RVs downstream of high pressure cylinders

[kyu]

Hi all,

I am looking at a relief valve that has been installed downstream of a high pressure cylinder. The cylinder contains a gas at a pressure of roughly 2400 psig. The cylinder is hooked up to a regulator that lets the pressure down to approximately 100 psig.

We have a relief valve downstream of the regulator to protect the downstream equipment in the instance that the regulator fails in the wide-open position.

I want to work out how much flow the relief valve will be required to handle. In order to do this, I need to know what the pressure upstream of the regulator will be.

Originally, I thought only of using the service pressure (2400 psig), but I have recently been wondering if perhaps it might make sense to use a fire-scenario pressure instead (i.e. there is a fire, and the pressure in the cylinder rises well above the service pressure, perhaps to a pressure corresponding to the cylinder's own relief setting -- it is this abnormally high pressure that would cause the regulator to fail in this instance).

Does anyone have any experience with a similar situation?

Thanks to everyone who reads this.

 

[EGT01]

In some cases, if I have no other info to support what a system's maximum upstream pressure might be, I will use 90% of the upstream relief valve set pressure as a likely maximum operating pressure. This is based on the upstream relief valve being a conventional type relief valve which, as a general rule, the operating pressure should be less than 90% of setpoint to ensure the relief valve does not simmer.

You indicate the cylinder relief valve set pressure is well above the service pressure. Then it is possible that the cylinder can be filled to a pressure above its service pressure but limited by the cylinder's relief valve. So I would say there is good justification to use something other than the service pressure.

Whether fire is a realistic scenario or whether fire will affect your regulator is up to you to evaluate. However, to say the regulator fails because the upstream pressure is abnormally high seems to imply you have not selected the proper regulator in accordance with the maximum possible upstream pressure.

 

[kyu]

Thanks for replying. I'll respond to each of your three paragraphs in turn.

1. Your 90% guideline makes a lot of sense to me. In my office, some of my colleagues will go all the way up to the relief valve set pressure, since the system upstream of the regulator is separate from the system downstream of the regulator, and you never know what could be going on over there. This always struck me as a) perhaps overly conservative and b) if you're going to do that, then why not toss in the overpressure as well? The 90% makes more sense I think.

2. These are the kinds of cylinders that we receive from someone else; we don't fill them ourselves. That said, my understanding is that when the cylinders are filled, the supply vessel to the cylinder has a relief valve that is set at 110% of the service pressure of the cylinder. So the cylinder can only be filled to (in this case) 2400 + 240 = 2640 psig. I think the only way the pressure could be higher than this is if there is a fire.

3. Yes, that implication is there isn't it? From my (rather limited) experience, I have found that these kinds of regulators are fairly standard, and that a 3000 psig max. inlet pressure is typical. Perhaps the relief devices on the cylinder ensure that 3000 psig can never be exceeded? I must admit I don't know very much about cylinder relief.

Anyway, thanks for your help.

 

[Montemayor]

kyu:

You don’t give specifics about your installation that would enable us to furnish a quicker and accurate answer to your query. For example:

1) What gas are you handling in the cylinders?
2) Is this a manifold installation? In other words, are several (or many) similar cylinders feeding a common manifold?
3) Are the cylinders in question common, industrial gas cylinders as found in the USA? It seems, from your description that they are. To an old industrial gas engineer like myself it sounds like you have a cylinder model 300cf, with DOT specifications 3AA 2400, a service pressure of 2,400 psig, and 9.27” diameter x 55” tall.
4) If we knew the gas, we would know the standard CGA (Compressed Gas Association) valve that is being used on the cylinders. This valve has a rupture disc built into it and it is pre-set for a given pressure.

The cylinders are equipped with the rupture disc in order to eliminate the possibility of rupture under pressure in the Fire Case. Therefore, I fail to see what you are worrying about. Each cylinder, if equipped with a CGA valve, has its own inherent safety device built in. It’s great that you have a relief valve downstream of the cylinder bank or manifold, but I fail to see your concern regarding the cylinders, unless I'm missing something.

Perhaps you have not read the applicable CGA literature on the gas, the cylinders, and the respective cylinder valves. If not, I highly recommend that you read this literature thoroughly first before trying to apply relief logic to the installation. You will find that the CGA and the DOT (as well as your industrial gas supplier) have already taken into consideration a lot of the hazards involved in handling and applying the cylinders of compressed gas.

I have created an Excel workbook with considerable information on what you are doing regarding the application and use of compressed gases. I downloaded the basic information from Liquid Air’s website some years back and transcribed it into a compact Excel workbook that is very handy to use. Tell me your email and I’ll send you a copy for your use.


Art Montemayor
Spring, TX

 

[kyu]

Montemayor, thanks for responding. I'll reply to each of your queries in turn:

1. We have a number of cylinders actually, but the one I first looked at was nitrogen.

2. No, it's not a manifold. Each cylinder is hooked up to its own distribution system.

3. Yes, those are the ones. I'm sorry I didn't specify that before, but I'm not too strong on my industrial gas cylinder terminology.

4. This part I didn't know. Are you saying that a given gas is always filled to the same service pressure? I *was* under the impression however, that the rupture disk is always set at 5/3 of the service pressure, whatever the gas may be. Having said that, it wasn't clear to me whether that rupture disk is standard equipment or not.

In response to your first paragraph, it's not the cylinders I'm worried about per se, because I know they have their own safety systems. I was just thinking about the extremely high burst pressure of the rupture disks, and wondering if I should be using those pressures as the inlet pressure to my downstream regulator when sizing my RV.

Just this week my office sent for and received a copy of CGA S-1.1. I admit to only giving it a cursory glance so far. I will go over it more thoroughly sometime soon. What is DOT?

Your spreadsheet does sound very useful indeed; it is most generous of you to offer me a copy. I would appreciate it very much. My e-mail is richard.a.haack@can.dupont.com

Thanks again.

 

[Montemayor]

Richard:

DOT = Department of Transportation, the official ruling body regarding the transportation of hazardous materials across state lines in the USA. I know you have a similar agency in Canada, but I forget its name or acronym.

The rupture disk on some of the cylinders are actually fuse plugs, set to melt at a certain temperature in case of fire.

All nitrogen, Oxygen, Argon, and air cylinders are filled to the same pressure limits as denoted on your cylinders - 2,400 psig. As you state, the pressure actually goes higher than this due to the heat of compression. But when the cylinder settles down to ambient temperature, it comes down to 2,400 psig or less. You must consider that maximum pressure as being the ultimate pressure available to your downstream regulator's inlet nozzle. That's what is going to initially happen the moment you connect a fresh, recently filled cylinder to your downstream regulator. As the contents are consumed, the cylinder pressure is reduced accordingly until the regulator can no longer sense any useful dP.

I hope this explanation helps. Your Workbook is in the mail.

Art Montemayor
Spring, TX