Shell&Tube Heat Exchanger Question 2

[Paulista]

The simplified theory pertaining to single-pass Shell&Tube Heat Exchangers, both parallel-flow and counter-flow types, is relatively straight-forward.

However, is there a rule-of-thumb approach to evaluating the benefits, if any, should the above basic scheme be modified, for example, by :

adopting a single-pass Shell & a double-pass Tube Heat Exchanger,

with "all other things being equal", viz., dimensions of Shell&Tubes, total n° of Tubes in Shell, mass flowrate of hot and cold fluids etc ?

 

[Montemayor]

Paulista:

There is no short cut, rule-or-thumb to use in optimizing a shell & tube heat transfer application. Yes, I said "optimization" - because that is precisely what you are describing when you want to consider or evaluate "a single-pass Shell & a double-pass Tube Heat Exchanger". There are fouling factors, temperature crosses, allowable pressure drop, etc. to take into consideration and these ultimately affect your final design.

You can't arbitrarily "fix" such factors as "dimensions of Shell&Tubes, total n° of Tubes in Shell" when you are comparing a single shell pass or a double shell pass. The baffle design and shell-side velocities that you may need are just some of the variables that force you to enlarge the shell diameter - or extend the exchanger's length.

Have you done a manual or simulated attempt to evaluate both shell designs? If you have, perhaps you can understand what I'm trying to communicate to you. Perhaps if you state the specific details and data of what you are proposing, we might be able to help out.

 

[Paulista]

Montemayor,
I fully appreciate your position, but if you will let me continue for a minute with my rather crude "seat-of-the-pants" engineering:

Consider a shell length of 1 metre, internal dia of 15 cms and 50 tubes.

It is performing well in a 150 HP marine diesel application,that is, the engine water temp does not exceed 85°C on a hot summer´s day at full load.

However, the 1 metre long shell is taking up precious space in the engine room and you would like to halve it´s length to say 50 cms, but without lowering the heat transfer capacity of the exchanger.

Could this be accomplished by now configuring a 50 cm long single-pass Shell and a n-pass Tube with 50/n (n= n°of passes)tubes each way ?

 

[Montemayor]

Paulista:

Thanks for furnishing the basic data. Now, I can offer some firm and concrete recommendations.

Yours is a good example where you want to seek out and apply a more practical and efficient heat exchanger to cool your marine diesel engine. I presume you're doing the cooling of the engine jacket water with external seawater, but this is of secondary importance for now. The important issue is that you lack sufficient and efficient space to inspect, operate and maintain your engine in top, safe shape. I have had the opportunity to do this very type of modification on boats while working in the Caribbean and we had excellent success.

Space is vital and very important in your application and I totally abandoned the idea of applying a conventional, straight length, TEMA shell & tube type of exchanger on this marine application simply because it just takes up too much room and is physically tough (or sometimes impossible) to clean and maintain. What I did for various of my friends was to replace the conventional TEMA type cooler with a Graham Helifow - a very old and proven type of compact, super-efficient, and unique exchanger that uses an Archimedes Spiral tube configuration to achieve extraordinary cooling using very little surface area. The key is the "spiral eddies" developed inside the tubes - which yield a superior convective film heat transfer coefficient. This, of course, means you will have about 10-20% more tube side pressure drop - but your film coefficient leaps up in efficiency gained. I first used them as reciprocating compressor intercoolers - on pressures up to 5,000 psig and found them to be very efficient, compact, and totally maintainable.

You can visit the Graham Corporation Website and also download Adobe Acrobat documents explaining and detailing how they work and how they are applied. In fact, you can copy the type of fabrication and make one yourself - if you're up to it. They are very simple and easy to fabricate. They are relatively less expensive than TEMA type exchangers and work for years and years. The ones I fabricated 35 years ago are still in service. This, in my opinion, is the way to go in your application. This type of device gives you more than the capacity you need, it is relatively cheap, it is compact-but yet easily dismantable and maintainable. It can be fabricated out of a variety of materials - including SS.

I hope this experience helps.

 

[Paulista]

Montemayor,
Thank you for the tip, I mean eng-tip !

However, are you willing - and able - to take a "seat of the pants" shot at what the value of n (n= ?) should be for the above problem ?

Halving the length of a satifactory heat exchanger, without changing the dia. of the shell, nor altering the total n°of tubes within the latter ?

Thanks

 

[Montemayor]

Paulista:

You have to feed us specific data, if you want us to follow your logic. Give us the specifics on the outside diameter of the tubes you have, the fluid in the tubes, the fluid in the shell, the temperatures of the existing streams (in & out), the TEMA type of exchanger, and the existing heat transfer area.

You ask if "this be accomplished by now configuring a 50 cm long single-pass Shell and a n-pass Tube with 50/n (n= n°of passes)tubes each way"; this statement is confusing in English. Please confirm that you mean to design a shell & tube unit (say, a TEMA BEM type) with a shell length of 50 cm and a diameter that will conform to containing 50 tubes per pass. The shell will have one pass and you want to have sufficient tube passes to do the duty. The diameter of the shell is what you require to evaluate.

If the above assumed translation is correct, you can't say anything about the application without knowing the diameter of the tubes - at least. Otherwise, how do you consider coming up with the diameter and partition plates required?

Please furnish the requested specific basic data and we'll have something concrete to discuss. "Seat of the pants" engineering is for engineers who have run out of options and basic data or don't have any ability to solve the problem except by guessing. If this is a real problem, it can be resolved using real basic data.

 

[25362]

Whenever you add tube passes, while keeping the same shell
diameter, the partitions at the channel (and bonnet) ends would reduce the space available to accomodate tubes on the tubesheets, hence the number of tubes for an originally "full" shell would perforce diminish.

 

[Paulista]

Chemical,
Yes, n° of tubes decreases as you increase n° of passes for a fixed diameter cylinder, as per the reasons stated in your post above.

Say you started with 50 tubes single-pass within a shell of internal dia. = 15cms.

The same space could accomodate, say, 40 tubes, double-pass (n=2), with the tube-carrying capacity of the assembly diminished by 20%.

For n=3, total number of tubes could probably be only 30-32.

Are you now able to make an intelligent guess and answer the original question ?

 

[25362]

A quick seat-of-the-pants assessment follows.

The original question called for equal "total number of tubes in shell", flow rates, etc., and then halving the length because of space restrictions...

Anyway, disregarding any effect on friction drops, a reduction in surface by 50% length-wise and then 20% by tube count wouldn't compensate for an increase in the tube-side heat transfer coefficient (HTC) expected improvement due to higher tube-side linear velocities.

Moreover, the shell-side HT resistance will probably grow due to the "by-pass" flow lanes opened by using less tubes. And, if this resistance is the dominant one, it would cancel any improvement on the tube-side HTC.

Remember that the overall HTC is always lower than the lowest of the individual HTCs.

One alternative that may be followed and could be checked having full details of the present unit, is to split the duty into two equal shorter single tube-pass units, one on top of the other. I assume one could save floor space while keeping performance at about the same level.

 

[Paulista]

25362 (Chemical),
Many thanks.That is what I was looking for.

Cheers

Paulista

 

[abeltio]

have you considered:
1. plate heat exchangers?
2. vertical shell& tube?
In the engine room vertical space is more affordable than horizontal (smaller foot print).

some piping re-arrangement is needed, of course.
cheers


saludos.
a.

 

[Paulista]

Abeltio,
What can you tell me about plate-type heat exchangers ?
How do they fare in reliability, cost and size, in comparison to shell and tube types ?
Are they at all serviceable, should something go wrong ?
Vertical shell and tube is out, as I will end up messing the engine-room sole - from the frying pan into the fire !

Thanks

 

[abeltio]

pound per pound in weight they are much more efficient than shell and tube heat exchanger.
they are used in all the industries from power plants to food processing.

in a power plant we've been using plate-type heat exchangers for cooling water, lube oil with sea-water as cooling medium and never regretted it.
they can be made of corrosion resistant materials.

for moderate temperatures and pressures they give the best kick for the buck.
they are very easy to service.

the only trouble with these HE is that you have to trust the OEM about performance... they never release the calculations.

HTH


saludos.
a.

 

[Paulista]

Thanks Abeltio.

I contacted a manufacturer in the US (Pennsylvania), an OEM supplier, that manufactures brazed-plate exchangers. The guarantee is only ONE YEAR. Does not seem to be field serviceable to me, on the face of it.

They also supply calculations for you to swap from a shell and tube to a plate type HE.

 

[TBP]

Brazed plate HXs are a one-shot deal. The only service you can really do is flush them out or pump a cleaner through them, if they haven't been completely plugged. If they leak or plug, they have to be replaced.

Abeltio - You don't get a calculation sheet from the HX supplier? We rep a HX manufacturers units, and purchasers ALWAYS get a calc sheet. The vast majority of them won't/can't do or check the calculations, but the main purpose is so that they can verify the operating conditions - fluids, flows, temps, pressure drops, etc. It's explained to them that the selected HX will perform as per the calculation, under the conditions stated, and while the calculations are very accurate, they're only as good as the input data. (Gargage in, garbage out.)

 

[Paulista]

TBP,
Thanks for the tip-off !

No wonder the guarantee is only 1 year !