LPG Storage Area - Flanged Connections

[NickVil]

Hi,
We have some flange connections within the LPG Storage bunded areas. This is to, in case of leaks the LPG will stay inside the bund. In the event of fire in such a bund, these flanges would be vulnerable to opening, and provide additional flammable material to the fire.

However some people have observed as current best practice as having no flanges on the bottom of LPG spheres, with minimal in-bund piping, and where present being fully welded with the remote operated isolation valve welded on the sphere side and located outside of the bund.

Is it recommendable to move such connections and piping outside of the bund? I think doing that, in case a leak the LPG will go to some other place instead of the bunded area.

Is there some standard that recommends the position of such valves?

Regards

 

[rneill]

What you want to review are the following:

API Std 2510, Design and Construction of LPG Installations, and
API Pub 2510A, Fire Protection Considerations for the Design and Operation of Liquefied Petroleum Gas (LPG) Storage Facilities

 

[dcasto]

can you explain how a gas will stay inside a diked (bunded) area?

 

[NickVil]

The C3, C4s and some C5 (up to 2mol%) have a molecular weight heavier than air. So, if there is not a ventilated area, this components will stay over the floor.

I know that a manual shutoff valve in the pipe connection should be provided between the tank and each liquid or vapor header. These valves should be kept closed, except for the connection in use to minimize the amount of gas that could escape through a leak.

I want to know if the valves described in lines above must be after the dike to allow the access for the operator or inside the dike to prevent that LPG from any leakage through the valve will go out of the dike and can originate fire risks.

 

[786392]

Dear dcasto Hello,
I was an eye witness in cold winter still air conditions travelling of LPG (C3,C4&fraction of C5)in white cloud form un-disturbed to as long as 150ft plus inside a bunded area.

Thanks God it could not find any Ignition source(s) and I am still alive to share this scenario.

Best Regards
Qalander(Chem)

 

[dcasto]

786392,

Yes, in risk management we know of these situations, but if we designed everything for things that could happen one every decade with 1,000,000 sites, we'd never be able to afford to build anything.

 

[tickle]

Dcasto

I guess that as engineers we have to apply ALARP and/or BATNEEC, or at least recommend. Considering the experiences of 786392 and other less fortunate examples the OP has identified an important scenario. Whatever risk mitigation measures are put in place, there will still be a residual risk and it is up to the owner of the facilities or insurance company to accept this risk. Which takes us back to my first sentence.

My comments to the OP; Will there be a over pressure protection (PSV)? Surely this will relieve lpg inventory. Is there a deluge system? The leaking flange will only add inventory if there is a source of LPG (eg truck load in)- this should be provided with its own isolation that is remote from the storage bund.

 

[786392]

Dear NickVil Hello/Good Morning,

Hopefully you have already studied/gone through NFPA 58 in this connection if not there might be some good guidance on this issue.

Moreover there is one guide for LPG bulk storage on "googling..." you should easily find this;that may have some useful info on the same issue.

Additionally if needed solar radiation hinderance shade should reduce underneath temperatures& a properly designed deluge system along with remote operated valves and sprayed water conservation in-place should prove helpful as pointed out by tickle.

Hope this helps in showing way forward.

Best Regards
Qalander(Chem)

 

[dcasto]

tickle,

if we accept without documentation, or without a resonable investigation of root cause analysis of 785392's scenerio, how do we ever move forward. every uneducated engineer or politically motovated person could stop any progress by claiming something happened. ol 7's witnessed event and claim that a white cloud (oh, HC's are colorless) could be traced to 10,000 other issues other than a dyke.

Bottomline, documented root cause analysis must be the basis, not some "ALARP and/or BATNEEC" precieved adjustment to good engineering design practise.

 

[rneill]

The root of the problem is obtaining consensus as to what is "reasonable" and "practical". In my experience, there can be a wide divergence of opinion as to what constitutes reasonable and practical. For that reason, I prefer to work to accepted industry practices such as the RP's published by API; these are developed by committees whose members represent some of the most experienced and knowledgeable people available in that particular subject area and they bring to the table sound factual based knowledge when the practice is written.

That said, my concern is more with applying standards that are less stringent than established industry practice, not with those that are more stringent. If a particular end user wishes to apply more stringent requirements, it is their money to do with what they choose...