Fire case validity for a tank in a remote location 1

[Motrack]

Hi folks,

I'm evaluating the need for fire sizing the PSV on a condensate tank that is located away (~1000 ft) from the general processing area. The closest source of a fire is 40 feet away. There are two other tanks that have their own containment. There are no other sources for a flammable pool of liquid to form / drain directly underneath the condensate tank. Any thoughts on whether liquid pool fire sizing would be a credible case?

Thanks.

 

[don1980]

There are two parts to this evaluation. One is to determine what shall be done for legal compliance and the other is to determine what should be done to satisfy the company's risk tolerance. For US facilities, the shall part is defined by OSHA (1910.106). Tanks containing liquids with a flash point < 200F shall have a fire-sized relief device. When you say "condensate" I assume you mean water and not petroleum condensate. If so, there's no legal requirement for fire relief. Next, assess the specific case from a risk perspective - assess whether fire relief is needed, even though it's not legally required. That's a separate question from the first one, and the answer depends on the details of the specific case. If flammable liquids are spilled from the tank 40 feet away, will this liquid flow toward the condensate tank? If so, is there enough flammable liquid to sustain a fire for the time it takes to boil the contents of the condensate tank? That will tank a number of hours - depending on the geometry and volume of the condensate tank. So, your evaluation of this second part is entirely a risk management exercise, and there's no black-and-white answer. Evaluate this based on the company's risk management practices and risk tolerance.

 

[Motrack]

Thank you Don for the response. This is not a storage tank, though. It is a "working" tank with a MAWP of 200 psig. I guess I should have called it a vessel. And yes, the condensate is mostly water, but it may on occasion contain hexanes through decanes.

 

[e43u8]

Motrack.

Is it oily water?

 

[Motrack]

e43u8,

The feed is coming off the high pressure dumps from the Amine, Dehy, and residue compressor systems. About 40% water, and the rest hexanes through decanes.

 

[georgeverghese]

There is some credit you could take for a well drained dyke at this vessel, but a fire case RV seems inevitable, since you could have a flange leak / instrument tubing leak or rupture at this vessel itself -the fluid certainly is flammable, given the composition of the hydrocarbon phase of the feedstream.

 

[mk3223]

First, to name as a "tank" for a MAWP of 200 psig "vessel" was odd. If it a "vessel" per ASME Code , a PSV protection is required by the Code. since it's for the flammable liquid, it should consider a potential fire case hazard. The "closest fire source is 40' away" doesn't mean it won't get a fire close to the "vessel".

 

[Latexman]

If it contains 60% "hexanes through decanes", then "closest fire source is 40' away" is not accurate. The condensate tank and piping could fail/leak to generate a pool of flammable liquid floating on water at the condensate tank itself. Did you mean closest ignition source is 40' away?

Liquid pool fire sizing is definitely a credible case.

Good luck,
Latexman

To a ChE, the glass is always full - 1/2 air and 1/2 water.

 

[Motrack]

If you had a flange leak at the vessel and the protected vessel was the source of the liquid pool, there would not be a whole lot left to relieve, since the contents of the protected vessel drained through the leak. Btw, the vessel does have a PSV which is not fire sized due to the location of the vessel. Any further thoughts?

 

[Latexman]

That's the scenario for a large leak. Which, by the way, is fire exposure to a vessel containing vapors. What about a medium size leak? A flammable pool is formed and there is liquid left in the condensate tank during the fire?

No one knows what size the leak will be, so consider the possibilities and go with the worst credible case.

Good luck,
Latexman

To a ChE, the glass is always full - 1/2 air and 1/2 water.

 

[georgeverghese]

API520 or 521 gives you some credit for the heat rate if you have a well drained dyke, so you could rate the existing PSV for this reduced firecase generated vapor rate due to liquid boiling - a 10% accumulation over vessel MAWP is applicable per recent practices or 20% accumulation per older practice. Use the lower of the 2 latent heats to determine the vapor rate (i.e for light condensate).
If you could model this on Hysis or Pro / II with the firecase relief utility, you'd get even lower vapor generation rates ( when compared to the old fashioned manual calc method).

 

[Motrack]

Thank you for the responses. The credit for a good drainage is, unfortunately, in conjunction with prompt firefighting facilities, which this plant does not have.