PSV sizing for liquids with high boiling points 4

[jmo83]

Hi,


Is it appropriate to size a PSV for liquids with high boiling points assuming thermal expansion only in a fire scenario? (Given the following)
BP = 400 deg F (at 14.7 psi)
Set pressure PSV= 75 psi
Flow pressure PSV = 82 psi (110%)
Tank = 30,000 USG of product, 98% full

Thanks in advance for your replies,

 

[Latexman]

Based on the limited info. given, that's quite a quantum leap to that conclusion.

What's the flash point?
Anything else in area that will burn?
What about other scenario's?


Good luck,
Latexman

 

[jmo83]

Latexman,

Thanks for the super fast response!


What's the flash point?
- 212 deg F

Anything else in area that will burn?
- It is assumed that a pool fire will engulf about 80% of the tank's surface area. Nothing else that will burn.

What about other scenario's?
- The tanks are for storage and transportations purposes only. It is assumed that a pool fire (torch fire?) will be the worst scenario.

My thoughts:
- The tank being 98% full will more than likely start by releasing liquid trough the PSV in a fire scenario due to thermal expansion ( 0.00086 coef TE)
- The BP will increase since the PSV setting is at 75 psi (not sure by how much).
-Lastly I have a hard time believing that the product will actually start vaporizing at a high rate, so that the tank will fail due to internal pressure (Burst pressure of the tank is 500psi). I think that a thermal tear will probably occur before the internal pressure of the tank is high enough to damage the tank.


Is there a rule of thumb for sizing the PSV for vapor or liquid release based on the boiling point of the product?

Regards,

 

[don1980]

Whether it's reasonable to design for fire exposure to a high-boiling fluid depends on the metallurgy of the vessel and the fluid temperature at relieving pressure.

It's not customary, or required, to specify the vessel MAWT based on fire exposure. During a fire, the wall temperature can exceed the MAWT, but it's up to the user to assess the resulting risks. For example, suppose you have mineral oil stored in a 150 psig pressure vessel. If the PSV is set at 150 psig, then it's not reasonable to expect the vessel to remain intact while the mineral oil boils at that pressure. In fact, it may not be reasonable to expect the vessel to remain intact while mineral oil boils at 20 psig. For CS, the yield stress drops off steeply when the temperature exceed 800-900F. That mineral oil tank probably can't be defended from fire exposure. The vessel will heat up and fail. A relief device is necessary for code compliance, but there's no point in sizing it for fire if the relieving temperature is higher than say 900F.

If you can't tolerate the failure of vessels like that, during a fire, then you need to apply other means of protections. In such cases, water spray, fire resistant insulation, and/or auto-depressurization will buy some time and possibly prevent failure.

For the vessel you mentioned (NBP = 400F, Set P=75 psig), I'm guessing the relieving temperature is less than 800-900F. If so, then it's reasonable to size the PSV for fire, assuming there's something in the area that can cause a pool fire around this vessel.

 

[jmo83]

Don1980,

Thanks for your reply, it is truly enlightening!

The tanks are made of carbon steel and as you stated they will more than like fail at about 800⁰F. (by the way you were dead on the releasing temperature = 753⁰F)

The goal is to prevent the tank (and its surroundings) from a catastrophic failure (meaning that the internal pressure of the tank will keep on rising until the tank explodes due to the PSV not being able to release enough vapor-pressure (assuming the liquid boils).
So the dilemma is the following:

1. Size the PSV for vapor release, since it will result on biggest PSV?
2. Size the PSV for liquid release? (I will not feel confident sizing only for liquid release)
3. Size the PSV for ????? (some kind of average situation that takes into account both liquid and vapor release)

Do any of the current codes “API, ASME” make any reference as of how to address high boiling point liquids when sizing the PSV for fire scenarios, where overpressure inside the tank is a concern?

Also some of the tanks have PSVs set to 165 psi (182 psi flow) which really makes thinks more confusing since the critical pressure is 196. Any thoughts?

 

[TD2K]

How long will it take for a fire to heat the contents up to set pressure? Depending how long it will take and the ability to fire a fire, you might decide that a vapor fire relief isn't a credible cause. If that is the case, I would size the PSV for the liquid swell (essentially a thermal relief) though you may have to adjust the sizing if the liquid will begin to flash at some point during this period.

The problem is how long do you take before you eliminate a fire? If it's 8 hours, I'd say you are pretty good. If it's 20 minutes, you need to think about how to protect this vessel such as depressuring. I think API 521 talks about 2 hours which I've seen referenced in some client standards as a reasonable time before significant firefighting is in progress. At that point, hopefully there is enough water coming down that any remaining fire is adding only a fraction of the sizing fire case duty. I'd also look at how big a fire case PSV is and make a judgment call.

It's a case that is always difficult to handle. If it's a hydrocarbon, you likely are going to start getting cracking much over 700F and how do you quantify that for the amount of lights produced and a MW/other physical properties.

 

[moltenmetal]

This problem is similar to the problem for gas filled vessels, except that the consequences of a failure are less dire.

You may need to size only for the liquid thermal expansion relieving rate based on the comments above, but the installation of a non-reclosing relief device, or a thermal relief device, are options worth considering.

 

[don1980]

jmo83 - when you're dealing with a vessel containing high boiling liquid, all-vapor, or a supercritical fluid, a re-closing device (e.g. PSV) doesn't offer any protection from fire exposure. My main point is this.....if such vessels need protection from fire exposure, then forget about the PSV and focus your atention elsewhere. These vessel still needs a PSV for code compliance, or for other scenarios if they exist, but for the fire scenario a PSV won't do any good, regardless of it's size. So, don't waste time worrying about how to size it for fire exposure (the three question s you asked). Look at your non-fire scenarios and size for the worst of those. If there are no other scenarios, then there's no basis for sizing - it's arbitrary. In such cases I simply pick a PSV based on the minimum allowable nozzle size for the vessel (typically 1" or 1.5"). In the documentation I clearly state that there are no sizing calculations because there is no sizing basis.

So, for cases like this, forget about how to size the PSV for fire. Instead, focus your attention on other protective measures - those that are truly effective. Refer to the three protective measures I mentioned above in my previous post.

Be aware, those three protective measures won't necessarily save the vessel from a long-term fire exposure, but they will buy some time. In addition to buying time, depressurization can significantly reduce the consequence of a failure if/when it does occur.