Use of API 2510 or NFPA 58 for design of LPG storage tanks

[Cubye1]

Hi, I would ask any help on interpretation of the right applicable standard for design demands. There are a group of spherical tanks each in the size of 1500 Ton LPG. This is a farm of containers that located inside the refinery premises, and used to store the final product before delivery to road tracks.
The argue is whether the tank design and safety means/accessories are to be as set in NFPA 58 or as API 2510.
NFPA 58 excludes refineries, and API5210 excludes tanks installations that are covered by NFPA 58/59
NFPA 58 demands an internal safety valve (as exessfllow valves- EFV) on outlets, and API 2510 doesn't mention it in any way.
Can you please help on interpretation what is mandatory here according to USA regulation and is it possible that API excludes Excess flow valves from use?
many thanks
Esra

 

[Duwe6]

Unless you are self-insured, ask your insurance company's Risk Management person. The answer will to install Excess Flow valves.

The risk of a hose break or other loadout problem is fairly high when running "bare". The consequences will be extremely bad -- Boiling Liquid Expanding Vapor Explosion BLEVE -- next to multiple vessels containing a huge ammount of LPG. Risk times Consequences equals unacceptable exposure in both dollars and dead operator(s).

 

[Cubye1]

thnks Duwe6 for your response. can you comment on 2 issues -
1- which standard is mandatory in USA for the design of LPG storage farms? API2510 or NFPA 58?

2- in reference to the liquid oulets of the tank (not the track loading station)- is it accepted not to use internal safety valves or internal Exessflow Valves according to API 2510 between the tank outlet and the steel pipes, at a refinery for example?
thanks for any comments
Esra

 

[Duwe6]

1 - Neither, really. Your company needs to "declare" which spec you will be using. My recommendation is API-2510 *plus* API Publication 2510A. The Pub is "Fire protection considerations for the design and operation of LPG facilities"

2 - Not cognizant enough on 2510 to give you an "API" answer. The risk of a BLEVE is enough for a prudent engineer to put them into the system. The alternative is to use an all-welded pipe system with plenty of strain relief - expansion/contraction loops, and 100% volumetric examination of all buttwelds. The 100% exam can be RT - x-ray, UT - ultrasonic shearwave testing, or PAUT - Phased Array UT. Also, all the high-stress welds -- first 2 90's out from each anchor, all 90's in expansion loops -- should be MT'd - magnetic particle exam for OD cracking. All piping, including flanges and flange bolting has to be rated for cryogenic temperatures.

For me, that is too much work and too much $$$. Just put in the Excess Flow valves.

 

[Cubye1]

Thanks Duwe6. I appreciate your response. The need for cryogenic design on piping is quite interesting and logical.
The reason of not using EFV was raised by engineers claiming that it blocks their need for higher capacities (I found that it exceeds 3 meter/s) and the need for frequently repairs.
The NFPA58 demands that the safety valves will be internal, and in that case they need to empty to "Gas Free" the vessel for each repair. I found it somehow a "chain" of faulty design that leads to it. I still cannot understand why (if it's so) the API 2510 doesn’t mandate or even doesn’t mention the use of EFVs.
Esra