Reactor Design For Cell Growth

Cellular kinetics and associated reactor design:
Reactor Design for Cell Growth
1
Cell Growth Kinetics
Using the population growth model, we could write the cell

growth rate (rX) as
rX = μ CX (43)
where
μ : specific growth rate (per time)
CX : cell concentration (dry cell weight per unit volume)
2
Batch Fermenter
Mass balance for the cell:
0 + (rX) V = 0 + d(VCX) / dt
which for a batch reactor with constant volume

reacting mixture gives
dCX / dt = rX (44)
V for volume of the reacting mixture at time t

CX for concentration of the cells in V at time t
(rX) for cell growth rate in V at time t
3
Batch Fermenter
Combining (43) and (44), we get
dCX
= μ CX (45)
dt
If μ is a constant then integrating (45) gives,
CX = CX0 exp[μ(t-t0)] (46)
where CX = CX0 when t = t0.
4
Cell Growth Kinetics
Mostly, however, μ is not a constant with time. It depends on
CS, the substrate concentration.
The most commonly used model for μ is given by the Monod
model:
μm CS
μ= (47)
KS + CS
where μm and KS are known as the Monod kinetic parameters.
Monod Model is an over simplification of the complicated

mechanism of cell growth.
However, it adequately describes the kinetics when the
concentrations of inhibitors to cell growth are low.
5
Batch Fermenter
Substituting μ in (45) by the Monod Model given by (47), we get

dCX μm CS
= CX (48)
dt KS + CS
Equation (48) could be integrated only if we know how CS

changes with either CX or t.
How to do that?
6
Batch Fermenter
It is done as follows:
Stoichiometry could have helped. But we don’t have such

a relationship in the case of cellular kinetics.
Therefore, we introduce a yield factor (YX/S) as the ratio

between cell growth rate (rX) and substrate consumption
rate (-rS) as follows:
YX/S = rX / (-rS) (49)
We know (rX) from (43) and/or (44). But we don’t know (-rS).
Therefore obtain an expression for (-rS) as shown in the next
slide.
7
Batch Fermenter
Mass balance for substrate:
0 = 0 + (-rS) V + d(VCS) / dt
which for a batch reactor with constant volume

reacting mixture gives
dCS / dt = -(-rS) (50)
V for volume of the reacting mixture at time t

CS for concentration of the Cells in V at time t
(rS) for substrate utilization rate in V at time t
8
Batch Fermenter
rX
YX/S = (49)
- rS
dCX / dt = rX (44)
dCS / dt = -(-rS) (50)
Combining the above equations, we get
dCX / dCS = -YX/S

which upon integration gives
(CX – CX0) = YX/S (CS0 – CS) (51)
9
Batch Fermenter
Substituting CS from (51) in (49) and integrating, we get
(
μm (t - t0) =
KS YX/S
CX0 + CS0YX/S )( ) +1 ln
CX
CX0
(
+
KS YX/S
CX0 + CS0YX/S )( )
ln
CS0
CS
(52)
where
(CX – CX0) = YX/S (CS0 – CS) (51)
10
Batch Fermenter
Exercise 1:
The growth rate of E. coli be expressed by Monod kinetics

with μm = 0.935 hr-1 and KS = 0.71 g/L.
Assume that YX/S is 0.6 g dry cells per g substrate.
CX0 is 1 g/L and CS0 = 10 g/L when the cells start to grow
exponentially (i.e., at t = 0).
show how CX, CS, and dCX/dt change with respect to time.
11
Exercise 1 worked out using the calculator/spread sheet:
CS is varied from 10 g/L to 0.

CX is calculated using (49) as CX = 1 + 0.6 (10 – CS)
t is calculated using (50) as follows:
( )( )
0.71 x 0.6 CX
0.935 t = +1 ln
1 + 10 x 0.6 1
+
( 0.71 x 0.6
1 + 10 x 0.6 )( ) ln
10
CS
CX is calculated using (48).
12
specify Calculate Calculate t Calculate

CS CX using using (50) dCX/dt
(49) using (46)
10 1 0
9.95 1.03 0.0317 0.9335
9.8 1.06 0.0624 0.9332
9.85 1.09 0.0923 0.9329
Continue
until CS
becomes 0
13
12
10
8
CS
6 CX
0
0 1 Time (in hr) 2 3
14
12
dCx/dt
10
8
CS
6 CX
0
0 1 Time (in hr) 2 3
15
Exercise 1 worked out using an ODE solver:
Programme written in MATLAB
[t,y] = ode45(@BIOCHEM,[0:0.01:3],[1; 10]);
function dydt =BIOCHEM(t,y)

%data given
mumax = 0.935; % per hr
Ks = 0.71; % g/L
YXS = 0.6;
%Monod model
mu = mumax*y(2)/(Ks+y(2));
%rate equations
rX = mu*y(1);
rS = -rX/YXS;
dydt=[rX; rS]
16
plot(t,y(:,1),'b',t,y(:,2),'r')
legend('Cell','Substrate')
ylabel('Concentration (g/L)')
xlabel('Time (h)')
17
mumax = 0.935;
plot(t,y(:,1),'b',t,y(:,2),'r')
Ks = 0.71;
legend('Cell','Substrate')
mu= mumax*y(:,2)./(Ks+y(:,2));
ylabel('Concentration (g/L)')
rX = mu.*y(:,1);
xlabel('Time (h)')
plot(t,rX,'g')
18
Plug-flow Fermenter at steady-state
F F
θ = V/F
CXi, CSi CX, CS
μm θ =
( )() KS YX/S
CXi + CSiYX/S
+1 ln
CX
CXi
where
( )( )
+
KS YX/S
CXi + CSiYX/S
ln
CSi
CS
(53)
(CX – CXi) = YX/S (CSi – CS) (54)

19
Continuous Stirred Tank Fermenter (CSTF) at steady-state
F
CXi, CSi
- also known as chemostat
V
CX, CS F
CX, CS
- Mixing supplied by impellers and/or rising gas bubbles
- Complete mixing is assumed (composition of any phases do
not vary with position)
- Liquid effluent has the same composition as the reactor
contents
20
F
CXi, CSi
V
CX, CS F
CX, CS
Mass balance for cells over V:
FCXi + rX V = FCX (55)
21
Equation (55) gives

V CX - CXi
= (56)
F rX
Introducing Dilution Rate D as

F 1
D = = (57)
V θ
in (56), we get
1 CX - CXi
= (58)
D rX
22
Since rX = μ CX, (58) becomes
1 CX - CXi
= (59)
D μ CX
If the feed is sterile (i.e., CXi = 0), (59) gives
CX (D – μ) = 0 (60)
which means either CX = 0 or D=μ
23
If D = μ, then
μm CS
D = μ = (61)
KS + CS
(61) can be rearranged to give CS as
KS D
CS = (62)
μm - D
To determine CX, we need to write the mass balance for

substrate over the CSTF
24
F
CXi, CSi
V
CX, CS F
CX, CS
Mass balance for substrate over V:
FCSi = FCS + (-rS) V
25
which is rearranged to give
(-rS) = D (CSi - CS) (63)
(58) gives
rX = D (CX - CXi )
Using the above equations in the definition of yield factor, we get
(CX – CXi) = YX/S (CSi – CS) (64)
26
Since the feed is sterile, (6 4) gives
CX = YX/S (CSi – CS) (65)
(62) is
KS D
CS = (62)
μm - D
Therefore, we have
CX = YX/S
(CSi -
KS D
μm - D ) (66)
27
KS D
CS = (62)
μm - D
which is valid only when D < μm
CX = YX/S
( CSi -
KS D
μm - D ) (66)
which is valid only when
CSi > KS D / (μm - D)
D < CSi μm / (KS + CSi)
28
Since D < CSi μm / (KS + CSi) < μm
critical value of the Dilution Rate is as follows:
DC = CSi μm / (KS + CSi) (67)
29
If μm equals or less than DC, then CX is negative.
That is impossible.
So, when μm equals or less than DC,
We need to take the solution

CX = 0 of (58), not D = μ
Substituting CX = 0 in CX = YX/S (CSi – CS) gives
CS = CSi
30
CX = 0 means no cell in the reactor.
CS = CSi means substrate is not utilised.
Since the CSTF has a sterile feed (CXi = 0),

no reaction takes place unless we inoculate with the
cells once again.
So, CSTF gets into a WASHOUT situation.
To avoid CSTF getting into WASHOUT situation,

we need to maintain D = F / V < DC
31
Exercise 2
The growth rate of E. coli be expressed by Monod kinetics

with μm = 0.935 hr-1 and KS = 0.71 g/L.
Assume that YX/S is 0.6 g dry cells per g substrate.
The feed is sterile (CXi = 0) and CSi = 10 g/L.
show CX and CS changes with dilution rate.
32
Plot the following using excel / MATLAB
From (60): CS = 0.71 D g/L

0.935 - D
(
From (64): CX = 0.6 10 - 0.71 D
0.935 - D ) g/L
From (65): DC = CSi μm / (KS + CSi)

= 10 x 0.935 / (0.71+10) = 0.873 per h
33
DC = 0.873 34
Near washout the reactor is very

sensitive to variations in D. Small
change in D large shifts in X and/or S.
DC = 0.873 35

Reactor Design For Cell Growth

Uploaded by

Copyright:

Available Formats

Reactor Design For Cell Growth

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Reactor Design For Cell Growth

Uploaded by

Copyright:

Available Formats

Cellular kinetics and associated reactor design:

Reactor Design for Cell Growth

Using the population growth model, we could write the cell

Mass balance for the cell:

which for a batch reactor with constant volume

V for volume of the reacting mixture at time t

If μ is a constant then integrating (45) gives,

CX = CX0 exp[μ(t-t0)] (46)

where CX = CX0 when t = t0.

where μm and KS are known as the Monod kinetic parameters.

Monod Model is an over simplification of the complicated

Substituting μ in (45) by the Monod Model given by (47), we get

Equation (48) could be integrated only if we know how CS

Stoichiometry could have helped. But we don’t have such

Therefore, we introduce a yield factor (YX/S) as the ratio

YX/S = rX / (-rS) (49)

Mass balance for substrate:

which for a batch reactor with constant volume

dCS / dt = -(-rS) (50)

V for volume of the reacting mixture at time t

dCS / dt = -(-rS) (50)

Combining the above equations, we get

dCX / dCS = -YX/S

(CX – CX0) = YX/S (CS0 – CS) (51)

Substituting CS from (51) in (49) and integrating, we get

(CX – CX0) = YX/S (CS0 – CS) (51)

The growth rate of E. coli be expressed by Monod kinetics

Assume that YX/S is 0.6 g dry cells per g substrate.

CS is varied from 10 g/L to 0.

t is calculated using (50) as follows:

CX is calculated using (48).

specify Calculate Calculate t Calculate

9.85 1.09 0.0923 0.9329

Programme written in MATLAB

[t,y] = ode45(@BIOCHEM,[0:0.01:3],[1; 10]);

function dydt =BIOCHEM(t,y)

(CX – CXi) = YX/S (CSi – CS) (54)

FCXi + rX V = FCX (55)

Equation (55) gives

Introducing Dilution Rate D as

Since rX = μ CX, (58) becomes

If the feed is sterile (i.e., CXi = 0), (59) gives

which means either CX = 0 or D=μ

(61) can be rearranged to give CS as

To determine CX, we need to write the mass balance for

Mass balance for substrate over V:

FCSi = FCS + (-rS) V

which is rearranged to give

(-rS) = D (CSi - CS) (63)

Using the above equations in the definition of yield factor, we get

(CX – CXi) = YX/S (CSi – CS) (64)

Since the feed is sterile, (6 4) gives

CX = YX/S (CSi – CS) (65)

which is valid only when

CSi > KS D / (μm - D)

D < CSi μm / (KS + CSi)

Since D < CSi μm / (KS + CSi) < μm

critical value of the Dilution Rate is as follows: