Highly Exothermic Reaction Temperature Control 3

[Camarena]

Hello,

I´m wondering how to eliminate in an efficient way the heat generated during a highly exothermic gas-liquid-solid reaction, with this limitations:

·The reaction takes place at low temperatures (not higher than 50ºC) and high pressures (80-100 bar)

·Temperatures are needed to be kept below 50ºC to avoid dissolvent evaporation with the risk of flammable mixtures formation and the reaction of dissolvent with the gases present in the reaction, leading to new gases formation and dissolvent decomposition, as well as, selectivity losses.

·Corrugated plate-type heat exchangers cannot be used since they will not resist this pressure (as I have seen in different plate-type heat exchanger manufacturers), and high plate thickness will be needed to resist the pressure, which results in heavy and expensive heat exchangers.

·Shell and tube heat exchangers will offer low U (mean overall heat transfer coefficient), since it will be a gas-liquid mix flow, which means that high areas will be needed.

·In case, water is used as the refrigeration fluid, large amounts will be needed owing to the low delta T that is available (temperatures not higher than 50ºC), which results in high expenses in refrigerating water.

So these are my queries:

Is there any way to eliminate the heat of reaction in an efficient way, in case the reaction takes place in a slurry reactor, what industrial alternatives or devices exist for this case and which will be more suitable?.

Which are the maximum U values that can be reached in shell and tube heat exchanger for gas-liquid mixtures?.

Does anyone know any plate-type heat exchanger manufacture which exchangers can resist this high pressures?.

In case any other refrigeration fluid is used (for example glycolated water). Where could I find some information about this devices and their maximum refrigeration capacity?

I will appreciate any bibliographic or web reference dealing with any of these matters, the same way as your own experience.

Thanks in advance,

 

[StoneCold]

Camerena
We do something similar to this all the time. We have batch reactors that we cool with cooling water (tower water)and it is very easy to stay below 50C. We perform the reactions in a semi-batch manner, meaning that reagent A is in the reactor with solvent and reagent B is fed to the reactor at a controlled rate to balance the heat of reaction with the rate of cooling. What reactant is fed and what reactant is in the solvent depends alot on the chemistry. Coupling reactions and undesireable side reactions can occur if you choose the wrong one.
A nice way to run the reactions is to use the boiling point of the solvent as the temperature limiter then you condense the solvent vapor and reflux it back to the reactor. We use shell and tube exchangers for our condensers which I believe is very common. We don't ever use plate exchangers due to gasket incompatabilities. We have not tried welded plate exchangers maybe some one else has. We don't have pump around loops on our reactors to speed up the reaction rates that can be achieved but I have seen that done in several batch reactor articles.

If you are going to run continuously you will have to take a different approach but I hope this helps some.

Good Luck
StoneCold

 

[jalvarez]

Camarena:
Maybe brazed heat exchangers could be the right solution; try Linde, AlfaLaval or Heatric. PCHEs from Heatric? See

http://www.heatric.com/pches.html

A spiral wounded heat exchanger should be the next option (Linde).
All these alternatives are expensive, I need to say. If it is possible for your case, StoneCold is right in his approach, we use a similar one...
Have a safe day
J. Alvarez

 

[Fzob]

Have you considered a jacketed reactor with a cooling circuit.

Meaning; you have a jacketed reactor that circulates water in a closed loop though a jacket and a heat exchanger. The jacket side stream absorbs the heat of reaction and rejects it into a heat exchanger located in the same circuit. The other side of the heat exchanger is a chilled or tower water supply. This way the heat exchanger and jacket can be designed to operate at lower pressures, and the reactors shell is the only one designed for the high pressure service. If the CT is a low enough temp, you don't need an additinal HX just run the tower water directly through the jacket. The control system for this can be a bit tricky, but it's done all the time.

We do this all the time for highly exothermic polymerization reactors.

 

[Camarena]

First of all thank you for your answers.

The process I was thinking of is a process where reagents are both in gas phase and are feed continuously and partially dissolved in the solvent to form the product in the liquid phase. The product will be taken continuously from the reactor together with the solvent and some dissolved unreacted gases in the liquid stream, the unreacted and undissolved gas stream is also taken from reactor. So, in this case heat of reaction cannot be controlled by reagent addition rate as in StoneCold case (semi-batch). In addition, as I mentioned, solvent evaporation should not take place in order to avoid flammable mixtures formation with the reagent gases.

J. Alvarez, I have been searching on Alfalaval website and their Complabloc plate-type heat exchangers can only resist around 40 bar, Linde and Heatric seem to be useful heat exchanger for high pressure applications. Do you know any free software for plate-type heat exchangers calculations?.

As Fzob says a jacketed reactor with a cooling circuit would be another possible solution but I’m not sure if this would be a suitable refrigeration configuration in an industrial CSTR reactor where large amounts of solvent and gases are involved and if it would be enough for having an homogeneous temperature in all the liquid volume, just cooling the reactor wall and using stirrers for trying to reach an homogeneous temperature or would it be necessary to use internal refrigeration devices.

My question is if there is any internal device or any other methods for heat of reaction removal in industrial CSTR reactors.? Do you know any web site, article or manufacturer dealing with these devices?
Which is the minimum temperature that can be reached for water in a refrigeration tower?. There is any free software for estimations?.

Thank you again,

 

[StoneCold]

Camarena
Continuous process. That is a big piece of information.
A comment on the flamable atmosphere. Unless one of the feed gasses is oxygen, and if it is that is a whole different game, you would purge the reactor with nitrogen prior to startup and there would be no flamable atmosphere.

I guess my first stab at a reactor design would be just to use a gas/liquid contactor design. Maybe sieve trays where the liquid and the desolved solid run down and the gas is introduced in the bottom. I would control the liquid flow rate and number of stages (residence time) to keep the temperature below 50C.

Depending on the chemisty you can either pump the liquid though a shell and tube heat exchanger and back to the top of the column or on to another column to continue the reaction. Unreacted gas would be recycled from the top of the column using recycle compressors.

You could have cooling loops built into the side of the column like a crude column but that might make the design extremely difficult.

Does this make any sence or am I way off base? Could you give us the types of gases and the flow rates of the materials?

StoneCold

 

[jalvarez]

Camarena:
I tried unsuccessfully to obtain a free soft to design these exchangers a couple of months ago. I didn't get anything from Internet. I finally gave the issue -I mean the technical design- to a heat exchange supplier (Alpha Laval), and I obtained a very satisfactory answer and economic proposal, and fast!!!
Maybe Alpha Laval doesn't fit your pressure requirements, but if you need a particular study for your case, the shorter way should be to go to one of the suppliers I've mentioned, or other, and pass them the issue/problem.
Another option is to contact them and ask for such soft. They should be interested in having satisfied clients, and when a calculation is done, it pointed out a product of their shelve. And I don't see here any violation of a construction secret or a particular technology.
If you don't ask, you don't get.
Good lack, have a safe day
J.Alvarez

 

[Camarena]

Sorry for may delay but I haven been a little bit busy.

I have been searching for industrial processes that deal with this problem and I have found that process for the production of H2O2 from H2 and O2 have this parameters:
High pressure, low temperature and highly exothermic reaction.
This reaction takes place feeding H2, O2 and N2 together with a solvent where gases are dissolved and get in contact with a heterogeneous catalyst. Flammability mixtures can be formed in case of solvent vaporization in presence of H2 and O2, and H2 should be fed below 4 % vol.
As StoneCold points out heat exchanger could be used cooling the liquid stream between reactors in case of serial connection, but you can’t allow high delta T between inlet liquid temperature and outlet liquid temperature (delta T 10-20ºC) so in case of trickle bed reactor, it would be necessary cooling the liquid phase through the reactor to control delta T, so I don’t know if it would be necessary to take the liquid out of reactor pass it through a heat exchanger and pump it again to the next catalyst bed of if there is any industrial device that can be installed inside the reactor (Heatric seams to have like a plate heat exchanger inside the reactor), does any one know about these devices, where could I find more information?. In this case you would have liquid-gas stream to be cooled so U will be lower than in liquid phase heat exchanger and bigger areas will be needed. Which are the bigger U reached on a gas-liquid stream to be cooled? Where could I find typical values of U for mixed streams?.I have found typical values of U for one phase streams, but I didn’t find values for mixed streams.
Any way this configuration seams a little bit complex. So I was thinking on a stirred reactor, but I don’t have much experience on the existing ways of cooling the reaction medium and I would appreciate any reference about this issue.

 

[pandora]

Camarena,

If you are still looking for a heat exchanger try looking at the Platular(TM). It is a half way house between a plate and the tube&shell. They may be able to give you what you want.

http://www.chemicals-technology.com/

contractors/heattransfer/barriquand/