10/17/2017

Comments on Productivity
of Batch & Continuous
Bioreactors (Chapter 9)

Topics

Definition of productivity
Comparison of productivity of batch vs flowing systems

Updated: October 17, 2017

2
John Jechura (jjechura@mines.edu)

1
 10/17/2017

Review Batch Reactor

Cell Balances (constant volume):

dNX dX
 VR
dt dt
dX 1 dX
  net X   g  kd  X   net 
dt X dt

Cycle time for a batch system (lump non-productive time as “lost”):

1  Xm 
tc  tlag  tgrowth  tharvest  t prep  tgrowth  tlost  ln    tlost
 m  X0 
Productivity (average rate of production):
1 NX Xm  X0 YX /S S0
rb   
VR t tc 1 m  ln  Xm X0   tlost
Updated: October 17, 2017
3
John Jechura (jjechura@mines.edu)

Review Steady State Chemostat (CSTR)

Cell balance:
 net  g  kd  D

Substrate balance:

D  S0  S  D  q
 app  Mg  P  m
X YX /S YX /S YP/S

Rate of cell production:

N X  FX  VR  DX 

Productivity is the DX term

Updated: October 17, 2017

4
John Jechura (jjechura@mines.edu)

2
 10/17/2017

Chemostat Productivity with Monod Kinetics

From Monod expression:

mS KS g
g   S
KS  S  m  g

Substrate balance (net = growth):

D  S0  S  g  KD 
  X   S0  S  YXM/S   S0  S  YXM/S
X Y M
X /S  m  D 

Productivity:
 KD 
rC  DX  D  S0  S  YXM/S  D  S0  S  YXM/S
 m  D 

Updated: October 17, 2017

5
John Jechura (jjechura@mines.edu)

Chemostat Productivity with Monod Kinetics

What happens with increasing D?

 At small D the cell mass concentration is
at its maximum
 KD 
X   S0  S  YXM/S  S0YXM/S
  m D

 As D increases X always decreases, but…

 … the productivity will increase since the increase in D is faster than the
decrease in X
 As D approaches µm the cell concentration goes to zero (cell mass 
washout) & the productivity DX will also go to zero
 S
Dmax  m 0
KS  S0
 There will be an optimum productivity in between
Updated: October 17, 2017
6
John Jechura (jjechura@mines.edu)

3
 10/17/2017

Chemostat Productivity with Monod Kinetics

Optimal productivity
when d(DX)/dD = 0
 KS 
Dopt   m 1 
 KS  S0 

Xopt
  KS  S0 
YXM/S
 KS  KS  S0 

 KS 
rC ,opt  YXM/S m 1
 KS  S0 

 KS  S0  KS  KS  S0  
Updated: October 17, 2017
7
John Jechura (jjechura@mines.edu)

Chemostat Productivity with Monod Kinetics

Normally in chemostat S0 >> KS so:

 KS 
rC ,opt  YXM/S m 1 1
 KS  S0 
 
 KS  S0  KS  KS  S0   YX /S  m S0
M

Usually greater productivity in chemostat than batch system

rC ,opt X 
 ln  m   mt lost
rb  X0 

Typically Xm/X0 10 to 20 & tlost 3 to 10 hours

 Example in text, Xm/X0 = 20, tlost = 5 h, µm = 1 h-1, then ratio is 8

Updated: October 17, 2017

8
John Jechura (jjechura@mines.edu)

4
 10/17/2017

Productivity – Chemostat with Recycle

Cell balance:
 net  D 1  1 C  

Substrate balance in growth phase:

D  S0  S  g  S0  S  YXM/S
  X1 
X1 YXM/S 1  1 C 
X2  1  1 C   X1   S0  S  YXM/S

Productivity is DX2:

rC  DX2  D  S0  S  YXM/S

Updated: October 17, 2017

9
John Jechura (jjechura@mines.edu)

Productivity – Chemostat with Recycle

Monod growth kinetics:

KS g KS D 1  1 C  
S 
 m  g m  D 1  1 C  

rC  D  S0  S  YXM/S
 KS D 1  1 C    M
 D  S0   YX / S

 m  D 1  1 C   

Mathematically similar to non-recycle case

D  KS D  M
rC   S0   YX /S where D  D 1  1 C  
1  1 C    m  D 

Updated: October 17, 2017

10
John Jechura (jjechura@mines.edu)

5
 10/17/2017

Productivity – Chemostat with Recycle

Washout dilution rate

mS0
  Dmax 1  1 C   
Dmax
KS  S0
Optimum at drC/dD =0 also drC/dD’ =0

 KS  m  KS 
   m  1
Dopt   Dopt   1 
 KS  S0  1  1 C   KS  S0 

X2,opt  YXM/S  KS  S0   KS  KS  S0  
 

D  KS 
rC ,opt   1   KS  S0   KS  KS  S0   YX /S
M

1  1 C   KS  S0 
Updated: October 17, 2017
11
John Jechura (jjechura@mines.edu)

Productivity – Chemostat with Recycle

Updated: October 17, 2017

12
John Jechura (jjechura@mines.edu)

6
 10/17/2017

Productivity – Multi-Stage Chemostat

1st reactor looks like a single reactor.

Great deal of flexibility in operating the
2nd reactor
 Additional substrate added?
 Different concentration added?
 Different volumes?
Remember that the downstream material balances must incorporate
the cell mass entering from the upstream reactors
Productivity must include the contribution of both volumes
 F 
N X  F2X2  r2stage   2  X2
 V1  V2 

Updated: October 17, 2017

13
John Jechura (jjechura@mines.edu)

Productivity – Multi-Stage Chemostat Example

Does order make a difference?

 Sometimes
Example – two vessels, 800 L & 200 L
10 g/L substrate @ 100 L/h
Max yield 0.5 g cell mass/g substrate
Monod growth, µm = 1 h‐1 & KS = 0.75 g/L
1st reactor, 800 L:
g ,1  D1  100 / 800  0.125

S1 
KS D1

 0.75 0.125  0.1071
 m  D1 1   0.125
X1   S0  S1  YX /S  10  0.1071 0.5  4.946

Updated: October 17, 2017

14
John Jechura (jjechura@mines.edu)

7
 10/17/2017

Productivity – Multi-Stage Chemostat Example

2nd reactor, 200 L:
 X   4.9464 
g ,2  D2 1 1    0.5 1 
 X 2   X2 
 S S2
g ,2  m 2 
KS  S2 0.75  S2
X2  YX /S  S1  S2   X1  0.5  0.1071 S2   4.9464

Solving iteratively:
g ,2  0.00516, S2  0.00389, X2  4.9981
F 100
rC  X2   4.9981  0.49981
V1  V2 800  200

Updated: October 17, 2017

15
John Jechura (jjechura@mines.edu)

Productivity – Multi-Stage Chemostat Example

What if we change the order?

1st reactor, 200 L:
g ,1  D1  100 / 200  0.500

S1 
KS D1

 0.75 0.5  0.75
 m  D1 1   0.5
X1   S0  S1  YX /S  10  0.75 0.5  4.625

2nd reactor, 800 L:
 X   4.625  
g ,2  D2 1 1    0.125 1  
 X 2   X2   g ,2  0.00929
 S S2 
g ,2  m 2    S2  0.00704
KS  S2 0.75  S2  X2  4.9965
X2  YX / S  S1  S2   X1  0.5  0.75  S2   4.625


Updated: October 17, 2017
16
John Jechura (jjechura@mines.edu)

8
 10/17/2017

Productivity – Multi-Stage Chemostat Example

What if we make the reactor sizes different?

1st reactor, 900 L:
g ,1  D1  100 / 900  0.1111

S1 
KS D1

 0.75 0.1111  0.09375
 m  D1 1   0.1111
X1   S0  S1  YX / S  10  0.09375 0.5  4.9531

2nd reactor, 100 L:
 X   4.9531 
g ,2  D2 1 1   1.0000   1  
 X 2   X2   g ,2  0.00872
 S S2 
g ,2  m 2    S2  0.0066
KS  S2 0.75  S2  X2  4.9967
X2  YX / S  S1  S2   X1  0.5  0.09375  S2   4.9531


Updated: October 17, 2017
17
John Jechura (jjechura@mines.edu)

Productivity – Multi-Stage Chemostat Example

What if we switch the order of these reactors?

1st reactor, 100 L:
g ,1  D1  100 / 100  1.0

S1 
KS D1

 0.751.0   
 m  D1 1  1.0 
This is a washout condition – this configuration will not work
properly!

Updated: October 17, 2017

18
John Jechura (jjechura@mines.edu)

9
 10/17/2017

Summary

Definition of “productivity” gives us a quantitative means to decide

how to size & operate a bioreactor
 Not just “what can be done” but also “what should be done”
Some of the simple configurations have simple relationships for
optimal productivity – some are much more complicated

Updated: October 17, 2017

19
John Jechura (jjechura@mines.edu)

Supplemental Slides

Updated: October 17, 2017

20
John Jechura (jjechura@mines.edu)

10
 10/17/2017

Other Configurations – Multi-Stage Chemostat

Updated: October 17, 2017

21
John Jechura (jjechura@mines.edu)

Material Balances – Multi-Stage Chemostat

Only growth & X0 = 0

Cell balance – transient to steady state
dX1 F
V1  FX1   net ,1X1V1  g ,1 
dt V1
dX2  F  F   F   X1 
V2  FX1   F  F  X2   net ,2X2V2  g ,2       
dt  V2   V2   X2 
Substrate balance – transient to steady state

V1
dS1  X
 FS0  FS1  gM,1 1 V1 
 F V   S0  S1   g ,1X1
dt YX /S X1 YXM/ S

V2
dS2  X
 FS1  FS0   F  F  S2  g ,2M 2 
 F V2  S1  S2    F V2   S0  S2   g ,2
dt YX / S X2 X2 YXM/ S

Updated: October 17, 2017

22
John Jechura (jjechura@mines.edu)

11
 10/17/2017

Material Balances – Multi-Stage Chemostat

For Monod growth

m S1 g ,1KS
g ,1   S1 
KS  S1  m  g ,1
 m S2 g ,2KS
g ,2   S2 
KS  S2 m  g ,2

… where µg,2 is more complicated than just the dilution factor

Updated: October 17, 2017

23
John Jechura (jjechura@mines.edu)

12