Cooling water Demands for Bio-reactor

[zamu2010]

Hi Guys,I am looking into the cooling water requirement for a bio-reactor where fermentation is taking place. Since the heat is generated in the fermentation reaction, I will have to keep the temperature to the desired value. Any suggestion/recomendations?Thanks in advance for your input on the issue.
Br.ZAMU

 

[duncan8943]

does your reactor have a cooler? D-Pipe? or What? How large is it and how much heat do you anticipate removing? Desired temperature? cooling water temperature? You really need to provide a little more detail.

 

[zamu2010]

Hi Duncan8946,
Thanks for your reply.
The Reactor is 500 litres and where the fermentation of C6 will take place.
The reactor is equiped with cooling water sytem in the outside jecket of the reactor.
Heat production is 23500 cal per mole of suger consumed and the reactor has 300 lire of suger.
The desired temperature to be mentained is 50C and cooling water is coming in at 13 C.
The major concern is heat of fermentation. I dont have the exact data with what rate the heat will be generation and how much cooling is required. temperature above or below the desired temperature has a great effect on the yield of the product. So it very critical to control the temperature very close.
Any suggestions on the issue.
Thanks & Regards
ZAMU

 

[FrancisL]

If you do not know how much heat will be generated you have no chance of sizing the cooling system!
And if you oversize it then it will become hard to control accurately. Back to the lab!

Francis

http://www.controldraw.co.uk

http://www.s88control.blogspot.com

 

[duncan8943]

You're saying that you don't have any data on the rate of heat generation. It looks to me like you need kinetic data on your fermentation. You can either find it where someone else has done the lab work or do an experiment with a small reactor. Fermentation of C6 data should be easy to find. Once you get this and determine your rate of heat generation, then you can fill your vessel with hot water and determine the heat removal rate (or ultimately U value).

 

[zamu2010]

Hi Duncan8943,
thanks for your reply.
Best available data to date is from Bouffard 1895 whci gives the no of 23500 cal per mole of suger consumed as the heat of reaction.
The main problem the rate of heat generation!!!
Is not unform over time spam.
Any recomendations....

 

[zorn]

Dear Zamu, is the cooling jacket already designed and built or do you need to design the whole system?

 

[zamu2010]

Hi Zorn,
Thanks for your reply.
Yes , reactor is equiped with cooling water jacket.
I presume to avoid under design, in the calculations all the heat generation should be absorbed with cooling water.
I hope it clarify the issue....

 

[zorn]

Just another question Zamu: is the reactor continuously or batch operated (i guess continuously)?

 

[zamu2010]

Precisely, Its continuous process....

 

[zorn]

Here it is an idea for having a very first approx of the solution that you can use for more precise calculations:

Hypotesis:

1)water like system (cp,k,d water like)
2)simple Michaelis Menten cinetic enzyme-substrate expression
3)CSTR reactor (well mixed)
4)fixed U derived from tables: 200-500 w/(m2°C)

known quantities:

Feed to the reactor
Exchange surface A
T_cin (cooling temperature in)
T_r (reaction temperature)
Volume V (wet surface)
DeltaH (heat of reaction)


HEAT TO REMOVE:
E=DeltaH*r=W*cp*(T_out-T_in) (1)
E=DeltaH*r=A*U*DeltaT_log (2)
DeltaT_log=
=[(T_r-T_out)-(T_r-T_in)]/ln[(T_r-T_out)/(T_r-T_in)] (3)

cp: water specific heat
W: cooling water capacity
r (mol/time): rate of reaction in Michaelis Menten form:

r=(v_max*cs)/(Km*cs) (4)

where cs is the glucose concentration, Km is the M-M costant and v_max is the maximum rate of reaction

r is evaluated at stationary state: r(cs_ss)

cs_ss is found numerically letting the material balance go to zero:

dcs/dt=Q/V*(cs_in-cs_ss)-(v_max*cs_ss)/(Km*cs_ss)=0 (5)

Km and v_max can be found here for the alcohol fermentation:

http://www.springerlink.com/content/tru96p7427476210/

Q is the volumetric feed

(they're at 20°C, but i couldn't find anything better)

once you have cs_ss you can evaluate r(cs_ss) and evaluate DeltaH*r(cs_ss) (heat generated) (6)

now we can find an approx solution for W and T_out:

- fix W and evaluate T_out in eq. (1)
- evaluate DeltaT_log in eq. (3)
- evaluate E=A*U*DeltaT_log
(with U fixed,for example at 350)
- check if the value |(E-DeltaH*r(cs_ss))/E| is less then a value you chose (for example 0.1)
- if not, choose another W and try again and so on..


i hope it helped somewhat

 

[zorn]

dcs/dt=Q/V*(cs_in-cs_ss)-1/V*(v_max*cs_ss)/(Km*cs_ss)=0 (5)

 

[Japher]

I would suggest setting up a recirc loop for the reactor through a heat exchanger. You can then control the recirc flow rate as a response to the heat removal required. This would be nice, especially since you a) don't know the heat of reaction, b) the heat released will probably not be constant, and c) the recirc will help with the agitation.