for Scale-up

8.2 Scaling of mass transfer equipment

8.3 Criteria for scaling up fermentors

8.4 Design estimation of various scale-up

8.5 Power estimation for Ungassed Liquids

8.6 Power estimation for Gas-sparged Liquids

2
8.1 Scale-up and Dimensionless
numbers for Scale-up
For the optimum design of a production-scale fermentation system
(prototype), we must translate the data on a small scale (model) to the large
scale. The fundamental requirement for scale-up is that the model and
prototype should be similar to each other.
Two kinds of conditions must be satisfied to insure similarity between
model and prototype. They are:
1. Geometric similarity of the physical boundaries: The model and the
prototype must be the same shape, and all linear dimensions of the
model must be related to the corresponding dimensions of the prototype
by a constant scale factor.
2. Dynamic similarity of the flow fields: The ratio of flow velocities of
corresponding fluid particles is the same in model and prototype as well
as the ratio of all forces acting on corresponding fluid particles. When
dynamic similarity of two flow fields with geometrically similar
boundaries is achieved, the flow fields exhibit geometrically similar flow
patterns.
3
Continued…

The first requirement is obvious and easy to accomplish, but the second is
difficult to understand and also to accomplish and needs explanation. For
example, if forces that may act on a fluid element in a fermenter during
agitation are the viscosity force , drag force on impeller , and gravity force ,
each can be expressed with characteristic quantities associated with the
agitating system. According to Newton's equation of viscosity, viscosity
force is
𝐹 𝑉=𝜇 ( 𝑑𝑢
𝑑𝑦 ) 𝐴 8.1

where is velocity gradient and A is the area on which the viscosity force
acts. For the agitating system, the fluid dynamics involved are too complex
to calculate a wide range of velocity gradients present. However, it can be
assumed that the average velocity gradient is proportional to agitation
speed N and the area A is to which results.

4
Continued…
2
𝐹 𝑉 ∝𝜇 𝑁 𝐷𝐼 8.2
The drag force can be characterized in an agitating system as
𝑃 𝑚𝑜
𝐹𝐷 ∝ 8.3
𝐷𝐼 𝑁
Since gravity force is equal to mass m times gravity constant g,
2
𝐹 𝐺∝ 𝜌 𝐷𝐼 𝑔 8.4
The summation of all forces is equal to the inertial force as,

∫ 𝐹=𝐹𝑉 +𝐹 𝐷 +𝐹 𝐺=𝐹 𝐼 ∝𝜌 𝐷 𝑁 4
𝐼
2
Then dynamic similarity between a model (m) and a prototype (p) is
8.5

achieved if
( 𝐹 𝑉 )𝑚 ( 𝐹 𝐷 )𝑚 ( 𝐹 𝐺)𝑚 ( 𝐹 𝐼 )𝑚
= = = 8.6
( 𝐹 𝑉 )𝑝 ( 𝐹 𝐷) 𝑝 ( 𝐹 𝐺)𝑝 ( 𝐹 𝐼 )𝑝

5
Continued…

or in dimensionless forms:

( ) ( )( ) ( )( ) ( )
𝐹𝐼
𝐹𝑉 𝑝
=
𝐹𝐼
𝐹𝑉 𝑚
𝐹𝐼
𝐹𝐷 𝑝
=
𝐹𝐼
𝐹𝐷 𝑚
𝐹𝐼
𝐹𝐺 𝑝
=
𝐹𝐼
𝐹𝐺 𝑚
8.7

The ratio of inertial force to viscosity force is

4 2 2
𝐹𝐼 𝜌 𝐷𝐼 𝑁 𝜌 𝐷𝐼 𝑁
= = = 𝑁 𝑅𝑒𝑖 8.8
𝐹𝑉 𝜇 𝑁 𝐷𝐼
2
𝜇
which is the Reynolds number. Similarly,
4 2 3 5
𝐹𝐼 𝜌 𝐷𝐼 𝑁 𝜌 𝑁 𝐷𝐼 1
= = = 8.9
𝐹𝐷 𝑃 𝑚𝑜 / 𝐷 𝐼 𝑁 𝑃 𝑚𝑜 𝑁𝑃
4 2 2
𝐹𝐼 𝜌 𝐷𝐼 𝑁 𝐷𝐼 𝑁
= = = 𝑁 𝐹𝑟 8.10
𝐹𝐺 3
𝜌 𝐷𝐼 𝑔 𝑔

6
Continued…

Dynamic similarity is achieved when the values of the nondimensional

parameters are the same at geometrically similar locations.
( 𝑁 ¿ ¿ 𝑅𝑒𝑖) 𝑃 =( 𝑁 𝑅𝑒𝑖 )𝑚 ¿
( 𝑁 ¿ ¿ 𝑃 )𝑃 =( 𝑁 𝑃 )𝑚 ¿ 8.11

( 𝑁 ¿ ¿ 𝐹𝑟 ) 𝑃 =( 𝑁 𝐹𝑟 )𝑚 ¿

Therefore, using dimensionless parameters for the correlation of data has

advantages not only for the consistency of units, but also for the scale-up
purposes.
However, it is difficult, if not impossible, to satisfy the dynamic
similarity when more than one dimensionless group is involved in a system,
which creates the needs of scale-up criteria.

7
Example 8.1:

The power consumption by an agitator in an unbaffled vessel can be

expressed as

( )
2 2
𝑃 𝜌 𝑁𝐷 𝐼 𝑁 𝐷 𝐼
=𝑓 , 8.12
3
𝜌 𝑁 𝐷𝐼
5
𝜇 𝑔

Can you determine the power consumption and impeller speed of a 10,000-
gallon fermenter based on the findings of the optimum condition from a
geometrically similar one-gallon vessel? If you cannot, can you scale up by
using a different fluid system?

Solution:
Since , the scale ratio is,
( 𝐷 𝐼 )𝑝 1/ 3
=1000 =10 8.13
( 𝐷 𝐼 )𝑚
8
Continued…

To achieve dynamic similarity, the three dimensionless numbers for the

prototype and the model must be equal, as follows:

( 𝑃 𝑚𝑜
𝜌 𝑁
3
𝐷 ) (
5
𝐼 𝑝
=
𝑃 𝑚𝑜
𝜌 𝑁
3
𝐷 )
5
𝐼 𝑚
8.14

( 𝜌 𝑁𝐷2
𝜇
𝐼
) ( 𝑝
=
𝜌 𝑁𝐷 2
𝜇
𝐼
) 𝑚
8.15

( 𝑁2 𝐷𝐼
𝑔 ) ( 𝑝
=
𝑁 2 𝐷𝐼
𝑔 ) 𝑚
8.16

If you use the same fluid for the model and the prototype, and . Canceling
out the same physical properties and substituting Eq. (8.14) to Eq. (8.16)
yields,

9
Continued…

( )
3
5 𝑁𝑝
( 𝑃 𝑚𝑜 )𝑝 =10 ( 𝑃𝑚𝑜 )𝑚 8.17
𝑁𝑚
The equality of the Reynolds number requires

𝑁 𝑃 =0.01 𝑁 𝑚 8.18

On the other hand, the equality of the Froude number requires

1
𝑁 𝑃 = 𝑁𝑚 8.19
√ 10

which is conflicting with the previous requirement for the equality of the
Reynolds number. Therefore, it is impossible to satisfy the requirement of
the dynamic similarity unless you use different fluid systems.

10
Continued…

If and , to satisfy Eqs. (8.15) and (8.16), the following relationship must
hold.

( 𝜇
𝜌 ) 𝑚
=
1
31.6 ( 𝜇
𝜌 ) 𝑝
8.20

Therefore, if the kinematic viscosity of the prototype is similar to that of

water, the kinematic viscosity of the fluid, which needs to be employed for
the model, should be 1/31.6 of the kinematic viscosity of water. It is
impossible to find the fluid whose kinematic viscosity is that small. As a
conclusion, if all three dimensionless groups are important, it is impossible
to satisfy the dynamic similarity.
As an example, for a fully baffled vessel when > 10,000, the power number
is constant. For a geometrically similar vessel, the dynamic similarity will be
satisfied by,

11
Continued…

( 𝑃 𝑚𝑜
𝜌 𝑁
3
𝐷
5
𝐼
) (
𝑝
=
𝑃 𝑚𝑜
𝜌 𝑁
3
𝐷
5
𝐼
)
𝑚
8.21

If the fluid employed for the prototype and the model remains the same,
the power consumption in the prototype is

)[ ]
5

(
3
𝑁𝑝 ( 𝐷 𝐼 )𝑚
( 𝑃 𝑚𝑜 )𝑝 =( 𝑃 𝑚𝑜)𝑚 8.22
𝑁𝑚 ( 𝐷𝐼)𝑝

where is equal to the scale ratio.

With a known scale ratio and known operating conditions of a model,
we are still unable to predict the operating conditions of a prototype
because there are two unknown variables, and N. Therefore, we need to
have a certain criteria which can be used as a basis.

12
8.2 Scaling of Mass-Transfer Equipment
 As discussed by Oldshue, the various quantities which may influence the
product in an agitated industrial reactor do not scale in the same way
with reactor size or impeller rate.
1. The turbulent Reynolds number , determines and thus bubble mass
transfer coefficients
𝜌 𝑢𝑟𝑚𝑠 𝐷 𝜌 𝐷 𝐷 𝑃
( )
1/ 3
Re 𝑡 = ∝ 8.23
𝜇 𝜇 𝜌 𝑉
2. The impeller tip velocity , determines the maximum shear rate , which
in turn influences both maximum stable bubble or microbial floc size
and damage to viable cells.
3. The power input per unit volumethrough , determines mass-transfer
coefficients and particulate sizes. In laminar and transition regimes of
aerators
8.24

13
Continued…

Figure 8.1: Different relationships among air bubbles (B), n-hexadecane droplets
(O), and yeast cells (C), at different stages of batch culture of petrophilum (1 = log
phase, 2 = first half phase, 3 = second half of exponential phase, 4 = after n-
hexadecane exhaustion).
14
Continued…

For turbulent regimes, the power number is constant; thus

8.25
Then taking to scale with gives
laminar, transition aeration
8.26
turbulent aeration
4. The power input during aeration is
8.27
Thus,

[ ]
0.45
3 0.44
𝑃𝑎 𝑃 (𝑁𝑖 𝐷 )
2
′ 𝑖 8.28
≈𝑚
𝑉 𝑉 2.22 𝑁 0.56
𝑎

which will determine the motor size needed during fermentation.

15
Continued…

5. If the vessel liquid is well mixed internally, a characteristic circulation

time exists. The liquid recirculation flow rate , through the impeller
region varies as a cross-sectional area and tank average impeller
velocity varies as Thus
3
Fl Ni Di −1
∝ 3
= N i time 8.29
V Di
a quantity of importance since it is inversely proportional to the time
that fluid may spend away from the homogenizing influence of the
impeller.\
Here by “basis for scale-up” we mean the quantity which, by choice of
operating conditions in the larger unit, will be maintained at the same value
as in the smaller scale unit.
For example, if we scale-up on the basis of constant power per unit volume,
for mechanical agitators3in the turbulent
2 regime
3 this2means
N i 1 D i 1= N i 2 D i 2 8.30

16
Continued…

where 1 and 2 denote the values in the small and large scale vessel,
respectively.
As indicated in Fig. 8.2a this gives very similar yields of penicillin for
vessels from 5 liters to 200 gallons. However, from the curves in Fig. 8.2a,
we see that, at different P/V values, there are significant differences in yield
at different scales.
Another frequently applied basis for scale-up is constant volumetric
transfer coefficient . Fig. 8.2b shows vitamin yields from bacterial
fermentations at different scales versus corresponding values of ().
The total pressure is included here to correct for the greater driving force
for oxygen transfer at higher pressures which are encountered in large-scale
bioreactors.

17
Continued…

Figure 8.2: (a) Penicillin yields vs. power input in various sized vessels
(After E Gaden, Super Sanita, vol. 1, p. 61, 1961). (b) Vitamin B12 yield
(ug/g) vs. mass-transfer group (

18
Continued…

If we scale up on one basis, we must be aware that other mixing and

flow properties are different. This point is dramatically illustrated in an
example of Oldshue which considers scale-up from an 80 L to a 10,000 L
agitated bioreactor. Here, increases fivefold and V increases by a factor of
125.
We must try to select as a scale-up basis the transport property most
critical to the performance of the bioprocess. If the bulk liquid composition
is uniform and the bubbles are uniformly dispersed throughout the vessel,
the mass-transfer rate is simply
∗
𝑉 k l 𝑎(𝑐 −𝑐 𝑙𝑖𝑞 )𝑒=𝑜𝑥𝑦𝑔𝑒𝑛𝑐𝑜𝑛𝑠𝑢𝑚𝑝𝑡𝑖𝑜𝑛𝑟𝑎𝑡𝑒 ,𝑚𝑜𝑙 𝑂 2 /𝑠 8.31

where subscript e refers to gas exit compositions.

19
Continued…

When the bubbles rise in plug flow through the vessel but the impeller
still maintains perfect mixing in the liquid phase, varies with position. Over
a differential reactor height dz, the instantaneous loss of oxygen from the
bubble is
𝑑𝑝 𝑂 1
𝐻 𝐴 𝑑𝑧 2
=
𝑔𝑎𝑠 𝑣𝑜𝑙
(
𝑑𝑖𝑓𝑓𝑟𝑒𝑛𝑡𝑖𝑎𝑙
)(
𝑐𝑜𝑛𝑐 . 𝑟𝑎𝑡𝑒 𝑜𝑓
𝑑𝑡 𝑅𝑇 𝑟𝑒𝑎𝑐𝑡𝑜𝑟 𝑣𝑜𝑙 𝑟𝑒𝑎𝑐𝑡𝑜𝑟 𝑣𝑜𝑙𝑢𝑚𝑒 𝑐h𝑎𝑛𝑔𝑒𝑖𝑛𝑏𝑢𝑏𝑏𝑙𝑒 ) 8.32

which equals
− k l 𝑎(𝑐 ¿ ¿ ∗− 𝑐 𝑏) 𝐴𝑑𝑧 ¿
the mass transfer rate into the liquid. Since (M is Henry’s law constant), we
have
∗
𝐻 𝑑𝑝 𝑂 𝐻𝑀 𝑑 𝑐 𝑙 ∗
= = − 𝑎 k l (𝑐 𝑙 − 𝑐 𝑏)
2
8.33
𝑅𝑇 𝑑𝑡 𝑅𝑇 𝑑𝑡

20
Continued…

For constant bubble-rise velocity , , and the variation seen to be

∗
( 𝑐𝑙 − 𝑐 𝑏) 𝑧 − 𝑎 k l 𝑅𝑇 𝑧 8.34
or 𝑙𝑛 ∗
=
(𝑐 𝑙 − 𝑐 𝑏)𝑖𝑛𝑙𝑒𝑡 𝐻𝑀 𝑢𝑏

(
(𝑐 ∗𝑙 − 𝑐𝑏 ) 𝑧 =( 𝑐∗𝑙 − 𝑐 𝑏)𝑖𝑛𝑙𝑒𝑡 exp
The overall mass-transfer rate in the volume Ah is therefore
)− 𝑎 k l 𝑅𝑇 𝑧
𝐻𝑀 𝑢𝑏
8.35

[ ( )]
h
𝐻𝑀 𝑢 𝑏 ∗ − 𝑎 k l 𝑅𝑇 h
∫ 𝑎 k l ( 𝑐 𝑙 −𝑐 𝑏 ) ( 𝑧 ) 𝐴 𝑑𝑧 =
∗
𝑅𝑇
( 𝑐 𝑙 −𝑐 𝑏 )𝑖𝑛𝑙𝑒𝑡 𝐴 1 − exp
𝐻𝑀 𝑢𝑏
0

8.36

21
8.3 Criteria for Scaling-Up Fermentors
Few steps are available for the so-called “scale-up” of fermentors,
starting from the glass apparatus used in fundamental research
investigations. Usually, chemical engineers are responsible for building and
testing pilot-scale fermentors (usually with capacities of ), and subsequently
for designing industrial-scale fermentors based on data acquired from the
pilot-scale system.
One criterion for scaling-up this type of bioreactor is the power input per
unit liquid volume of geometrically similar vessels, which should be
proportional to for the turbulent range and to for the laminar range, where
is the rotational stirrer speed and L is the representative length of the
vessel.
In order to minimize any physical damage to the cells, the product of the
diameter and the rotational speed of the impeller - which should be
proportional to the tip speed of the impeller and hence to the shear rate at
the impeller tip - becomes an important criterion for scale-up.
22
 Aerated stirred tanks, bubble columns, and airlifts are usually used for
aerobic fermentations. One criterion of scaling-up aerated stirred tank
fermentor is , approximate values of which can be estimated. For the
turbulent range, a general correlation for in aerated stirred fermentors is of
the following type:
𝑚 3 2 𝑛
𝑘 𝐿 𝑎=𝑐 𝑈 𝐺 (𝑁 𝐿 ) 8.37

where is the superficial gas velocity over the cross-sectional area of the
tank, is the rotational stirrer speed, and is the representative length of
geometrically similar stirred tanks.
For the turbulent regime () should be proportional to the power requirement
per unit liquid volume.

23
Continued…

The main operating factor of a bubble column fermentor is the superficial

gas velocity over the column cross section, which should be kept equal in
scaling up.
in bubble columns cm in diameter will increase with the column
diameter to the power of . As this trend levels off with larger columns, it is
recommended that values estimated for a cm column are used.
If heat transfer is a problem, then heat transfer coils within the column,
or even an external heat exchanger, may become necessary when operating a
large, industrial bubble column-type fermentor.

24
Continued…

Example 8.2: Two geometrically similar stirred tanks with flat-blade turbine
impellers of the following dimensions are to be operated at 30°C as pilot-
scale and production-scale aerobic fermentors.

Scale Tank diameter and Impeller Liquid volume ()

liquid depth (m) diameter (m)
Pilot 0.6 0.24 0.17
Production 2.0 0.8 6.28

Satisfactory results were obtained with the pilot-scale fermentor at a

rotational impeller speed N of and air rate () of . The density and viscosity
of the broth are and , respectively. Data from the turbine impellers showed
that can be correlated by Equation 8.37, with values .

25
Continued…

Using as the scale-up criterion, estimate the impeller speed and the air rate
for the production-scale fermentor that will give results comparable with
the pilot-scale data. 𝑚 3 2 𝑛
𝑘 𝐿 𝑎=𝑐 𝑈 𝐺 ( 𝑁 𝐿 )
Solution:
In the pilot-scale fermentor:.
At equal , the air rate to the production-scale fermentor should be

0.0295 × 60 × ( )
𝜋
4
×(2.0)2=5.56 𝑚3 𝑚𝑖𝑛−1
With the pilot fermentor .
With this equal value of and for the production fermentor, the production
fermentor should be operated at

1.215
3
𝑁 = = 0.304 𝑁 =0.672 𝑠 −1
4

26
Continued…

Incidentally, the impeller tip speeds in the pilot and production fementors
are calculated as and , respectively.

27
8.4 Design estimation of various Scale-
up Parameters
Most often, power consumption per unit volume is employed as a criterion
for scale-up. In this case, to satisfy the equality of power numbers of a
model and a prototype,

) ( )[ ]
2

( ) (
3
𝑃 𝑚0 𝑃 𝑚0 𝑁𝑝 ( 𝐷𝐼)𝑝
= 8.39
𝐷
3
𝐼 𝑝 𝐷
3
𝐼 𝑚
𝑁𝑚 ( 𝐷 𝐼 )𝑚

Note that represents the power per volume because the liquid volume is
proportional to for the geometrically similar vessels. For the constant ,

[ ]
2

( )
3
𝑁𝑝 ( 𝐷 𝐼 )𝑚 8.40
=
𝑁𝑚 ( 𝐷 𝐼 )𝑝

28
Continued…

As a result, if we consider scale-up from a 20-gallon to a 2,500gallon

agitated vessel, the scale ratio is equal to 5, and the impeller speed of the
prototype will be,

[ ]
2 /3
( 𝐷 𝐼 )𝑚
𝑁 𝑝= 𝑁𝑚
( 𝐷𝐼 )𝑝
𝑁 𝑝 =0.34 𝑁 𝑚 8.41

which shows that the impeller speed in a prototype vessel is about one
third of that in a model.
For constant , the Reynolds number and the impeller tip speed cannot
be the same. For the scale ratio of 5,
( 𝑁 𝑅𝑒 )𝑝 =8.5 ( 𝑁 𝑅𝑒 )𝑚
𝑖 𝑖
8.42

( 𝑁 𝐷 𝐼 )𝑝 =1.7 ( 𝑁 𝐷 𝐼 )𝑚 8.43

29
Continued…

Table 8.1 shows the values of properties for a prototype (2,500 gallon)
when those for a model (20-gallon) are arbitrarily set as 1.0 (Oldshue,
1966).
The parameter values of the prototype depend on the criteria used for
the scale-up. The third column shows the parameter values of the
prototype, when is set constant. The values in the third column seem to be
more reasonable than those in the fourth, fifth, and sixth columns, which
are calculated based on the constant value of , and , respectively.
For example, when the Reynolds number is set constant for the two
scales, the reduces to 0.16 percent of the model and actual power
consumption also reduces to 20 percent of the model, which is totally
unreasonable.

30
Continued…

As a conclusion, there is no one scale-up rule that applies to many different

kinds of mixing operations. Theoretically we can scale up based on
geometrical and dynamic similarities, but it has been shown that it is
possible for only a few limited cases. However, some principles for the
scale-up are as follows (Oldshue, 1985):
1. It is important to identify which properties are important for the optimum
operation of a mixing system. This can be mass transfer, pumping
capacity, shear rate, or others. Once the important properties are
identified, the system can be scaled up so that those properties can be
maintained, which may result in the variation of the less important
variables including the geometrical similarity.
2. The major differences between a big tank and a small tank are that the
big tank has a longer blend time, a higher maximum impeller shear rate.

31
Continued…

3. For homogeneous chemical reactions, the power per volume can be

used as a scale-up criterion. As a rule of thumb, the intensity of agitation
can be classified based on the power input per 1,000 gallon as shown in
Table 8.1.

Table 8.1: Properties of Agitator on scale up

32
Continued…

4. For the scale-up of the gas-liquid contactor, the volumetric mass-transfer

coefficient kLa can be used as a scale-up criterion. In general, the
volumetric mass-transfer coefficient is approximately correlated to the
power per volume. Therefore, constant power per volume can mean a
constant kLa.
5. Typical impeller-to-tank diameter ratio for fermenters is 0.33 to 0.44. By
using a large impeller, adequate mixing can be provided at an agitation
speed which does not damage living organisms. Fermenters are not
usually operated for an optimum gas-liquid mass transfer because of the
shear sensitivity of cells.

Table 8.2: Criteria

of Agitation
Intensity

33
8.5 Power Estimation for Ungassed Liquids

There are well-established empirical correlations for stirrer power

requirements. Figure 7.8 is a log-log plot of the power number N, for
ungassed liquids versus the stirrer Reynolds number (Re). These
dimensionless numbers are defined as follows:
𝑃
𝑁 𝑃 = 3 5 8.44
(𝜌 𝑁 𝑑 )
2
𝑁 𝑑 𝜌
𝑅𝑒 = 8.45
𝜇
where is the power required , is the number of revolutions of the impeller
per unit time , d is the impeller diameter , p is the liquid density , and is the
liquid viscosity . Since the product disproportional to the peripheral speed
(i.e., the tip speed of the rotating impeller).

34
Continued…

Figure 8.6: Correlation between Reynolds number (Re) and Power number ().
Curve (a): six-flat blade turbine, four baffles,
curve (b): two-flat blade paddle, four baffles, =0.1 D,
curve (c): three-blade marine propeller, four baffles, =0.1 D.

35
Continued…

In Figure 8.6, the curves a, b, and c correlate data for three types of
impellers, namely, the six-flat blade turbine, two-flat blade paddle, and
three-blade marine propeller, respectively. It should be noted that, for the
range of (Re) >, is independent of (Re). For this turbulent regime it is clear
from Equation 8.44 that
3 5
𝑃 = 𝑐1 𝜌 𝑁 𝑑 8.46

where , is a constant that varies with the impeller types. Thus, for a given
type of impeller varies in proportion to , and liquid density ,but is
independent of liquid viscosity .
For the ranges of Re below approximately 10, the plots are straight lines
with a slope of ; that is, , is inversely proportional to . Then, for this laminar
regime, we can obtain Equation 7.33 from Equations 8.44 and 8.45:

𝑃 = 𝑐2 𝜇 𝑁
2
𝑑
3 8.47

36
Continued…

Thus, for the laminar regime varies in proportion to liquid viscosity , , and
to a constant , which varies with impeller types, although is independent of
the liquid density .
It is worth remembering that the power requirements of geometrically
similar stirred tanks are proportional to in the turbulent regime and to in
the laminar regime.
In the case of scaling-up of geometrically similar stirred tanks, the equal
power input per unit liquid volume, which should be proportional to for the
turbulent regime and to for the laminar regime, is sometimes used as a
criterion for scale-up, because it leads to the same values of in a larger
stirred tank with that in the smaller stirred tank.

37
8.6 Power Estimation for Gas-Sparged Liquids

The ratio of the power requirement of gas-sparged (aerated) liquid in a

stirred tank,, to the power requirement of ungassed liquid in the same
stirred tank, , can be estimated using Equation 8.48.
This is an empirical, dimensionless equation based on data for six-flat
blade turbines, with a blade width that is one fifth of the impeller
diameter , while the liquid depth , is equal to the tank diameter. Although
these data were for tank diameters up to , Equation 8.48 would apply to
larger tanks where the liquid depth-to-diameter ratio is typically in the
region of unity.

( ) ( ) ( ) ( )
0.115 2 1.96 (𝑑/ 𝐷)

( )
𝑃𝐺 𝑑
4.38 2
𝑑 𝑁 𝑑𝑁 𝑄
log =−192 8.48
𝑃0 𝐷 𝑣 𝑔 𝑁𝑑
3

where is the impeller diameter (L), is the tank diameter (L), is the
rotational speed of the impeller (), g is the gravitational constant (), Q is the
gas rate (), and is the kinematic viscosity of the liquid ().

38
Continued…

The dimensionless groups include () Reynolds number (Re);

Froude number (Fr);
aeration number (Na),
which is proportional to the ratio of the superficial gas velocity with respect
to the tank cross section to the impeller tip speed.
The ratio for flat-blade turbine impeller systems can also be estimated by

( ) ( ) ( blade width) (). All other

𝑃𝐺 𝑄
− 1/ 4
𝑁 𝑑
2 4 −1 /5
log =0.10 8.49
𝑃0 𝑁𝑉 2
where V is the liquid volume () and is the impeller 𝑔𝑏 𝑉 3

symbols are the same as in Equation 8.48.

39
Continued…

Example 8.3: After a batch fermentation, the system is dismantled and

approximately 75% of the cell mass is suspended in the liquid phase (2 l),
while 25% is attached to the reactor walls and internals in a thick film (ca.
0.3 cm). Work with radioactive tracers shows that 50% of the target product
(intracellular) is associated with each cell fraction. The productivity of this
reactor is 2 g product/l at the 2 l scale. What would be the productivity at
20,000 l scale if both reactors had a height-to-diameter ratio of 2 to 1?
Solution:
Both tanks are geometrically similar, so we can calculate the diameter and
the resulting surface area and volume in both tanks
1 1 1
𝑉 = 𝜋 𝐷 . 𝐻 = 𝜋 𝐷 .2 𝐷= 𝜋 𝐷 3
2 2
4 4 2
2
𝑆=𝜋 𝐷 . 𝐻 =𝜋 𝐷 . 2 𝐷=2 𝜋 𝐷

40
Continued…

For the system

For the systems,

At the bench scale the amount of product made by surface-attached cells is

, for of surface area

41
Continued…

The amount of product formed at due to surface-attached growth is

2
342600 𝑐𝑚
× 2 𝑔= 928 𝑔
738 𝑐𝑚2
The overall yield in the system is
𝑔
2 × 2 𝑙= 4 𝑔
𝑙
The overall yield in the system is

928 𝑔+ 2 ( 𝑔 1
)
. .20000 𝑙 =20928 𝑔
𝑙 2
If no wall growth had been present, the tank would have yielded . Thus,
wall growth, if present, can seriously alter the productivity of a large-scale
system upon scale-up.

42
Continued…

Example 8.4: Consider the scale-up of a fermentation from a to vessel. The

small fermenter has a height-to-diameter ratio of . The impeller diameter is
30% of the tank diameter. Agitator speed is 500 rpm and three Rushton
impellers are used. Determine the dimensions of the large fermenter and
agitator speed for:
a. Constant P/V
b. Constant impeller tip speed
c. Constant Reynolds number Assume geometric similarity

Solution:
Assume the vessel is a cylinder.
Thus,
𝑉 = ( 𝜋
4 ) 𝐷 𝑡2 𝐻

43
Continued…

but , so
3 3
𝑉 =( 𝜋 / 4 )3 𝐷 𝑡 =10,000 𝑐𝑚
Solving for , , and gives:

The scale-up factor is the cube root of the ratio of tank volumes, or

( )
1/ 3
10000 𝑙
=10
10𝑙
To maintain geometric similarity the larger vessel will have its dimensions
increased by a factor of 10. That is:

44
Continued…

a. For constant to apply must be the same in both vessels. Let subscript
refer to the small vessel and to the large vessel.

( )
2/ 3

( )
2 /3
𝐷𝑡1 1
𝑁 2=𝑁 1 =500 𝑟𝑝𝑚 =107 𝑟𝑝𝑚
𝐷𝑡2 10
b. For constant tip speed to apply must be the same in both vessels.

𝑁 2= 𝑁 1
( )
𝐷𝑡1
𝐷𝑡2
=500 𝑟𝑝𝑚
1
10 ( )
=50 𝑟𝑝𝑚

c. For constant Re to apply must be the same in both vessels.

( )
2
𝐷𝑡1
( ) =5 𝑟𝑝𝑚
2
1
𝑁 2= 𝑁 1 =500 𝑟𝑝𝑚
𝐷𝑡2 10
Scale-up on the basis of constant would be the most likely choice unless
the culture was unusually shear sensitive. Scale-up based on constant Re
would not be used.

45