Relieving temperature fire case

[hesanchez]

Hi everybody,
Currently I am designing the relief system for a gas plant, my question is about the relieving temperature in fire case.
The vessel I am protecting is protected at the top of the rating 600#, the operationg temperature is 90°F, then I set the design temperature 50°F above (140°F) and the corresponding design pressure is 1432 psig (following the tables of the API B16.5). However, the operating pressure is around 326 psig. The vessel must be designed at rating 600# because there is a chance to feed the plant with additional high pressure streams further.
The whole system is protected upstream by a PSV for blocked outlet case, then, the vessel mentioned must only be protected by fire case.
While determining the relieving temperature (Eq 11. API 521 2007) with a pressure set point at 1432 psig, the result was high enough to realize that when reaching the pressure set point, the vessel was already melted. Then, I decreased the pressure set point (then the relieving temperature was reduced) to find a point where the pressure and temperature coincident for rating 600# met the relieving pressure and temperature.
The results are:
Relieving pressure: 759 psig
Set Point Pressure: 627 psig
Relieving Temperature: 819 °F

With these conditions defined I followed to size the relief load (Eq 12. API 521 2007) and then the orifice size (Eq 2. API 520 Part I 2008) and I realized that at increasing relieving temperature the orifice size gets smaller.
After all this tale, my concerns are:
- The way I determined the relieving pressure and temperature is adequate or is there another better way to do it?
- I don't fully understand the physical effect why the increasing relieving temperature decreases the PSV orifice size.
Thank you in advance for all the help you can give me, I know the subject is kind of confusing so I'll give you further information if needed.

Best regards,
Eduardo Sánchez
Process Engineer

 

[don1980]

There's not much technical support for Eq 12 in API 521, 5th ed.

Unless you have a significant amount of liquid in the vessel, and that liquid will boil at a reasonable temperature (below the yield T of the metal), then a relief valve isn't protecting the vessel from fire exposure. Go ahead and install a PSV for code compliance, but don't pretend that the valve is protecting from fire exposure. Explain this in the relief documentation and inform the owner that the PSV won't do any good during a fire. They should do a risk assessment and consider whether other layers of protection should be used.

 

[LittleInch]

It looks to me as if you have a gas filled vessel which melts before the gas expands to a pressure above the design pressure??

This is an issue seen and addressed before and see the thread below for some very useful info and search gas filled vessel fire case. If you set the PSV at 627 psi, how are you going to operate it at "high pressure" later on. I think you need a blowdown system here or some other way of dealing with the fire case (can the vessel be exposed to a pool fire, can you fire protect it in another way - that sort of thing).

You don't say but I assume you're using ASME VIII for the vessel design - looks to me iike you need a pressure vessel designer to help you through this one.

http://www.eng-tips.com/viewthread.cfm?qid=231963

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[PaoloPemi]

if you know fluid composition and initial (operating) conditions you can estimate the temperature inside the vessel when the PSV opens solving a V-P flash operation,
the procedure is discussed here

http://prodesoftwareapplications.blogspot.com/

in the same way you can simulate the following conditions with a series of flash operations,
in the same page there is the description of a rigorous procedure for sizing / rating a PSV with critical (possibly your case) and two-phase flow.

 

[hesanchez]

don 1980. In fact the relief valve would be a second layer of protection in a gas filled vessel, as long as the applied standards require a depressurization system linked to the fire & gas system. However, in such standards it is also required the installation of relief valves in ASME VIII vessels, then I'm finding out the best way to get an useful protection using a relief valve.
LittleInch, it is not expected to change operational conditions in the short term, it is only a possibility in the future, then, in that case, the valve should be validated for the new conditions.
Thank you for your answers.

 

[don1980]

hesanchez - for fire exposure to a vapor filled vessel, a PSV is essentially useless in protecting the vessel. As a layer of protection, it's essentialy no more effective than crossing your fingers.

The vessel temperature is rising unabated. The only thing happening inside the vessel is vapor thermal expansion. Unless the starting pressure is close to the PSV set pressure, the PSV probably won't even open before the vessel fails. Even if the PSV does open, it releases a trivial amount of heat before reclosing, meanwhile, the temperature continues to climb. From the perspective of the vessel's time-to-failure, there's no noticeable difference between an installation with a big PSV and one with a samll PSV. So, there's no value in us discussing relieving temperature. Spend your time focusing on layers of protection that can actually do some good (auto depressurization, water spray, fire-resistant insulation). Explain all this to the clients so they can make an informed risk management decision. Too often, engineers are either unaware of these facts or they just sweep this risk under the rug. That's not good.

 

[zdas04]

Good advice above. For an API 521 analysis, I look for a credible scenario that maximizes liquid inventory in the vessel. Then I do a boiling calc to see if turning all the liquid to vapor results in overpressure. If not, then fire case is not credible. The real gas law just doesn't have enough hp to significantly raise pressure before the metal flows.

David Simpson, PE
MuleShoe Engineering

"Belief" is the acceptance of an hypotheses in the absence of data.
"Prejudice" is having an opinion not supported by the preponderance of the data.
"Knowledge" is only found through the accumulation and analysis of data.
The plural of anecdote is not "data"