CHEMICAL ENGINEERING TRANSACTIONS

VOL. 79, 2020

A publication of

The Italian Association

of Chemical Engineering
Online at www.cetjournal.it

Guest Editors: Enrico Bardone, Antonio Marzocchella, Marco Bravi
Copyright © 2020, AIDIC Servizi S.r.l.
ISBN 978-88-95608-77-8; ISSN 2283-9216

Determination of Volumetric Oxygen Transfer Coefficient to
Evaluate the Maximum Performance of Lab Fermenters

Lucia Paciello*,Palma Parascandola

Department of Industrial Engineering, University of Salerno, Italy
lpaciello@unisa.it

The biomass productivity of some fermentation processes is limited by the oxygen availability in the medium.
Consequently, it is necessary to know the microorganism’s oxygen demand as well as the volumetric oxygen
transfer rate of the fermenter. In this study, the oxygen transfer performance of two lab fermenters was
evaluated from the volumetric oxygen transfer coefficient (KLa). Here the aeration system of these two
fermenters was investigated by operating with two different impellers, i.e., the Rhuston turbine and the
pitched-blade impeller, according to the stirring speed, the air flow rate, the broth-culture rheological
properties, and the presence of antifoam agents in the medium. The experimental results show that the trends
of KLa against stirring speed, obtained by using the two impellers, are different. Furthermore, the results
highlight the strong lowering of KLa with increasing the viscosity of the medium. The formation of foam by
using a typical rich medium, composed by peptone and yeast extract, reduces the transfer of oxygen. In these
conditions the addition of antifoam into the medium is fundamental to improve the oxygen transfer and
consequently the fermentation duration in the case of high agitation speed.

1. Introduction

In the fermentative processes the choice of the most suitable bioreactor is of fundamental importance. This
choice is already made at the laboratory scale taking into account that the main function of a properly
designed bioreactor is to provide a controlled environment in order to achieve the optimal growth and/or
product formation (Falco et al., 2014; Landi et al., 2014, Paciello et al., 2013).
In aerobic fermentation processes, oxygen transfer is the key problem. As a matter of fact microorganisms
growing in submerged cultures can use only dissolved oxygen. Therefore all oxygen must first be dissolved in
the broth and then transported to the cell. In this perspective, no limitation should affect oxygen transfer from
the gaseous phase to the microorganism.
The concentration of dissolved oxygen in a suspension of breathing microorganisms generally depends on the
rate of oxygen transfer from the gas phase to the liquid phase, the rate of oxygen transfer to the site of
utilization, and the rate of its consumption by the microorganism. In the conventional processes utilizing water
soluble carbohydrate as substrate, it has frequently been found that the rate of oxygen transfer from dispersed
air bubbles to liquid phase can become a limiting factor of the microbial growth rate (Enfors and Haggstrom,
1998). In the bioreactor, the aeration system supplies the oxygen and the agitation system maintains uniform
conditions within the bioreactor. Altogether, the aeration and agitation are important in promoting effective
oxygen transfer to the liquid medium in the bioreactor, but the oxygen transport is however strongly limited by
various factors, for example the type of impeller, the aeration rate, the foam formation and the rheology of the
broth-culture.To evaluate the performance of an aerobic bioreactor as a whole, the volumetric oxygen transfer
coefficient (KLa) is the most important parameter because it represents the capacity of oxygen supply and
transfer in the fermenter which is related to agitation speed, aeration rate, geometrical characteristic of
fermenter and rheological character of the medium. In the scientific literature are present many works where
correlations for kLa were proposed for the scale-up, design and performance optimization of agitated vessels
(Labík et al., 2017). In this study, the oxygen transfer performance of two lab fermenters (Biostat® B - B.Braun
Biotech Int., and Bioflo110 - New Brunswick Scientific) was evaluated from the determination of volumetric
oxygen transfer coefficient (KLa). Aeration system of the two fermenters was investigated employing two

DOI: 10.3303/CET2079013

Paper Received: 13 September 2019; Revised: 29 January 2020; Accepted: 23 February 2020
Please cite this article as: Paciello L., Parascandola P., 2020, Determination of Volumetric Oxygen Transfer Coefficient to Evaluate the
Maximum Performance of Lab Fermenters, Chemical Engineering Transactions, 79, 73-78 DOI:10.3303/CET2079013

different impellers, i.e., the Rhustun turbine and the pitched-blade impeller, according to the stirring speed, the
air flow rate, the broth-culture rheological properties, and the presence of antifoam agents and cells in the
medium. This experimental work is part of a wider work on heterologous protein production by engineered
yeast cells. The main goal of this research is to know the maximum performance of fermenters in terms of
oxygen transfer in order to exclude the engineering limitations as the cause of yeast growth arrest, a
phenomenon frequently observed during fermentation runs in fed-batch reactor (Mazzoleni et al.,2015, Landi
et al., 2011).

2. Materials and methods

2.1 The fermenters and impellers employed

Tests for the determination of KLa were carried out with two fermenters, the Biostat
®B (B.Braun Biotech Int.,

Melsungen, Germany) and Bioflo110 (New Brunswick, Edison, NJ). A working volume of 1 litre was chosen in
a 2-litres total volume bioreactor. Air was supplied to the fermenter with a standard sparger located at the
base of the agitator shaft. The two-phase (liquid-gas) dispersion was agitated with one stirrer whereas four
equally spaced baffles were used to enhance mixing. As regards Bioflo 110, the efficiency of two different
stirrers i.e. a pitched blade propeller and a Rushton-style impeller (Figure 1) was investigated, whereas in the
case of the Biostat®B-Braun only the Rushton-style impeller was taken into consideration (Figure1, right). The
pitched blade propeller is constituted by three 45 degree inclined blades and the diameter and the height are
60 and 46 mm respectively. The design of Rushton-style impeller is based on cylinder with six flat blades
vertically mounted. The impeller has a total diameter of 50 mm and a singular blade has dimensions of 18 x 12
mm.

Figure 1: Pitched blade (left) and Rushton-style(right) stirrers

2.2 kLa determination

kLa (volume of oxygen per volume of liquid per time, time
-1) determination, at a fixed agitation speed, with the

gassing out method (Enfors and Haggstrom 1998) was carried out without cells in the medium. Oxygen
concentration in the medium was initially reduced to zero by flushing with nitrogen gas after which aeration
was started and the increase in Dissolved Oxygen Tension (DOT) signal was monitored through the Clark
electrode until air saturation was achieved (Figure 2).

Figure 2: Dissolved Oxygen Tension (DOT) signal vs. time

The data so obtained have been manipulated to calculate the slope (-kLa) of curve ln(DOT*-DOT) vs. time
where DOT* is the saturation value in the medium and DOT the current value at time t. KLa determination

100

120

0 100 200 300 400 500 600

D
O

T
[%

ai
r

sa
tu

ra
tio

Time [s]

tests have been made in duplicate for each examined stirring value. The examined media were the following:
distilled water, YEP broth, a typical rich complex medium for yeast growth containing 1% Yeast Extract
(BDTM), 2 % Peptone (BDTM) and 5% w/v Destrose (BDTM), and YEP medium described above added with
1.4 ml antifoam C (diluted ten times) (Dow Corning). In the case of tests with broth-culture, yeast cells of
Saccharomyces cerevisiae CENPK113.7D strain, subjected to thermal death (10 min at 80°C), were added to
a final concentration of 20 mgd.w.ml

-1.

2.3 Results

As regards Bioflo-110, first experimental results, shown in Figures.3 and 4 for pitched blade propeller and
Rushton-style impeller, respectively, were obtained employing distilled water as liquid, varying stirrer speed in
the range of 200-900 rotations per minute (rpm) and fixing, each time, the aeration rate at two different values,
1 and 2 air volumes per liquid volume per minute (vvm), that are the values most utilized during fermentation
runs. In fact it is not possible to work over 2 vvm because if the air volume that passes through the liquid
phase is more than twice that of the liquid phase the dragging phenomenon occurs.
For both the stirrers the kLa increases with increasing the aeration rate (Figures. 3 and 4). In fact, raising the
aeration rate increases the parameter “a” of the product KLa, since the total volume of gas suspended in the
liquid increases with the increase of aeration rate, and for high aeration rate the kLa values almost doubles.

Figure 3: Pitched blade propeller: KLa values vs. stirrer speed using water as medium

Figure 4: Rushton-style impeller: KLa values vs. stirrer speed for water as medium

When a pitched blade impeller was employed (Fig.3) KLa exhibited a maximum value at 500 and 800 rpm in
correspondence with 1 and 2 vvm, respectively, then it decreased. This means that the pitched blade impeller
gives a low contribution to turbulence and oxygen transfer at the highest stirrer speed. Consequently, it was
not suitable for ensuring an optimal gas-liquid mass transfer rate required to satisfy the oxygen demand of
respiratory metabolism during the bioprocess. On the contrary, by employing the Rushton-style impeller
(Fig.4), KLa values increased as a monotone function of stirrer speed. The behavior exhibited by the two

stirrers is ascribable to the different flow patterns they promote. The propeller pumps the liquid axially and
gives, as above mentioned, a low contribution to turbulence and oxygen transfer whereas the impeller pulls
most of liquid to flow from the axis towards the periphery causing more turbulence and therefore a higher gas-
liquid mass transfer rate.

Figure 5: Effect of liquid composition in the vessel on KLa employing Rushton-style impeller (Bioflo110)

The study of KLa as a function of stirrer speed gave interesting results also when, employing a Rushton-style
impeller, distilled water was replaced by rich medium (YEP) and YEP containing siliconic antifoam (Figs.5 and
6). As expected, KLa values of Bioflo 110, obtained with distilled water, drew higher values with respect to
those obtained employing broth with antifoam (Fig.5). If the antifoam was not added, a maximum KLa value
was achieved in correspondence of about 750 rpm agitation speed, then KLa diminished. The composition of
YEP, especially the presence of proteins in the peptone made YEP a mixture similar to a colloidal solution
which easily gave rise to foam bubbles capable of hampering oxygen transfer from gaseous into liquid phase.
The addition of antifoam favored the coalescence of bubbles so that the dissolution rate of gas into the liquid
phase raised up from 25 to 130 m3 of oxygen per m3 liquid volume per hour at 900 rpm. The addition of
antifoam is a common procedure employed during the fermentation processes.

Figure 6: Effect of liquid composition in the vessel on KLa employing Rushton-style impeller (Biostat B)

The same behaviour was obtained with the Biostat® B (Fig.6), where KLa was evaluated employing the broth-
culture as well. In this case, KLa values resulted the lowest encountered until then. As a matter of fact the
viscosity of the broth-culture caused by dead cells in addition to the medium composition affected oxygen
transfer further on. In the figure 7 it is possible to appreciate, through the pictures taken during the kLa
determination tests for Bioflo110 bioreactor, the behavior of the different media at various stirring rate when
the air bubbles come out from the sparger.
At low stirring speed, the oxygen bubbles exiting from the sparger go up almost undisturbed. In these
conditions the oxygen transfer coefficient values are low because the permanence times in the liquid medium
are low such as the “a” value, that is the amount of interfacial surface area per unit volume of oxygen.
With increasing the impeller speed, the oxygen permanence time in the liquid phase increases and, already in
the case of distilled water, it is possible to see the little bubbles of air that are chopped by the impeller and the
foam arises proving the large air volume trapped in the liquid phase.

100

150

200

0 200 400 600 800 1000

K
La

[h
-1

]

Stirring [rpm]

Broth
Broth with antifoam
Water
Cells with antifoam

Water

Broth

Broth
with
anti-
foam

200 rpm 500 rpm 750 rpm 900 rpm

Figure 7: Effect of liquid composition in the vessel by using Rushton-style impeller (Bioflo110)

Interesting is the behavior of the rich medium without and with the addition of the siliconic antifoam. It is worth
noting that the kLa value at low rpm is the same in both the media. However, when the stirring reaches the 500
rpm the foam formation is huge and it is clearly noticeable the separation between the foam volume and the
volume with the gaseus and liquid phases. With the increasing of stirring the situation is getting worse and at
900 rpm stirring rate, the whole internal volume of bioreactor is filled with foam and the culture medium risks
going outside: the medium goes up along the condenser to the filter at the extremity of the exhausted gas
tube, leading to the wetting of the filter itself and system blocking due to overpressure.By antifoam addition,
the foam is reduced even at high stirring speed, and the air volume is kept inside the liquid medium, resulting
in longer permanence times of oxygen bubbles in the liquid phase and a more effective oxygen transfer to the
liquid phase.

2.4 Determination of critical biomass concentration

Once KLa values are known together with that of the specific microorganism oxygen consumption rate,qO2 (h
-1)

it is possible to calculate the critical biomass concentration, X (mg ml-1), that is the biomass concentration in
correspondence to which the oxygen consumption rate (rO2) equals the maximum oxygen transfer rate
(OTRmax) of the fermenter, that is the transfer rate, at a fixed stirring value, where the oxygen gradient is
maximum (C* is the saturation concentration of oxygen in the liquid). From the following equation it is possible

(1)

to determine the critical concentration value for each microorganism used for the aerobic bioprocess and
therefore the maximum performance of bioreactors in terms of oxygen transfer. In these conditions, to avoid
oxygen limitation which could determine a severe limitation of cell growth rate, the enrichment of air with pure

( )
22max

*
OOL rXqOTRaCK =⋅==

oxygen is required so that, C* enhances and as a consequence OTRmax increases as well. A different way to
by-pass the problem would be to reduce the feeding rate of carbon source so that the growth rate significantly
diminishes so as to obtain a volumetric oxygen consumption rate again below OTRmax.

3. Discussion

To evaluate the maximum performance of the fermenters in terms of oxygen transfer it is necessary to
evaluate the volumetric oxygen transfer coefficient, kLa, in the real conditions of microorganism growth. In fact
the experimental evaluation of kLa must be made not by using distilled water as liquid medium but employing
the same broth used for the microorganism growth because the presence in the culture medium of little
amounts of salts, precursors, antifoam and cells lower the oxygen transfer efficiency. As expected, the
agitation and aeration influence strongly the oxygen transfer in the fermenter. The agitation is the most
influential parameter, affecting positively the oxygen mass transfer according to the experimental results
obtained for the two lab fermenters. Also the aeration rate has a great influence on kLa. The choice of stirrer is
the first step to set up the aerobic fermentation not only because it must guarantee the perfect mixing but also
it must ensure a very good oxygen transfer. The stirrer frequently used in aerobic bioprocesses is the
Rushton-style impeller for its vigorous stirring but for more gentle mixing, for example for mammalian cells
growth, is often used the pitched-blade stirrer. The results indicate that trend of kLa versus agitation speed is a
monotone increasing curve for the Rusthon impeller while for the pitched-blade impeller the curve have a
maximum for central values of agitation rate range under investigation. The rheology of growth medium greatly
influences the oxygen transport and for this reason it is fundamental to know the kLa value for each temporal
value of bioprocess to allow to operate promptly to optimize the productivity and to reduce the bioprocess
costs.

4. Conclusions

This work presents a valid method to evaluate the maximum performance of fermenters in terms of oxygen
transfer. The results highlight the importance to evaluate the volumetric oxygen transfer coefficient (kLa) in the
real bioprocess condition in order to not overestimate the maximum performance of lab fermenters.
The importance of this work compared to other experimental works is based on the possibility to know the
maximum performance of the growth system in terms of oxygen transfer so that making possible a deeper
investigation of other factors which can be responsible for the yeast growth arrest observed during
fermentation runs in fed-batch reactor.

References

Enfors S.O. and Haggstrom L. 1998, Bioprocess Technology: Fundamentals and Applications.
Hogskoletryckeriet, Stockolm.

Falco F.C., Landi C., Paciello L., Zueco J., Parascandola P., 2014, Fed-Batch production of endoglucanase
with a recombinant industrial strain of the yeast Saccharomyces cerevisiae, Chemical Engineering
Transactions, 38, 379-384, DOI: 10.3303/CET1438064.

Labík L., Moucha T., Petříček R., Rejl J.F., Valenz L., Haidl J., 2017, Volumetric mass transfer coefficient in
viscous liquid in mechanically agitated fermenters. Measurement and correlation. Chemical Engineering
Science 170, 451-63

Landi C., Paciello L., De Alteriis E., Brambilla L., Parascandola P., 2014, Effect of auxotrophies on yeast
performance in aerated fed-batch reactor. Biochemical and Biophysical Research Communications, 414:
604-611.

Landi C., Paciello L., De Alteriis E., Brambilla L., Parascandola P., 2014, High cell density culture with S.
cerevisiae CEN.PK113-5D for IL-1β production: optimization, modeling, and physiological aspects.
Bioprocess and Biosystems Enginnering, 38(2), 251-61

Mazzoleni S., Landi C., Cartenì F., De Alteriis E., Giannino F., Paciello L., Parascandola P., 2015, A novel
process-based model of microbial growth: self-inhibition in Saccharomyces cerevisiae aerobic fed-batch
cultures. Microbial Cell Factories, 14,109

Paciello L., Landi C., Zueco J., Parascandola P. 2013, Production in fed-batch reactor of Bacillus
subtilis lipase A immobilized on its own producer Saccharomyces cerevisiae cells. Chemical Engineering
Transactions, 32:907-912, DOI: 10.3303/CET1332152