Microsoft Word - Contents.doc


HUNGARIAN JOURNAL 
OF INDUSTRIAL CHEMISTRY 

VESZPRÉM 
Vol. 34. pp. 35-39 (2006) 

OPTIMIZATION PROBLEMS OF FERMENTOR AERATION-AGITATION SYSTEM 

L. KOZMA1, L. NYESTE2 and A. SZENTIRMAI3 

1Chartered engineer, former leader of the Technical Division of BIOGAL Pharmaceuticals, 
Postal address: Őrség u. 17., H-1112 Budapest, HUNGARY, E-mail: multiplan@axelero.hu 

2Professor of Department of Biochemical Engineering, Budapest University of Technology, Szt. Gellért tér 4., 
H-1521 Budapest, HUNGARY 

3Professor of Microbiology and Biotechnology, University of Debrecen, Haraszt u. 21., H-4010 Debrecen, HUNGARY 
 

After four decades of almost exclusive application of Rushton turbines a number of types of impellers were developed for 
the fermentation industry in the last 20 years which were proven to be more efficient despite their lower power 
requirements. 

The efficiency of an impeller is affected strongly by the degree of their hydromechanical properties corresponding 
with the specific characteristics and requirements of a certain fermentation process. Evaluation in case of non-Newtonian 
broths cannot be carried out with proper accuracy, and their optimization becomes difficult. 

An increase in yield even by a few percent can be of great importance in large-scale fermentation vessels. 
Consequently, the optimization of agitation system is a very important factor but it is only partially provided by scale-up 
using pilot plant data and similarity criteria. This is the reason why we need newer methods for optimization of aeration-
agitation systems for large-scale fermentation vessels by agitators equipped with changeable flow modifying parts. 

Keywords: power number, non-Newtonian broth, Rushton turbine, shear power, scale-up, flooding 

Advantages and Disadvantages of Newer Agitation 
Systems 

The disadvantages of flat-blade Rushton turbines – less 
axial circulation capability and large power requirement 
- applied in penicillin fermentation for many years due 
to their excellent dispersion capability became more and 
more obvious with the increase in size of fermentation 
vessels. Due to these disadvantages newer types of 
impellers and complex agitation systems were developed. 
 
Turbines 

The efficiency of Rushton turbines (Fig. 1) was 
increased by the application of impellers with parabolic 
profile instead of flat-blades in Scaba 6SRGT system 
(Fig. 2). Their power number has decreased from 5,5-
6,0 to 3,2. The number of higher energy peaks around 
the impeller endangering more sensitive microorganisms 
decreased and by increasing impeller’s diameter larger 
volumes could be blended with the same power 
requirements. The impellers’ sensitivity to “flooding” 
phenomenon and increased viscosity had decreased, but 
still they had lower circulation capability. In the event 
of applying more impellers compart-mentalization 
(inadequately blended areas) may occur. That is why in 
newer agitation systems these are used only at the 
lowest, dispersion level. (A. Baker et al. 1) 

Closed turbines have better circulation capability but 
these are rarely applied due to their little dispersion 
capability. 

Propeller agitators 
Propeller agitators have excellent axial circulation 

properties but weak dispersion capability. Due to their 
little power number of 0,5-1,1 their diameter could be 
larger with the same power requirement and this 
facilitates full blending of viscous broths and filamentous 
microorganisms. 

First EKATO has developed propeller type impellers 
with pitched blades called MIG and INTERMIG (Fig. 
3). Recently streamlined propeller agitators with twisted 
surface have been applied at the upper levels of 
complex agitation systems. Impellers with larger 
diameter ratios of 1:0,5-1:0,6 narrower blades such as 
Lightnin 310 are applied for blending of lower viscosity 
broths. Impellers with less diameter and diameter ratio 
of 1:0,45 and broader blades which have power number 
of 1,0-1,1 such as Lightnin A 315 and Prochem Maxflo 
(Fig. 4 and Fig. 5) are applied for blending of higher 
viscosity broths. 

According to A.W. Nienow propeller agitators 
provide better blending efficiency for both the lower 
and higher viscosity broths and better mass and heat 
transfer than Rushton turbines. Other advantages of 
these agitators are their power number and indulgence 
with sensitive microorganisms (A.E. Nienow 2.) 
 
Vacuum agitators 

Vacuum agitators has low power requirement, good 
dispersion capability but lesser circulation properties. 
These types can be used for blending less air volumes. 



 

 

36

They are used only in certain technological processes 
such as in flotation devices and in yeast production. 

Newer Complex Agitation Sysems 

Merely the last few decades the researchers and 
manufacturers have realized that the efficiency if 
agitation systems can be increased by development of 
complex systems including more agitators of different 
types and properties which satisfy better the 
requirements of the particular levels. Despite many 
published paper literature dealt not so much with the 
problems of agitation of large-scale fermentation 
vessels. Perhaps on the Symposium in Firenze in 1993 
data on the mass transfer problems due to differences 
between the levels of large-scale fermentation vessels 
were published for the first time. These differences are 
stemming partially from the position of levels and 
partially from their different functions i.e. could be local 
or functional differences. 

Local differences are mainly caused by the pressure 
differences due to 8-12 m height of fermentors and this 
may affect bubble size and the density of foaming 
broths, etc.  

Functional differences are because the function of 
the lowest agitator is efficiently disperse air input, the 
function of the middle agitator(s) is the best intensity 
circulation of the broth-air mixture and the function of 
the upper agitator is recirculation of the foaming broths 
on the surface with less further foam formation possible. 

In the nineties the increase in differences due to 
larger and larger-scale fermentation vessels had led to 
the development of complex agitation systems 
considering the differences between levels. 6SRGT 
modified turbine agitator with good dispersion 
capability is generally used on the lowest level, and high 
efficiency propeller agitators e.g. Lightnin, Prochem are 
applied on the highest level. (K. Myers, 3.) These 
complex systems have better energy dissipation, 
dispersion and circulation capabilities, they are more 
efficient and more sensitive to flooding than the older 
systems built from components of the same type and 
size. 

Optimization Problems of Agitation Systems 

Sizing and development of large-scale agitation systems 
is still based mainly on data from and experiences with 
pilot plant fermentors, and relations developed through 
the theory of similarity and dimension analysis. Lately 
industrial measurements have been used more and more 
often.  

The application of the results of experimental 
measurements during scale-up is limited very much by 
the significant differences in the hydrodynamic fields 
and flow patterns of large-scale fermentors mainly due 
to the following reasons: 
a) because of the nearness of the baffles and impellers 

the large velocity gradient between the flowing 

layers in the experimental device results in large 
shear velocity and shear power, while in large-scale 
devices the much less velocity gradient due to larger 
sizes results less values. 

b) unlike in the large-scale vessels no free turbulence 
facilitating mass transfer is evolved because of the 
less Reynolds number value due to the less size of 
the experimental device.  

c) flow of high viscosity broths can slow down so 
much in the large-scale vessels that inadequately 
mixed areas are formed even when newer agitation 
systems are applied. No secondary dispersion can 
occur along the baffles which may mask the 
deficiencies of the impeller type itself in pilot plant 
fermentors. 

d) Due to the high pressure of large-scale fermentation 
vessels bubble size affecting oxygen transfer is 
decreased, solubility of gases, the density of liquid-
gas mixture and coalescence of bubbles are 
increased. 

e) In large-scale vessels the agitation time and cross-
sectional air flow velocity are increased with the 
same specific air volume and v/v input. 
 
Kipke’s example can be cited for demonstration of 

the increase in agitation time, namely if a given 
agitation time can be produced in a laboratory fermentor 
of 5 liters with P/V = 1 kW/m3 power/unit volume, in a 
large-scale fermentation vessel of 50 m3 the same 
agitation time can only be achieved with 5000 kW 
power! The differences in magnitude show the problems 
of scale-up and the limitations of the application of 
experimental results. 

The scales are changed considerably during scale-up 
even during entirely proportioned geometric scale-up. 
For example, if the size of a model is increased only by 
tenfold, its surface increases hundred-fold but its 
volume increases thousand-fold. That is why even the 
name of similarity criteria is false since their application 
provides merely partial similarity. Due to the unequal 
change in size and value ratios, physical, geometric, 
kinetic and dynamic similarity criteria cannot be 
selected simultaneously. 

Due to the lack of a generally valid procedure many 
scale-up processes had been developed. 
The most often is to rely on power requirement per 
volume (P/V), volumetric oxygen transfer coefficient 
(kLa), gas-holdup and shear stress (viscosity/velocity 
gradient). 

The variation of chosen considerations may lead to 
great differences. That is why many researchers’ opinion 
is that results do not comply with the technical and 
economical requirements of biotechnology and can only 
be informative data for developers of sizing procedures. 
According to M. Charles: “in practice, scale-up strategies 
tend to be »mixed bags« engendering art enpricism, 
conventional wisdom and (frequently) wishful thinking” 
(4). 

For the optimization of fermentation process i.e. for 
the achievement of largest possible yields even 
distribution of the dissolved oxygen (DO), medium and 
ingredients added during fermentation and optimum 



 

 

37

mass transfer conditions should be provided besides 
application of high productivity microorganisms and 
adequate mediums. 

Adequate oxygen level can be achieved by both 
proper air volume input and its best possible dispersion 
i.e. the least oxygen bubbles and their most even 
distribution. To achieve this, adequate agitation power, 
air volume and an agitation system is necessary which is 
suitable for effective dispersion of air, for creation of 
intensive circulation and for even distribution of bubbles. 

The level of dissolved oxygen (DO) can be measured 
during fermentation and can be adjusted by the regulation 
of power input and/or air volume – if there are adequate 
quantities. 

Considering the sometimes high values e.g. in case 
of penicillin fermentation the efficiency of the process is 
a significant factor and it is affected by the structure of 
the agitation system besides the power input and 
adequate air volume. The problem is that however, we 
can calculate – at least approximately - the diameter and 
power requirement of the agitators and the air volume 
by the available procedures and relations, these data 
provide very little information on optimum design 

Similarly to the added nutrients oxygen transfer 
occurs on the interfaces of air bubbles and medium 
particles and through the walls of microorganisms’ 
cells. According to the double layer theory thinning of 
the laminar layers on the interfaces by creating turbulent 
liquid flow and shear stress due to this turbulence is 
necessary for the acceleration of mass transfer. 

It is well known that vortexes are arisen during real 
liquid flow due to their viscosity and because of the 
collision of these vortexes turbulence proportional to the 
velocity of flow occurs. Shear stresses proportional to 
the velocity of flow occur between turbulent liquid 
layers which have important role in oxygen (DO) and 
mass transfer: these stresses thin the laminar layers of 
transferring interfaces, micromix the components of 
broths, disperse oil particles and air bubbles facilitating 
and accelerating mass transfer processes, disintegrate 
clots and in some cases cause morphological changes in 
the structure of microorganisms as in the case of 
penicillin fermentation. 

The magnitude of hydrodynamic forces created by 
agitation can be seen from the fact that according to the 
calculations of Van’t Riet and Smith the centrifugal 
acceleration behind the vortexes created by impellers 
can be seven-hundredfold of the gravity (5). 

Shear powers may, however, damage microorganisms 
which are especially sensitive, contribute to the creation 
of stable liquefied foams which generally decrease 
oxygen transfer, and aeration of carbon-dioxide and 
other gases partially on direct way and partially due to 
antifoaming oils. 
Microorganisms on the interface of vortexes can be 
disrupted while those in the centre of the vortex may 
abrade each other. 

Consequently the intensity of agitation should remain 
within a narrow range for keeping damaging effect at a 
minimum level while maintaining maximum advantages 
and this is the purpose of optimization. 

The characteristics of fermentation processes may 
vary due to the differences in viscosity, foaming 
properties, density, etc. A typical feature is that while 
foaming generally decreases the oxygen transfer, in 
certain cases the increased persistence of bubble may 
raise the rate of oxygen transfer in liquefied foams, and 
antifoaming agents may decrease it. 

Some microorganisms such as oxytetracycline 
producers do not need agitation, and their fermentation 
can be made in slim vessels without agitator which are 
much cheaper. 

Air inflated into the broths dispersed, distributed and 
circulated in the medium of the fermentor by the 
agitation system proportionally to power input. Due to 
this procedure the volume of the medium is increasing 
and oxygen will be dissolved in the medium depending 
on the intensity of agitation, characteristics of medium 
and surface gas velocity vs. The degree of oxygen 
transfer is depending on the viscosity of the medium the 
characteristics of air, medium and microorganism system, 
and coalescing properties of air bubbles. The entrapment 
of air and this way oxygen fusion can decrease greatly 
due to increased viscosity and bubble coalescence 
(Van’t Riet, Smith 5., and Buchholz et al. 6.) 

Besides the mentioned air entrapment broths volume 
can also be increased by the often very intensive foam 
formation depending on broths characteristics. Stable so 
called liquefied foams are formed on more viscous 
mediums such as in penicillin fermentation. Contrary to 
the air entrapment mentioned above this foam formation 
is detrimental since it limits oxygen transfer partially 
directly and partially through antifoaming agents, 
however rarely the opposite situation may occur. 
Consequently maintaining the air input and power 
within a narrow range based on continuous instrumental 
measurement of fermentation parameters is an important 
requirement for dissolving oxygen and nutrients and 
also their transfer to microorganisms with adequate rate. 

Considering economical importance of mass transfer 
problems arising from increasing size of fermentor 
vessels more efficient complex agitation systems were 
developed with lower power requirement, better 
dispersion and including far better circulation levels. In 
their paper published in 1987 B.C. Buckland et al. had 
revealed that application of Lightnin and Prochem 
propeller agitators providing better “top to bottom” 
blending of viscous broths is cost effective due to 
saving power input and/or by the application of these 
agitators the production can be increased because of the 
higher cell concentration due to better agitation 
(Buckland et al. 7). Papers on complex agitation systems 
have been published more often since the beginning of 
1990s (Chemineer, 8).  

During their developmental activities manufacturers 
besides the relations for calculation of main sizes and 
powers could mainly rely on experimental results 
which, however, provided merely informative results 
due to the above reasons. It should also be noticed that 
uniformization, development of systems which can be 
distributed widely and production of their own types 
and licensed products are the main interests of 
manufacturers. All these factors eventually lead to 



 

 

38

negligence of specific requirements of fermentations. 
Exerting themselves to protect their trade secrets, 
factories generally provide very little possibilities for 
carrying out profound studies fermentation process by 
the professionals of manufacturers. 

According to the above characteristics and 
requirements of fermentation processes may differ very 
much. It follows from the above written that besides 
applied technologies and materials the success of 
fermentation processes also depends upon whether the 
agitation system used during fermentation is adequate 
for the specific requirements of fermentation. 
Consequently the characteristics, dispersion and 
circulation capabilities of agitation levels should be 
adjusted to the features and requirements of the 
fermentation which may, conversely, vary because of 
the differences between the experimental and industrial 
levels. In case of viscous liquids experimental levels do 
not provide data and indications of adequate accuracy 
for the adjustment. Although the analysis of experimental 
data has been improved very much since V. Charles 
through the application of computers, lesser changes 
may also be of significance due to the large volume of 
industrial fermentors and these changes cannot be 
designed with adequate accuracy. 

It follows from the above that there is no 
adequate procedure available for actual optimization 
of industrial agitation systems and for establishment 
how much an agitation system can be considered 
optimal for a certain fermentation procedure. 
The efficiency of an agitation system is depending on its 
structure and considering fermentation it is depending 
on how the agitation system’s levels use power input for 
dispersion and circulation and how adequate this is for 
the requirements of a given fermentation process. 

A solution for this problem may be if manufacturers 
provide special separated parts for the particular levels 
of the agitation system which could be fixed on the 
system by screw this was changing the characteristics of 
agitation. It would not be especially difficult to solve 
since power input is proportionally changed with the 
fifth degree of the diameter of the agitator and the 
characteristic of flows can be modified within a wide 
range merely with changing the shape and angle of 
blades of the impeller. 

The application of this idea requires some change in 
viewpoint according to the following: 
1. It cannot be expected that a manufacturer will 

provide an “optimal” agitation system, but it is 
expected to provide an agitation system of which 
perfusion properties can be modified within a wide 
range with auxiliary parts. It would be, of course, the 
obligation of the manufacturer to provide detailed 
user manuals and information sheet for the 
expectable effects of these auxiliary parts and 
provide professional assistance for testing on 
demand. 

2. The obligation of the user would be the actual 
optimization of the agitation system according to 
provided directives and thorough analysis of the 
effects of the auxiliary parts. Inclusion of factory 
professionals in the selection of the most efficient 

system may solve the problems of scale-up sometime 
mentioned as dream by M. Charles (4) and may assist 
the establishment really optimal agitation systems. 

Biogal 

Pharmaceuticals established for the production of 
antibiotics together with research centers has endeavored 
to develop its devices since the beginning. According to 
the knowledge learned in the international symposium 
in Prague in 1964 where both European and US 
professionals attended, BIOGAL Pharmaceutical was 
the first pharmaceutical company applying two-turns 
driver engine which increased power utilization by 30-
40%. At the beginning of 1970s the company changed 
the systems with Rushton agitators which had 
asymmetric structure, and 20% better power on the 
lowest level. This was due to the cognition of the fact 
that in the applied asymmetric systems the efficiency of 
the lower agitators compared to the upper ones was 
considerably decreased by the function of dispersion. 
Since the beginning of seventies the company had started 
to apply a complex system including propeller agitators 
and Rushton turbines and with this method narrow OTC 
fermentors without agitators could successfully be 
adapted for penicillin fermentation. 

Based on these experiences also considering the 
construction of BIOGAL’s newer complex agitation 
systems it can be concluded that there are more 
possibilities for the increase of efficiency and 
optimization of agitation systems through the application 
of modifiable impellers recommended above. 

Conclusions 

Conditions of optimization of the aeration-agitation 
systems of large-scale fermentors in case of viscous 
broths: 
1. Providing flow modifying parts for the agitations 

system for variation of dispersion and circulation 
capabilities and adjustment for the requirements of a 
specific fermentation process. 

2. Evaluation of the results of variation by fermentation 
professionals and choosing optimum variation. 

 
Considering these on a long term basis may lead to 

gain profound knowledge about specific requirements of 
fermentation processes and industrial optimization may 
become unnecessary in the future. 

NONATION 

DO dissolves oxygen 
P/V power/volume 
kLa volumetric oxygen transfer coefficient 
vs gas velocity 



 

 

39

REFERENCES 

1. BAKKER A., SMITH J. M., MEYERS K. J., Chemineer, 
PO Box 1123, Daytona, OH45401, Reprinted from 
Chemical Engineering 

2. NIENOW A. W.: 9th Biotech Symposium, Crystal 
City, USA, 1992 pp. 196-196 

3. MEYERS K., REEDER M., BAKKER A., RIGDEN M.: 
Agitating for Success, The Chemical Engineering 

4. CHARLES M.: Trends in Biotechnology, Vol 3., No.6 
180 

5. VAN’T RIET K., SMITH J. M., Chemical Eng. Sci. 
1975, 30. 1083 

6. BUCHHOLZ H., BUCHHOLZ R., NIEBENSCHUTZ H., 
SCHÜGERL K.: Eur. J. Appl. Microb. And Biotechn. 
6. 1978, 115. 

7. BUCKLAND et al. Bioengineering Vol. 31. 70. 737-
742. I. 1988 

8. Chemineer, Inc. Reprinted for Chemical Engineer 
Crammer Road, West Meadows, Daby, DE 21 6XT, 
England