Undersized relief valves 1

[Stickers]

Here's an interesting question that I've been thinking about:

What would happen to the pressure in a vessel which is protected by an undersized PSV? Obviously, it would go up but to what value?

For compressible choked flow, as pressure increases mass flow would increase also so, for some cases, it would seem logical that the vessel pressure would increase to a pressure above 10%/21% overpressure. The PSV would then either pass the required flow or a flange/the vessel would fail. Do you agree?

How about in a fire case? As pressure increases, the heat of vaporization decreases and hence the relief load increases. Would this situation be a positive feedback loop where pressure goes up -> relief load goes up -> pressure goes up etc. ?

 

[dcasto]

The pressure would increase until you reached eqaulibrium of PSV capacity and flow being generated. For example, you may reach that eqaulibrium at 120% of pressure for an inflow case to a vessel, not a fire case. The vessel would not be damaged.

 

[Latexman]

My understanding is, when the pressure gets to about 2 x MAWP there is a significant chance of failure.

Good luck,
Latexman

 

[djack77494]

Especially if considering a fire case, there may be a significant possibility of vessel failure. Not only would the pressure be greater than expected (>121% of design pressure), but the temperature would also be higher because saturated vapors at the higher pressure would be hotter. Adding to this is the lower heat of vaporization which would mean even more vapor is generated, If the PSV undersizing were significant, I think you could have a real problem. (That's why we're typically quite conservative in PSV sizing.)

 

[Stickers]

Thanks Guys.

Any tips for calculating what pressure the vessel would rise to? The case I'm struggling with is a fire case. This is tough since the relief load increases with falling heat of vaporization as the vessel heats up.

 

[iainuts]

Assuming you're in the US: If it's a fire case, CGA S-1.3 2005, para. 6.3 is applicable. It gives a very simple equation for the flow capacity needed for a relief device on a tank exposed to fire. If the relief device doesn't meet that, don't waste your time figuring out what pressure the vessel will go up to. It doesn't meet the code, so you'll need to enlarge the relief system.

You can purchase the code from CGA for about $109.

http://www.cganet.com/publication_detail.asp?id=S-1.3

 

[iainuts]

You could also check out this link. Starting on page 41, they have some info covering fire cases.

http://www.tycoflowcontrol-na.com/ld/Tech Sem Manual.pdf

 

[Latexman]

It's easier to calculate the nozzle area you need using 1.21 x MAWP as the sizing pressure and the proterties of the single component that gives you the most flow area. If this nozzle area is greater than what you have, you don't meet code. Otherwise, you need to do a dynamic simulation.

Good luck,
Latexman

 

[robsalv]

Our process guys would use process simulation software to determine the answer to the question you're asking. I imagine it would be a painful manual iterative process determining what the system pressure would rise to for a given contingency and given sized SV...

For a non fire case, failure of the vessel is unlikely if the vessel is in as original condition and the system pressure stays within the original hydrotest envelope (most likely up to 150%DP). Failure is still not very likely even if the system pressure goes outside this envelope, but now we're in uncharted territory - any previously blunted undetected defects buried in a weld seam may become active.

Depending on the exact scenario, there is a good chance flange joints will leak before the vessel fails.

If your vessel is in poor shape, with pits, some defects and/or areas of corrosion losses, then a significant overpressure could cause failure.

This is all supposition though.

It's a different story in a fire case. A vessel will fail if the metal temperature gets hot enough for the metal to flow - irrespective of DP or system pressure.

Interesting stuff.

 

[JimCasey]

iainuts got it right. Don't even bother trying to calculate how much overpressure you'll get if the relief system does not meet the sizing criteria.

Most states recognize the ASME Boiler and pressure code as law. Since you apparently know, or even suspect , that the relief system is undersized as defined by the code, you are prohibited from operating the unit. If you are caught operating it by the boiler inspector, your company gets a BIIIG fine and you get to apply for unemployment. If, God forbid, there is an accident related to this, you go to jail. The phrase "Criminally negligent" comes to mind.