che_april-2024 - 36

cence behavior of the
gas-liquid system [2].
Following the abovementioned
scale-up
rule and target, the
power-to-volume ratio,
P/V, can clearly be
decreased
with
increasing
scale. This is
first of all because vsg
increases with scale
FIGURE 2. This 2-m3 fermenter model is used for research and development
with model fluids
which varies with batch time. To
comply with the mass balance, the
volumetric air feed rate, vvm, is usually
kept constant during scaleup, as
shown in Equation (1):
(1)
with the normalized gassing rate, ˙qN,
the liquid phase fermentation broth
volume V and the internal fermenter
diameter d1. Under these conditions,
scaling up with geometric similarity
results in a linear increase of the
superficial gas velocity, vsg, with the
vessel diameter, d1, as
(2)
where A is the cross-sectional area
of the fermenter. The volumetric oxygen
transfer rate, OTR, through the
gas-liquid Interface is determined
from Equation (3):
(3)
OTR must satisfy the volumetric oxygen
uptake rate OUR of the microorganisms.
The difference between
the saturation concentration, c*, and
the
liquid
bulk
concentration, cL,
represents the driving force for gasliquid
mass transfer. The gas-liquid
mass-transfer capacity for agitated
processes, kLa, can be correlated by
Equation (4):
(4)
where k, α and β vary over a wide
range depending on the gas-feed
device, impeller type and material
properties, such as the coales36
(~
d1) and also since
the average hydrostatic
pressure in the
larger
tanks
operated
at higher liquid level results
in an increased saturation concentration,
c*, and therefore driving
force c*- cL.
This leads to operation windows,
according to Figure 1, with reduced
specific agitator power and higher
gas velocities going from laboratory
through pilot to production scale.
This trend can lead to a drastic
change in hydrodynamics. Figure 1
shows a typical scenario with a transition
from
the homogeneous-flow
regime and complete dispersion condition
in laboratory and pilot scale to
an operation in the heterogeneousflow
regime and loading condition
in production scale. This means
that the bubble-size distribution is
narrow, and the bubbles are rather
evenly dispersed in the small-scale
fermenter. In production scale, large
bubbles appear, and an increasingly
inhomogeneous local distribution
of the gas phase can occur. Finally,
flooding must be strictly avoided, as
gas-liquid mass transfer would drastically
break down. The flooding limit
can be shifted towards higher gas
flowrates when
using modern
concave-type
impellers instead
of the traditional
ones, such as,
for example, flat
blade disc (Rushton)
turbines or
pi tched-blade
turbines. The effect
of the interaction
of the impeller
type with,
for example, the
feeding device
or internal heat exchangers and the
influence of the operation range on
the agitator system performance
can be investigated with model fluids
in sufficiently large scale of ~1 m3
as shown in Figure 2. This allows a
comparison of agitator and fermenter
systems regarding the general
flow pattern with possible stagnant
areas, gas-liquid mass transfer and
blend time.
Process design for pure gases
Fermentations with pure gases have
recently become increasingly important.
Examples are power-to-food
or power-to-gas processes. These
processes have both similarities to
traditional aerobic fermentations
and significant differences that are
more reminiscent of pure gas reactions
such as hydrogenation. Many
of these new processes have so far
been operated on a pilot scale or
smaller production scales. To benefit
from " economies of scale, " the understanding
of these processes still
needs to be significantly improved.
However,
the
experience
already
gained with aerobic fermentation and
reactions with pure gases can be
helpful. Reactors are therefore used
for pure gas reactions, using gas recirculation
from the headspace. This
variant is particularly suitable for processes
in which the gas supplied to
the process needs to be completely
converted. The fresh gas is usually
introduced into the bottom of the reactor
and split into small gas bubbles
by a primary disperser. At the same
time, unreacted gas is redispersed
from the headspace via a hollow
shaft using a self-inducing impeller
FIGURE 3. The impact of heterogeneous-flow conditions on the kLa value can
be seen here
CHEMICAL ENGINEERING WWW.CHEMENGONLINE.COM
APRIL 2024
http://WWW.CHEMENGONLINE.COM
che_april-2024

Table of Contents for the Digital Edition of che_april-2024
