1000 psi O2 piping material 3

[PagoMitch]

Been awhile since I have been on the site...
I have a client with a research related need for piped O2 at 1000 psi. Source is (16) 2000 psi O2 H series tanks. My usual domain is Hospitals, and this is way outside of that.
I've researched some ASME docs - but cannot seem to find the applicable for high pressure O2.
I'm leaning towards seamless 304 or 316L, but I need confirmation.
Our local Airgas folks are useless.
Any recommendations for info?
TIA

 

[EdStainless]

There are a number of NASA publications on this subject.
Yes, all SS is the most common approach to this.
Flow and velocity control are important.
Selection of valves and regulators is critical.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[shvet]

As per my experience SS X6CRNITI18-10 EN 10217-7 is applicable up to PN100 in range from DN6 to DN400 in ambient-temperature oxygen environment.

Beware of design specification as details are crucial - the same pipe may be safe or not depending on specification details. See ASTM G88, ASTM G94, EIGA IGC 13 and references in those. Also table F.3.4.2 in NFPA 53 may seem useful.

 

[EngrPaper]

Monel piping would be the most common, especially for higher pressure systems which often see higher velocities on press up or blowdown. If you go with stainless, you need to be [more] careful with velocities in the system causing ignition events.

 

[EdStainless]

Monel is often used for breathable O2 service, and SS for other uses.

I don't know, does the CGA have a high pressure O2 doc?

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[1503-44]

O2 system design, see attached.

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[EngrPaper]

Monel is often used for breathable O2 service, and SS for other uses.

That too. In some previous aerospace work, we used stainless for all liquid oxygen systems where we would have slow chill in to prevent high GOX velocities and then used monel for all gaseous systems so that we could have faster/lower risk tank fill during loading. These were oxidizer loading systems, not the cabin O2. You mention NASA, hence I wanted to point out that those kind of systems favor monel.

 

[georgeverghese]

As far as I can remember, all of our O2 equipment within the cold box and O2 compression external to the cold box, going up to 500psig, was in 316L. None in 304L. Velocity limits apply for GO2 with SS316L nevertheless.

 

[PagoMitch]

Thanks all for the info - especially 1503-44!
Much appreciated.

Interesting stuff indeed.
Looks like we will be using 0.12" wall 1" copper tubing. Even 316L seems to have issues with high (we are now at 1300 psi) pressure SS. In my HVAC-land endeavors over 4 decades, I never knew that SS could ignite!

 

[1503-44]

We wouldn't want any nasty surprises.

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[LittleInch]

A key issue is the number of things that burn in oxygen service that are normally inert.

Especially any traces of grease, cotton or other seals.

1300 psi for copper must be pretty small diameters....

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

 

[shvet]

Copper-based alloys used in components for oxygen piping systems generally contain at least 55 weight % copper. Included within this group are the coppers, brasses (copper alloyed primarily with zinc), bronzes (copper alloyed with aluminium, silicon, manganese, tin, lead, etc.) and copper nickels (copper alloyed with nickel). These have had an outstanding application history in oxygen service. Caution should be exercised in the use of aluminium bronzes. Aluminium bronzes containing more than 2.5% and up to 11% aluminium (by weight) have been extensively used for cast components (e.g., valve bodies, pipe fittings, etc.) in oxygen pipeline duty for many years without a significant history of failure. However, use of aluminium bronze is not recommended as flammability tests show that it will support burning if ignited, even at low pressure.
Aluminium content in copper alloys should be limited to 2.5% (by weight).

5.2.3 Incorrect design of oxygen system
The proper design of oxygen installations is critical. Inadequate designs can and have lead to serious accidents.

Examples of inadequate design
a) Use of rapidly opening (ball) valves. This can lead to ignition caused by the heat generated by high velocity gas or adiabatic compression, (see below).
b) Allowing too high a gas velocity which can cause ignition of incompatible materials in the system due to particle impact, etc.
c) Opening the main shut off valve in an oxygen supply pipeline without equalising the pressure before.
d) High pressure gas in the presence of sharp edged orifices, rapid expansions or reductions.
e) Poorly located vents causing accumulation of oxygen in the vicinity.

5.2.4 Incorrect operation and maintenance of oxygen equipment
Incorrect operation and maintenance of oxygen equipment is one of the most frequent causes of fires in oxygen systems.

Examples of Incorrect Operation
a) Failing to reset pressure regulators to the closed position when the oxygen cylinder valve has been closed. This results in extremely high oxygen gas velocities when pressurising the regulator next time it is used.
b) Rapid opening of valves can result in momentarily high oxygen velocities, sufficient to project any debris present in the system through the system, at sonic velocity causing frictional heat, sparks, etc..
c) Opening a valve rapidly against a closed valve (or pressure regulator) downstream in a system – high heat can be generated through adiabatic compression of the oxygen causing a fire.
d) Start-up of an oxygen compressor erroneously with oxygen. (This is incorrect operation only relevant to special cases – see references 7 and 8)

Examples of Incorrect maintenance
a) Working on pressurised systems.
b) Venting oxygen into restricted, enclosed or confined spaces.
c) Allowing systems to become contaminated. Contamination by particulate matter, dust, sand, oils, greases or general atmospheric debris creates a potential fire hazard. Portable equipment is particularly susceptible to contamination and precautions shall be taken to prevent ingress of dirt, oil, etc.
d) Failure to completely remove cleaning solvents from components which are to be used in oxygen service. The solvent residues are not compatible with an oxygen enriched atmosphere.

5.2.5 Use of incorrect materials
Design of oxygen equipment is very complex and the “why and how” is not always obvious. In essence nearly all materials are combustible in oxygen. Safe equipment for oxygen service is achieved by careful selection of suitable materials or combination of materials and their use in a particular manner. Any modifications to a design must be properly authorised to prevent incompatible materials being used.

Substituting materials which look similar is extremely dangerous and many accidents are reported where the cause was incompatible replacement parts. Examples of this practice could be:
a) Replacing o-rings and gaskets with similar looking items. There are hundreds of different types of elastomers and most are not compatible with oxygen.
b) Replacing a metal alloy with a similar type of alloy. The composition of particular alloys has a significant effect on its mechanical properties and oxygen compatibility. “Bronze”, which covers a wide range of alloys, has several varieties that are compatible with oxygen and even more which are not; e.g. tin bronze is used in liquid oxygen pumps while aluminium bronze is considered hazardous.
c) Replacing PTFE tape with a similar white tape. Not all white tape is PTFE and not all brands of PTFE tape are safe for use in oxygen., see EIGA Doc 138, PTFE Tape as a sealant for Cylinder/Valve Connections
d) Replacing parts/components with non-approved equipment is not allowed. The geometry of certain components is sometimes critical and approved manufacturer’s parts shall always be used when maintaining oxygen equipment.
e) Replacing or installation of combustible material in filters e.g. plastics, paper, adhesives. Filters in oxygen systems are very sensitive to ignition due to presence of particles and complicated flow conditions. Therefore filters should be made of materials that demand very high ignition energy e.g. Monel.
f) Lubricants are generally not allowed in oxygen service except for special applications.
Specialist expert advice shall always be obtained before applying such lubricants.

 

[1503-44]

Oil and grease free valves and no oil filled pressure gages.

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[FacEngrPE]

Some normal materials become shock sensitive when saturated with oxygen.

https://sciencing.com/spillage-asphalt-pavement-potentially-hazardous-6509943.html

This is an important consideration in the design of liquid oxygen bulk delivery unloading stations.

 

[PagoMitch]

Thanks again all for such informative responses.

The EIGA Doc 13 "Oxygen Pipeline Systems" is extremely useful; yet it is a Euro doc, and I cannot seem to find any NFPA docs with similar info. NFPA 53 "Recommended Practice on Materials, Equipment, and Systems Used in Oxygen-Enriched Atmospheres" would - by the title - appear to encompass similar criteria. Alas, IMHO it is almost useless, and seems to be written by gents who did not want to commit to anything; lots of vagaries where you would expect concrete "though shalt" data. As well, the current NFPA system of "online viewing" is maddening. Is anyone aware of a NFPA - or other US based - doc that covers this same info?

shvet - Great info. Thanks for citing that doc. Again, it would be great to find a US-centric doc that provided similar info.

Back to the project - we ended up at 15 CFM at 1300 psi, for which the 1"/0.12" wall meets the pressure requirements, at a nominal velocity of 80 fps. Our source has been reduced to a 2-bank system of 2 tanks each; it seems the volume required has been reduced significantly.

Thanks all.

 

[1503-44]

No nasty surprises.

80 fps seems a bit high for tubing.
There could be high vibration and associated work hardening on a relatively very thin wall.

https://www.sciencedirect.com/science/article/abs/pii/B9780080439594500225

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[PagoMitch]

1503-44,

Now that is interesting. Although, with the significant disparity in O2 vs crude oil density, I would not expect O2 to exhibit the same amount of vibratory force on fittings. I think. F=mA and all that.

Re: the velocity. I was also a bit concerned with that; although I am (barely) within the 30m/sec limitation I have seen listed.
However, we are kind of boxed in to this size.
Any smaller and velocity is out of bounds.

The 1" tubing and 1.125" and 1.25" data is based upon Swagelok engineering data.
1.125" tubing with 0.12" wall has a pressure limitation less than our operating.
1.25" tubing with 0.12" wall has the same problem; as pipe size goes up, with the same wall thickness, the safe operating pressure reduces.
We would need to go to 1.5" tubing with 0.25" wall to get the pressures to be satisfactory, using ASME B31.1 calcs using similar data as Swagelok.

So, 1.5" tubing works; but Swagelok does not offer any fitting in this size that are not stainless. A call to about half a dozen aerospace tubing/fitting vendors did not come up with ANY that could supply copper based mechanical fittings for 1.5" copper. Everybody seems to use only stainless. Rather a puzzle to me...as per everything I have read, SS tubing/fittings at these pressures is an accident waiting to happen.

At this point, I still need to confirm another couple sources re: max velocity. If we absolutely must reduce below our nominal 80 fps, about all we can do would be to run (2) 1" lines in parallel. Crude, but functional.

Sigh.

 

[1503-44]

OlK, maybe false alarm.

Medical app

https://www.copper.org/applications/plumbing/cth/design-installation/cth_3design_medgas.html

Also see attached (work hardening)

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[PagoMitch]

1503-44 - Thanks for the info.
I have the copper.org info, but eliminated silver solder/brazing due to the low (250-350 psi, depending upon whose data you choose to believe) operating pressure limitations.

The work hardening info is also interesting; but I sincerely hope that the installer is not bending our 1"/0.12" wall or 1.5"/0.25" wall tubing when he is installing.

I am still somewhat bothered by the design decisions we have been forced to make; if we end up running the (2) 1" lines in parallel because 1.5" fitting are "not available", it just seems so...amateurish. Like someone did not think this all the way through...

Thanks again all. This site represents the best of what the Inet has to offer.

Regards all.

 

[1503-44]

2 tubes.
Right. It could be telling us something, just not sure what.

Maybe it's just telling me that I'm not used to transporting a gas through 1" tubing at rates equal to cross country transmission pipelines and the usual flow rates I see for tubing are standard cfm at a few inches of water. But remembering way back to my Big 3 DAYS, Our O2 piping was steel and we were filling banks of O2 gas cylinders through some kind of tubing at the manifold (prob Ss, definitely not copper) at relatively high rates, 300 psi/min, 10min to 3000 psig. It would heat up the cylinders quite a bit and enough to require water cooling. I'll have to figure out what velocity was. Oh, not too difficult, 75 ft2/10m = averages to 7.5 Acfm, but I don't remember the tubing size. It definitely was not 1", so it would seem to be in the be around 30 to 50 Afps.