Heat Exchanger Design-Choice fo TEMA type

[djack77494]

I (think that I) understand the factors involved in selecting which side of an exchanger to put the two fluids, and I know the limitation due to thermal expansion of fixed tube design. But that leaves choices between expansion joints, floating heads, and U-Tube designs when temperature differences are beyond fixed tubesheet use. I would appreciate any guidance that can be offered on the relative costs and benefits of these approaches.

Thanks for any assistance you can provide,
Doug

 

[sshep]

djack,

The U-tube exchanger is the cheapest exchanger per equivalent sized unit, cheaper even than a BEM regardless of whether an expansion joint is needed, because there is only one shell head flange. U-tubes (barring an F-shell which you should avoid if possible) do not have the counter current flow advantages possible with straight tubes, so multiple shells are required in some situations.

The choice of which side to use for each fluid usually a based on the nature of the fluid (fouling, corrosive) and the function of the exchanger (reboiler, condenser, etc). Sometimes the choice is made based on the fact that you have more design control of velocity (major impact on heat transfer) on the shell side (baffle spacing). Sometimes the pressure rating can be a factor in choosing sides.

Generally it is nice to have a fouling fluid in straight tubes as these are easy to clean. U tubes can be hydroblasted, but the bend can cause a problem getting clean and if a tube is totally plugged then you may not be able to do anything with it (a straight tube can be drilled). fixed tube sheet exchangers do not give access to the bundle, so the only cleaning option is chemical. Floating head exchangers give the advantage of straight tubes which can be pulled, but are the most costly and shell diameter must be larger in order to pass the rear head tube sheet.

My experience is that an expansion joint will add about 20% to the cost of the fixed tubesheet exchanger, but of course every case is different.

These are my brief comments. Best wishes, sshep

 

[SeanB]

To sshep,

Why do you recommend avoiding F-shells? There seems to be a lot of reluctance to use this design. I want to educate myself on what are the issues. I thank you for your insight.

 

[Longinthetooth]

In an F shell you have to prevent short circuting and it is often very difficult to maintain the longitudinal seal between the long baffle and the shell wall. The seal has to be welded or be a complicated mechanical joint to be effective. Just not worth the expense and trouble.

 

[djack77494]

Thank you for your comments, all. Though U-Tubes are very popular, I've seen quite a few exchangers with expansion joints and floating heads. That leaves me wondering when they would be the best choice. Is it only for severely fouling services? Is the cost penalty that significant? One last question. If I use a type S rear head, can I pull the bundle out the front head? Please comment.

Side note: I concur with what was said about F shells. The seal's supposed to be a real bugger to get right.

Thanks,
Doug

 

[Longinthetooth]

Type S rear heads. Yes, You undo the packed gland and the rear head pulls straight back allowing you to disconnect the internal head. The split backing ring is in 2 halves which you remove leaving the tubesheet which is smaller than the shell diameter. The bundle then pulls through.
Having said that, I've only seen one instalation using an S head. I think they used it to get very low DP on the tubeside (single pass). I dont know of any other reason for doing this. The unit I saw dated from the 1960s and was fairly low pressure. There was an internal expansion bellows on the tube side outlet as well.