Relief valve for a plastic carboy with MAWP less than 15 psig

[jari001]

Hi everyone,

I am designing a small system that will use a 10 gallon plastic carboy as a receiving vessel for IPA that we are using in the process. The source vessel for the IPA is a stainless steel canister (5 gal, filled with 2 gal of IPA) that is rated above 100 psig, though my house N2 pressure is 90 psig and I can regulate down further at the point of use; I will use nitrogen to pressurize the headspace of the source canister to deliver IPA into the system. I am assuming the carboy has a MAWP of 5 psig until I get the specs. Can I use the guidelines in API 520 to determine the relief valve size in this situation? When I read the Scope for API 520 (7 ed.), it says that the minimum MAWP of the equipment in need of relief protection is 15 psig. Is there other guidance for relief protection w.r.t low MAWP equipment?

Edit: The relief scenarios I want to protect from are the pressure buildup from the IPA collecting into the carboy (IPA vapour pressure plus N2 that was used to push to the IPA through the system) and accidentally blowing N2 straight through the system into the carboy.

Edit: IPA means isopropyl alcohol (not pleasant for barside consumption)

 

[met11]

Can you just vent the plastic carboy during filling? It shouldn't be pressurized at all, in which case no relief valve required. If you can't vent it to atmosphere you could just put a really low cracking pressure check valve on the thing and run the outlet line somewhere safe? Or alternatively just the equivalent of a water seal (with IPA) to vent off if the head pressure gets too high.

 

[Latexman]

This isn't Code, agree with met11. Check-All is your friend.

Good luck,
Latexman

To a ChE, the glass is always full - 1/2 air and 1/2 water.

 

[jari001]

Asking this question made me realize I put the cart before the horse because now I realize that my only reliable way to vent the carboy is to provide a vent header connection within the system. I am hesitant to do this because I have included an IPA LEL monitor with process interlocks in the vent header as part of my design and I wouldn't want the carboy "outbreathing" to activate the interlocks. My process would fit inside an enclosure like a gas cabinet, so if I had a dP based interlock (that has its own questions I'd have to run down) on the exhaust rate I could allow the carboy to relieve into the cabinet instead of the vent header directly - but that's an additional level of complexity that I'd like to avoid.

I will approach my reviewers with the check valve idea - maybe they have used this method of overpressure protection before. I don't know what met11 means regarding the water/IPA seal, do you mean a device that the venting vapour would go through that would strip out the IPA vapour?

 

[LittleInch]

You need to properly sketch this out and think about gas pressures and flows.

You have one cylinder filled with IPA (?? IPA in my world is Indian Pale Ale, a fine beer, but I suspect not what you have) which you are subjected it to an unknown pressure of Nitrogen to force it into your plastic vessel, akin to a jerrycan?

I doubt this plastic vessel will be officially rated for any pressure at all, 1-2psi at most.

So where does the gas go that is being expelled by the IPA?

why do you have a change in pressure rating?

I think you need to look at tank venting systems and / or a bursting disc to protect the plastic vessel which if subjected to any level of N2 flow is liable to simply explode.

This is all far too small beer (pun intended) for the codes, but I would look at API 2000 or search in the tank forum for pressure vents. A large weighted metal flap device might be easiest to calibrate and vent any excess pressure.


Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[jari001]

A few updates:

The carboy I have on hand to use is designated as UN 1H1/Y1.8/100, so it was hydrotested to withstand 100 kPa (~14 psig, rounded down) if I understand the UN markings system. This is much better than my assumption of 5psig as the "MAWP"

My manager would like for me to go through using API 520 as we have done so for similar situations where the "MAWP" was too low for API 520, his rationale was that the PSV is a more tightly manufactured device compared to a check valve and in flammable liquid/vapor service that is the best practice.

I am revising my system sketch before I submit for P&ID creation, I will have to focus on how to integrate the venting of this carboy into the system or decide to provide some other means of collecting the spent IPA.