Mixing: Aeration and Agitation in a Stirred Tank Reactor

• Maintain uniform conditions in the vessel (solid, liquid, gas
concentration, Temperature, pH).
• Disperse bubbles throughout the liquid, promote bubble break-up,
increase gas-liquid interfacial transfer (bigger the interfacial area
for diffusion, the better)
• romote mass transfer of essential nutrients
Mixing is effected by
• !eration and agitation in a "tirred Tank #eactor
• !eration (and consequent fluid circulation) in an !ir $ift #eactor
"chematic of "tandard tank configuration
Agitators in Bioreactors
Rushton Turbine Impeller in Glass Bioreactor
Types of agitator
• µ (apparent viscosit%) & '( c, high ) (rotational speed)
⇒ turbine (rushton or inclined blade) like above
#emote clearance* D (agitator diameter) + T (tank diameter)
* (.,'-(.')
-essel baffled (in general, four strips of metal running parallel to the
.all of the bioreactor, protruding into the liquid) to prevent vorte/
(similar to flo. behaviour about a sink plug hole) formation at high
agitation speeds
The impact of turbine blade pitch on flow pattern
0lat blade ⇒ #adial flo. (radial means perpindicular to the shaft of
the bioreactor. - out.ards)

"ketch and measure*
itched+inclined blade+propeller ⇒ a/ial component (a/ial means that
a proportion of the primar% flo. is parallel to the shaft 1
up+do.n.ards)
"ketch and measure*
Marine propellers ⇒ three blades, .ide range of ), high shearing
effect at high rotational speeds
"ketch and Measure*
High iscosity Solutions
• High µ⇒ anchors.helical ribbons ( and propellers)
!nchors, helical ribbons*D+T 2(.3
$o.er speeds, vessels generall% not baffled
• 4ntermig agitator ⇒ a/ial pumping impeller requires less energ%
and lo.er gas through-put to produce same mass transfer
coefficient as turbine.
4nsert 4ntermig icture Here*
• 0or adequate particle suspension and dispersal, ma% require
profiled vessel base5 inclined-blade agitators preferable
!imensionless "umbers in Agitated#Aerated Systems
6e use dimensionless numbers in agitated+aerated s%stems to help
us characterise the design and performance of the process, ho.ever
in a scale independent manner.
The first dimensionless number presented is the po.er number, )

' 7
D N
P
N
P
ρ
·

This number in con8unction .ith 4mpeller #otational "peed ()),
4mpeller Diameter (D) and $iquid Densit% (ρ) allo. us to calculate the
Mechanical o.er () being transmitted to the fluid b% a
turbine+impeller of a given design.
#e%nolds )umber is the second ke% number in the set of
dimensionless numbers. !gain similar to applications in pipes, etc.,
the #e%nolds number indicates the degree of turbulence e/perienced
in a stirred tank reactor.
µ
ρ
,
#e
ND
N ·
6here µ is the viscosit% of the liquid in .hich the agitator is turning.
0lo. )umber ()
9
) 1 :seful measure of the pumping capacit% of an
impeller. !gain the number is design specific and independent of
scale.
7
ND
Q
N
Q
·

!eration )umber ()
9g
) 1 :seful measure of the gas dispersion
capabilities of the impeller.
7
ND
Q
N
g
Q
g
·

; agitator po.er (6) ().<. "haft po.er onl%)
D ; impeller diameter (m)
ρ ; fluid densit% (kg m
-7
)
) ; impeller speed (s
-=
)
µ ; fluid viscosit% ()s m
-,
)
9 ; fluid flo. rate (m
7
s
-=
)
9
g
; gas flo. rate (m
7
s
-=
)
The Relationship of $ower "umber and Reynolds "umber
#elationship has three phases 1 each phase corresponding to the
three phases of liquid flo., laminar, transition and turbulent
! plot of $n )

vs $n )
#e
⇒ straight line, slope 1=
Turbulent flo., )
p
independent of )
#e
(also constant)
<ioreactors are, in the main, in turbulent flo.. This means that the
po.er number is constant for a given impeller design. o.er
numbers for a variet% of impellers in turbulent flo. have been .ell
characterised, therefore if .e kno. the impeller diameter and the
rotational speed of the impeller (both eas% to measure) .e can
subsequentl% estimate the mechanical po.er input to the bioreactor.
It is important to note that all of the correlations presented apply to
ungassed, single phase fluids only ⇒ no allowances for aeration or
suspensions
4n general the Gassed Power is less than the calculated ungassed
power. ! general rule of thumb for the calculation of gassed po.er is

g
; (.>

%xample
?alculate the specific po.er requirement (+-) for a standard
configuration "T#, full% baffled, fitted .ith a #ushton turbine and
containing .ater at ,'
(
?. The vessel diameter is (.'m. The impeller
speed is 7((rpm.
Solution
"tandard "T#⇒ T ; (.'m
D ; T+7 ; (.=>@m
H ; T ; (.'m
- ; ∏ T
7
+A ; (.(3Bm
7
( )
7
,
,
#e
=( =
=>@ . (
>(
7((
=(((

,
_

¸
¸
· ·
!
ND
N
µ
ρ

'
#e
=( A . = =73AA' ! N ≈ ·
⇒ full% turbulent flo., therefore from the o.er )umber #e%nolds
)umber correlation graph, (curve = is a #ushton turbine 1 remember
not to misread the log scaleC)
)

;'
;)
p
ρ)
7
D
'
; (')(=((()(7((+>()
7
((.=>@)
'
; B=6
Power input per unit "olume is a useful compariti"e measure
between bioreactors of different scales
7 7
+ = + B,B
(3B . (
B=
m #$ m $
%
P
≈ · ·
Typical Specific $ower &onsumptions '$#( k)#m
*
Mild agitation (.=
"uspending light solids
<lending of lo. viscosit% liquids
Moderate Agitation (.A
Das dispersion, liquid-liquid contacting
"ome heat transfer
+ntense Agitation =.(
"uspending heav% solids, emulsification
<lending pastes, dough A.(
+ndustrial,scale fermenters (.'-'
-ab,Scale fermenters '-=(
Reynolds "umber ranges for Rushton turbine
#e & =(
=
laminar flo.
=(
=
& #e & =(
A
transitional flo.
#e & =(
A
⇒ turbulent flo.

Mixing %ffecti.eness
• Mi/ing time t
m
1 time required to achieve specified degree of
homogeneit%, starting from the completel% segregated state
• ! sub&ecti"e quantit%
• Measured b% tracer studies
4n8ect a tracer pulse into the agitated vessel
Monitor concentration at a single point
• ?olouring+decolouring method
- e.g. meth%lene blue, iodine+starch
- simple to implement
- monitor b% e%e+spectrophotometer
- good for detection of stagnant regions
but - d%e ma% adhere to biomass
- ?oloration is irreversible (disposalE)
- vessels seldom transparent ⇒ sampling
• conductivit%
- electrol%te tracer e.g. F?$ added to vessel
- monitor response using conductivit% probe
- fast probe response time
- cheap and reliable for small scale s%stems using .ater
But - bubbles interfere .ith measurement
- addition of electrol%te to broth ⇒ changes in osmotic
pressure ⇒ rheological effects
- not suitable for actual fermentation s%stems
• pH
- acid added
- one (or more) pH probes to monitor response
- pH probes steriliGable, .idel% available
- acid addition circuit available for pH control
- most suitable for large-scale applications
- suitable for three-phase s%stems
but - pH signal requires careful interpretation
&orrelations for t
m
in Stirred Tank Reactors
Single,phase li/uids
0or full% turbulent flo., the energ% delivered to the fluid b% the
impeller P, is completel% transformed into kinetic energ% of the liquid*
,
,
' 7
u
Q D N N P
P P
ρ
ρ · · '0(
6here Q
P
is the pumping capacit% of the impeller (m
7
s
-=
) and u is the
liquid velocit% as it leaves the impeller. 0or an impeller blade .idth w,
Dw u Q
P
Π ·
'1(
The circulation time t
circ
is defined as
circ
circ
circ
Q
%
t ·
'*(
0or an agitated vessel, Q
circ
, the circulation capacit% is greater than
the pumping capacit% Q
P
due to liquid entrainment b% the impeller.
H/perimentall% it has been determined that*
P circ
Q Q , ≈
'2(
The mi/ing time t
mi!
is related to t
circ
as follo.s*
circ mi!
t t A ≈
'3(
!ssuming %
circ
; % ; ΠT
'
()* and that
,
,
u u · '4(
Hquations (=)-(') %ield
77 . (
,
7
1
1
]
1

¸

,
_

¸
¸

,
_

¸
¸

,
_

¸
¸
·
D
w
N
T
(
D
T
N
c
t
P
mi!
'5(
0or the assumptions made above cI(.>.
0rom equation (@), for full% turbulent flo. (i.e N
P
constant)
Nt
mi!
; constant '6(
0or (+T and w;(.,D,
( )
[ ]
77 . (
7
+ J
P
mi!
N
D T
N
c
t ·
'7(
6here c
K
I =.@', in this case.
Ln the basis of e/perimental evidence for a .ide range of impellers
and assuming a mi/ing intensit% of I3(M, c
,
I7 (for single phase
s%stem, #e2=(,((()
0or #e &= / =(
A
, Nt
mi!
as #e↓
8or aerated systems '1 phases(
• ?omparativel% little e/perimental data available
• $imited range of reactor+impeller design
• Fno.ledge of the flo. mechanisms limited
• 0or gas flo. rates near the flooding region, influence of gas phase
ma% be significant
• Ln the basis of data available
⇒ mi! mi!
t t ,
, ,

for equation (3), c
K
I >.
Significance of t
mix
for bioreactor operation9
• H 1 measurement and controlE
• DL concentrationE