Time to fire driven relief 1

[hollerg]

I have an ASME code vessel running 85% liquid full, which fire exposure would result in the PRD lifting. The material would be just below the pseudocritical condition during the relief. Downstream there is and an undersized flare, when I compute the valve size by API-521 with the minimum insulation credit. It’s a 4000 gallon vessel with adequate drainage. The valve is grossly oversized, by API-521 with the minimum insulation credit. Depressurization rate would be too high for the flare.

Although I can mitigate the required flow and could downsize the valve by additional insulation, it’s not enough to impact the fire case load from the fire pool. However, if I can insulate and make fire a non credible scenario & not get the wall hot enough to need depressurization, then the existing valve can stay and the flare system meets requirements.

1. Is it appropriate to insulate sufficiently that the insulation prevents a fire driven release for what is declared to be beyond a “reasonable” fire duration? E.G.

a. Assume -- Proper installation of sufficient insulation
b. Assume – HAZOP review agrees that the fire would extinguish before the PRD would lift

…. then does Fire Case become a non-credible scenario, and the PRD would be sized to meet any remaining credible scenario?

2. IF there were no other credible scenario, then could the PRD be completely omitted?
3. Is the method in the spreadsheet an adequate estimate of the time to relief?
4. Do I need to perform another calculation to examine the wall temperature criteria, or is it safe to assume the wall temperature is the same as the temperature when the PRD is ready to POP?

 

[RaRo]

1. Yes, at least in Europe, provinding that:
0. your "reasonable" fire duration is at least one hour (recommended) and covers: fire built-up, detection, emergency response team arrival, first emergency assessment & response, evacuation of personnel & population potentially affected in pressure vessel/system blast, safe evacuation of the emergency response team to safe location against pressure system blast
1. you do not consider the fire wil extinguish itself
a. your insulation is fireproofed/fireproofing (seea API521 requirements and other sources)
b. your risk management philosophy agree with the principle of letting it burn (and explode) to the ground
2.

2. Yes, provided 1 is acceptable
3. you should consider the liquid thermal expansion (& that of the gas if there is no venting) you have the heat absorbed by the liquid from API521, do not use any additional transfer coefficient/DTLM (minor impact but conservative) t heat-up= M*cm*(t2-t1)/Q ; if thermal expansion is significant, break down the calculation to consider the wetted area change (and so the Q change) over time
4. if unwet wall is within the flame impigement, and your liquid has a high boiling point (preventing cooling down by evaporation long enough for the wall to get very hot), then you should. With full wall fireproofing or evaporating liquid, you probably do not need to.

 

[don1980]

All pressure vessel codes leave it to the user to determine the sizing basis for the relief device. That includes the decision of whether fire exposure is a risk of overpressure, and whether the relief device can actually defend the vessel from a credible risk of fire (in lots of cases it can't).

So, if your assessment concludes that fire isn't an overpressure risk, and the equipment owner agrees, then document that decision. That's acceptable.

With one exception, codes (ASME and European PED) still require overpressure protection (relief device), even if there are no credible causes of overpressure. The exception is in ASME jurisdictions you have the option of omitting the relief device per UG140. I have not found a European code that has an option equivalent to UG140 - I don't think there are any.

 

[RaRo]

The European PED, so far as I understand it, does not require the systematice relief devices. It just gives some requirements if you fit one.
The basis of safety is left to the user. If you are not handling any flammable/reactive materials, you can rely purely on having design pressures (MAWP) greater than the maximum pressure you could get in your system, considering failure modes.

 

[don1980]

RaRo - I investigated this in the past, searching European harmonized codes for justification to apply that logic in European facilities, but didn't find anything equivalent to ASME UG140. Like ASME Sec VIII, European application codes lists the options (PSV, POSV, disk, CSPRS, etc) for protecting from overpressure, but none have a provision for protection by system design.

Your comment above got me thinking that maybe I'm looking in the wrong place. Perhaps the provision allowing this practice is found in the PED Articles or Guidelines. I searched the Guidelines (interpretations) and didn't find anything, but in the Articles/Annexes I found what appears to be the answer, confirming your comment.

Annex 1 (Essential Safety Requirements) point 2 says "The essential requirements laid down in the Directive are compulsory", but the next sentence says, "The obligations ..... apply only if the corresponding hazard exists for the pressure equipment in question...".

Annex 1 point 2.10 further underscores this point.

Thanks you very much!