Evaporator design criteria 1

[c2sco]

Hi,

I'm designing a new evaporator as part of a plant modification in the UK. It will replace an existing evaporator that will be too small for the modified process.

The evaporator removes water from organics. It has forced circulation, the external pump pushing a high flowrate of liquid through a steam heated calandria. The process is kept under static and dynamic pressure so that the calandria has no boiling in it, the liquid flashing across a restrictor orifice plate immediately before re-entering the evaporator. It is a vertical cylindrical vessel.

The evaporator acts like a knock-out pot. Previously vapour-liquid separation performance was not critical (and was unknown) as the steam went into a distillation column on another part of the plant where organics were recovered. We can't copy that in the new process and so need to condense and treat the condensate before discharging it. Hence the new evaporator must have good performance in avoiding excessive carry-over. Demister type units are not preferred due to risk of fouling. Ideally we want to make as few fines as possible and then do a good job of removing all but the finest, whilst of course minimising capital cost.

The existing evaporator has a radial entry for the recirculating liquid into the vapour space. It has been suggested that a tangential entry would act like a cyclone and improves liquid droplet separation. However the situation is unlike a typical cyclone, in that all cyclones I've ever seen have a volumetrically much larger gas or vapour flowrate then liquid, whereas this one has a ratio of about 17:1 gas:liquid (ie over 5% liquid by volume). The tangential inlet velocity would be about 20 to 25 m/s (although the bulk liquid might not attain this velocity). The superficial vapour velocity of the gas in the evaporator body assuming plug flow would be 0.7 m/s. The vapour is essentially water vapour at 1 barg, and the liquid is an organics/water solution. The vessel is 2m diameter.

Has anyone come across such conditions in an evaporator or cyclone before? Are there any useful design criteria I can use to decide if the cyclone would carry out a useful separation? I don’t know the droplet diameters. Can anyone offer suggestions to get the best performance for the evaporator design?

Thanks,

Stuart

 

[MJCronin]

c2sco..

Sometimes there is a reason to contract the services of wise, gray-haired older engineers and thier consulting firms. Not all of their judgement, experience and common sense can be conveyed in a few paragraphs on an "engineering hints" bulletin board....

The specification, design, installation and operation requirements of large heat exchangers involves financial risk for both you and your organization

Is this your first design effort for a large heat exchanger ?

-MJC

 

[rmw]

Go to this site and make some contacts.

http://finance.groups.yahoo.com/group/sugartech/

or go to this site;

http://www.sugartech.com/

These guys have experience with Roberts type calandria as well as fallilng film tubular and plate types and PHE plate types.

The falling film plates use the direction change from down to upward flow for exit at the top of the vessel for separation of the large moisture droplets. Chevron vanes do the final mist elimination.

Google "GEA Ecoflex" to find a site where you can see how they do it.

The PHE plate types have to have a separator vessel after similar to your application. They don't want carryover because it contains sugar solids which is product loss as well as a bloody mess.

You might find some information on the Ecoflex site or if not, google "Alfa Laval".

I used to design evaporators that used a similar evaporation stage as yours. We aimed our flash device down so that the vapor had to do almost a 180 degree turn to pass through the mesh ME's at the top of the chamber.

In your case, since you can't use mesh, you will have to use chevron types and maybe several stages if you want to get all the moisture out. Pay attention to the velocities and the face loadings.

I've rambled a bit, but I have done what you want to do as have the guys that lurk on the sites that I gave.

rmw

 

[c2sco]

Hm, thanks for those.
MJC. Actually I am a "wise, gray-haired older engineer" - well maybe not all that wise, as I've not come across this exact configuration before! I'm only looking for real experience hints.
RMW. I'll try the sources you suggested - I've not seen the sugar one before. I'm flushing more experience from past plants in this company now. They had a big improvement about 20 years ago when they swapped a straight radial entry to a sort of single pipe distributor which basically points downwards. I think we'll end up with the tangential inlet design - it's hard to see how it could be worse than any other, it's just that it's not what to me is a "normal" cyclone configuration due to the high liquid load.