9 replies to this topic

[MGruber]

I have read with some interest discussions on whether a relief device is needed on the tube side of an ASME exchanger. I can see where you could do thermal expansion equations for a heater but my application is using chilled water to cool hot Water-for-injection in the tubes from 185 F to 77 F. Is there still a credible scenario for thermal expansion in this case? It could be blocked in. Also, what about a fire case?

Thank you.

[pleckner]

A common system for bio-pharm installations. If the hot stuff is in the tubes and especially if the exchanger is indoors, you don't need to be conerned about thermal expansion relief in the tubes.

By the way, there is nothing in any U.S. code that specifically requires you to put a thermal relief on the tube side of a heat exchanger. I would as good engineering practice if it warrants one. The PSV is usually cheap insurance against the potential of having to repair a heat exchanger, not to mention the down time.

And this goes into the fire scenario. We historically do not protect the tubes from fire as the tubes are pretty well protected by the shell and its contents.

[djack77494]

Phil,
I would take issue with your statement that "We historically do not protect the tubes from fire as the tubes are pretty well protected by the shell and its contents." I agree that you probably needn't consider heat gain into the shell fluid and from the shell into the tubeside fluid. However, you do have heads/channels that contain the tubeside fluid. Assuming a credible fire scenario exists, the channels will be exposed to heat gain due to the fire. Though the exposed area is likely much less than that of the shellside, you cannot neglect this scenario.
Regards,
Doug

[pleckner]

I did a little research again (and to refresh my dying barin cells) on ASME and API and come to the same conclusion; there is no requirement to protect the tube side of a heat exchanger. The only required protection is from an internal failure and this pertains to the shell.

Saying this, there is good engineering practice, which I am all for, and which is why I do include PSVs on the tube side of heat exchangers and will argue that point until the cows come home...for thermal relief. But a fire case? And affecting the heads/channel? I'm afraid that I don't think there is much protection from a tube side PSV if the liquid in the heads/channel begins to boil. The generated vapor will have to go through the restriction of the tubes before it reaches the PSV. I'd bet you would have a failure in the gaskets or even in the heads before then. The liquid would tend to act as a plug as well.

And it is precisely because of this type of problem in the expected response of the PSV that many designers put rupture disks on the shell of the exchangers rather than a PSV. The fear is that even on the shell, the PSV won't operate fast enough to prevent a catastrophic failure.

To be honest, I can't remember if I had specified a tube side PSV for fire in the past, probably not for the reason I state above. I'll reserve that until the opportunity once again arises.

[djack77494]

O.K., and thanks for the education. I can see a lot of reasoning for NOT putting PSV's on the tube side of an exchanger, especially for a multi-pass tube side. (The more passes, the less effective the PSV would be.) Like you, however, I will probably continue to err on the side of being conservative; the thermal relief PSV sounds like a good solution.

In those rare instances when you don't have many tube passes AND the shell side happens to be at higher pressure than the tube side, I do think a PSV could serve a useful purpose in the event of a fire. (Maybe the best approach in those cases would be to just raise the tubeside design pressure.)
Doug

[pleckner]

Doug, we are in agreement!

[sofkadi]

Hi,

I am a process engineer working in Paris, actually i want to know how to determine the scenario for the heat exchanger ( shell and tube ) exposed to fire.
the first case:
- heat exchanger with ( hot oil -dowtherm- in the shell side ) and (oil-water in tube side ) and the design pressure is 15barg on each side.
for this case , what is the realistic scenario to take into account for both sides
the second case
- heat exchanger with (oil with 2% BSW in the shell side ) and (oil-water in tube side) and the design pressure is 15 barg on each side.
for this case , I consider that I have a fire case on the shell side and a thermal expansion on the tube side .
is it realistic ?

because , when i read your explanation for relief case , you consider only a thermal expansion on the tube side. is it right ?

Regards

[pleckner]

You really need to read the previous posts more carefully concerning the tube side PSV and scenarios. The reason the discussion was about thermal expansion is because the scenario is blocked-in, liquid filled tubes. If the liquid isn't blocked-in or the tubes just happen not to be liquid filled, then there is no thermal expanson overpressure scenario as the expanding liquid volume has a place to go.

Now, you have to realize why you are sizing a relief valve for a fire scenario in the first place. The PSV in a fire scenario isn't to protect the equipment; the equipment will be destroyed (or at the very least damaged and requiring some type of repair work)! It is to allow for a controlled release of the vapor generated during the fire. The tubes are immersed in liquid and are already protected. The liquid in the tubes will heat and want to expand (but not boil). As I stated above, as long as the piping is open, there is no thermal expansion overpressure. So, on the tube side, what are you protecting? You certainly aren't protecting the heat exchanger, it's gone after the fire anyways!

So it appears to me that you have a credible fire scenario for the shell side in both of your cases. The tubes don't enter the picture for this fire scenario.

[sofkadi]

Hi,

thank you for your answers, but i am wondering about the shell side protection for my cases,
why is it realistic to put a PSV fire for the hot oil , how can i boil the hot oil, for me I put a TSV on the shell side and i route this liquid to recovery hot oil drum.
what do you think ?

but if I put a TSV on the shell side, what about the tube side

thanks a lot

[pleckner]

There is indeed an issue with fire and hot oil because I guess you can actaully get decomposition of the hot oil before it would boil. I had this problem many years ago for a Lube Oil facility. We ended up using a real low latent heat (can't really remember; 15 btu/lb or so).

Decomposition of the hot oil could be a worse condition than just boiling. The key question is how do you tell if you get decomposition? I would talk to the vendor about this if I was you.

If you are convinced that you can't get boiling and there is no decomposition (which I doubt is a true statement), then yes, thermal expansion is all you would need on the shell side.

I'm not changing my mind, you don't need anything on the tube side. But if blocked-in tubes is a credible scenario, then you can put in a PSV for themal relief on the tube side as good engineering practice.