Applicability of the Fire Overpressure Scenario
Applicability of the Fire Overpressure Scenario
(OP)
I am trying to determine if API 521 speaks to how long a fire needs to last to be considered a valid overpressure scenario. In the particular case that I am reviewing, a small 1 gallon vessel (MAWP = 1,800 psig) may be exposed to an external fire if a hexane spill occurs and ignites near the vessel. The issue is that there is only 4-6 gallons of hexane available, which I am guessing won't sustain a fire for very long. Does anyone know if API 521 provides any guidance on how much flammable material is needed to constitute a valid pool fire? Thanks.





RE: Applicability of the Fire Overpressure Scenario
Actual survival time of the vessel depends on the heat flux, exposure time, and material of construction. Refer to paragraph 4.4.1.3 and Figures 1 and 2 in API 521, latest edition.
Dejan IVANOVIC
Process Engineer, MSChE
RE: Applicability of the Fire Overpressure Scenario
I have just gone through this. Look at your area and determine how thick the liquid would be if it pooled. If it's an unconfined area, the liquid will spread to a very thin layer (imagine throwing a bucket of water on the driveway). If the area is confined, such as if there are walls, then it could be a thicker layer. I can't remember the reference, but I believe an unconfined pool height (where it will no longer spread due to gravity) is about 2-4 mm.
Use the burn rate provided by EmmanuelTop to get the fire duration, and the API fire equations to calculate the heat flux. Take the normal operating temperature, then calculate the change in temperature of the vessel contents (mCpdT) over that time frame. If the vapor pressure of the contents at the elevated temperature is less than the relief set pressure, and the liquid swell is less than the vessel volume, then you should be OK for eliminating the scenario.
With 4-6 gallons, I think you'll find a very short fire. However if the vessel is only 1 gallon, and the liquid inside is volatile, then it may be an issue.
RE: Applicability of the Fire Overpressure Scenario
For your specific case, my attention is drawn to the size of the vessel (1 gal), more so than the possibility that the fire will extinguish itself. Does this 1 gal vessel justify a fire sized relief valve? In my opinion, the answer is generally no. The high surface-to-volume ratio means that the heat flux is high. The liquid is going to be gone in very short period of time (probably less than a minute). So, after installing a fire-sized PRV, have you really accomplished much with regard to risk reduction? Again, my answer is no. Once this vessel is empty of liquid, it's going to quickly heat up and fail due to high temperature - the PRV isn't providing any real protection from fire exposure once all the liquid has vaporized. So, this vessel is probably going to fail pretty quickly, regardless of whether you've sized the PSV for fire exposure or not. There are exceptions, but for most cases like this I see little to no value in sizing the PRV for fire, and that's a judgment decision the user can make.
RE: Applicability of the Fire Overpressure Scenario
RE: Applicability of the Fire Overpressure Scenario
Also, what if a nearby larger hydrocarbon vessel ruptures, spills oil near this small drum, and catches fire?
RE: Applicability of the Fire Overpressure Scenario
The original question was for whether or not the scenario is applicable. If the scenario legitimately isn't plausible (due to low inventory), then it makes no sense to size for it. Including it for all vessels in the area arbitrarily would lead to oversized headers, potentially large flare and disposal systems etc. It can most certainly be worth working around in some cases, especially if a flare/oxidizer has already been purchased and is at site.
I agree a fire calculation takes minimal time, but things can get very expensive very fast for something that literally can't happen with minimal inventories / adequate drainage. My two Canadian cents.
RE: Applicability of the Fire Overpressure Scenario
My concern with the original post's statement is that it is not clear that a pool fire could not form from a source other than the protected vessel. In refinery process units it is generally assumed, at least until proven otherwise, a fire can spawn anywhere, regardless of what is in the vessel.