Code Compliance for relieving diesel / fish oil vapor 4

[yene]

Hi all,

I'm reviewing a past relief analysis where ~15 vessels relieve to a common header. Contents of 4-5 vessels are fish oil but we are able to assume diesel oil for modelling purposes.

Long story short: the worst-case scenario was assumed to be an external fire scenario, however I'm learning that the calculated relieving flows for the diesel-oil containing vessels were based on formulas prescribed by API RP 521, as opposed to NFPA 30. I'd have expected it to be based on NFPA 30 since diesel oil is, to my knowledge, a combustible liquid.

I'm wondering if anyone can point to any particular reason why diesel oil might be exempt from NFPA 30? Fire exposed area for each of these vessels is in the ~100 sq.ft magnitude. This exemption would be very convenient as the relieving flows are a lot higher when using NFPA 30 formulas, which might present issues with respect to back pressure and vessel internal relieving pressures.

I am starting to think it was an error that it was exempt from the start but just want to make sure I'm not missing something.

Thanks to anyone who can help!

 

[Latexman]

Backpressures (individual PSV outlet pipe and fittings pressure drop to the header) are based on the capacity of the existing PSV at 10% over-pressure. It really doesn't involve a sizing flow rate and it's method at all.

Now, if the PSV's are re-sized, the sizing flow rate and it's method are quite important. If you end up with a smaller or larger PSV, then backpressures will change from the existing PSV.

Header(s) and disposal system capacity is more complex and may depend on sizing flow rates and it's methods, plus the method used to determine the maximum load to the disposal system is important.


Good Luck,
Latexman

 

[The Obturator]

As far as I understand, NFPA-30 is for low pressure storage tanks.

Is your issue with storage tanks or pressure vessels ?

If pressure vessels, then the code should be ASME VIII, in which case API-521 should be used.

Note also that the PRV's should be sized on 21 % overpressure. 21 % will also be the maximum limit of built up back pressure should conventional PRV's be installed.

(You have also double posted the same issue in the chemical plants forum - double posting is a no-no. Please delete)




*** Per ISO-4126, the generic term 'Safety Valve' is used regardless of application or design ***

*** 'Pressure-relief Valve' is the equivalent ASME/API term ***

 

[don1980]

For pressure vessels, regardless of the contained liquid, you can use either API 521 or NFPA 30. Both are perfectly acceptable RAGAGEP-compliant standards for calculating the heat-input from fire exposure. NFPA 30 states that a fire-sized relief device is required if the liquid has a flash point of less than 200 F (93C). API 521 has no such requirement - it leaves combustibility guidance (what fluids are likely to cause fire) entirely to the user. Note that ASME VIII does the same - it has no guidance for what causes the fire, nor does it tell you how to calculate the fire heat input. Those decisions are intentionally left to the user. For pressure vessels, some companies use NFPA 30, but most use API 521. They're both equally acceptable. The API 521 method will almost always result in a lower heat input (smaller relief device) if the vessel is insulated with a fire-resistant material. NFPA-30 will usually result in a lower heat input if the vessel has an approved (NFPA-15 compliant) water-spray installation.

For low-pressure tanks located at a US facility, there's a legal requirement (OSHA 1910-106) that mandates which tanks must be equipped with a fire-sized relief device, and how to size that device (heat input calculation). These requirements were lifted directly from NFPA 30 (and same 200F criterion for mandated fire relief). So, if you're using NFPA 30, then you're compliant with the OSHA reg. The NFPA/OSHA heat input equation is also identical to that in API 2000 (exact same equation). However, API 2000 is different from NFPA/OSHA in the use of heat reduction credit factors. Specifically, API 2000 doesn't have a water-spray credit while NFPA/OSHA do, and API-2000 uses a calculation to determine the insulation credit while NFPA/OSHA use a fixed value (0.30). For those of you in the US who are using API 2000, there's no need for concern because that's a globally recognized standard, and it's the one most commonly used in the US. I think there's very little risk that an OSHA inspector would cite someone for using the credit factors in API-2000, despite them being different from those in OSHA 1910.106.

There are no requirements in Europe (PED jurisdiction) which mandate a fire relief design. That's true for pressure vessels and low pressure tanks. The decision to design for fire exposure is a risk decision for the owner to make. And, to my knowledge, that's true in other countries too (outside the US). I'm not saying that there are no other countries that prescriptively require fire relief design. Just saying that if there are any such requirements, I haven't seen them.

 

[yene]

Thanks everyone.

So the theme seems to be that API RP 521 can be applied. This is quite the discovery as my understanding was always that any flammable/combustible liquid must have its required relieving flow calculated using the NFPA 30 formulas. Is this really not the case? Operating in Canada by the way.

Also in reply to Obturator: These vessels have MAWPs ranging from 15 to ~100 psig so I believe they would be pressure vessels. But I just had a skim through NFPA 30 and can't find any indication that pressure vessels are exempt from it with respect to relief requirements. There is a section that describes construction requirements for pressure vessels being aligned with some ASME Section but nothing on relief requirements being exempt from NFPA 30.

 

[The Obturator]

15 PSIG Is where ASME VIII kicks in. Anything below 15 PSIG is considered by some as low pressure.

*** Per ISO-4126, the generic term 'Safety Valve' is used regardless of application or design ***

*** 'Pressure-relief Valve' is the equivalent ASME/API term ***

 

[yene]

Right, but ASME VIII applies to construction of vessels. I'm concerned with sizing a vent header / rupture disk system based on the required relieving flow in a fire scenario.

 

[The Obturator]

ASME VIII Also includes for the provision of overpressure protection - 15 PSIG and above. As of this month July, 2021 specific Pressure Relief Device requirements are now in new ASME XIII Section as referenced in new 2021 ASME VIII. In any case, API-521 is a user standard which you can use.

*** Per ISO-4126, the generic term 'Safety Valve' is used regardless of application or design ***

*** 'Pressure-relief Valve' is the equivalent ASME/API term ***

 

[don1980]

This is quite the discovery as my understanding was always that any flammable/combustible liquid must have its required relieving flow calculated using the NFPA 30 formulas. Is this really not the case? Operating in Canada by the way

Yes, it's really not the case. The required relief load doesn't have to be calculated using NFPA 30.