hi Can someone tell me whether the 2

[Antonia74]

hi
Can someone tell me whether the pressure level is critical for adiabatic compression of oxygen in the pipe to happen? An oxygen hose exploded at the client site, in which the oxygen supply pressure from the header directly to the hose at the moment of explosion was about 6barg. Pressure range that we use is about 5barg - 20barg. In other words, is adiabatic compression occurs at also low pressures, or only at pressures higher than 150bag.
Thank you

 

[1503-44]

How?


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

 

[Antonia74]

well, we have a skid supplying O2, N2 and LPG to the blast furnace. The oxygen was connected with a hose (1-1/4 in) to an oxygen line (2 in). We used client's hose , probably not purged and perhaps also not certified for oxygen (we can;t trace back the cause of explosion). Anyhow, upon start up, the explosion happened and the hose was shredded to pieces and 3 people were injured. The rest of the SS piping in the skid was burned on the inside but not ruptured. I checked the design and the only thing that I can deduce is that the smaller size pipe can create high velocities (I am not sure why the design is as it is because we want to have 1200 Nm3/h of O2 at say 6 barg, so 1-1/4 in is too small for sure). So do I have adiabatic compression or exceeding of the impingement velocity with the presence of impurities in the hose (normally we use our own hose but this time it was too short so we took one from the client - probably mistake number 1). I have read in another thread that adiabatic compression is more likely to occur at pressures > 200 bar.

 

[Prometheus21]

Hello,

"Can someone tell me whether the pressure range is critical for adiabatic compression of oxygen in the pipe to happen?" - Yes.

"I checked the design and the only thing that I can deduce is that the smaller size pipe can create high velocities" - correct, as an example: with 6 mm tubing you get sonic/ near sonic velocities during filling of cylinders. Smaller hose/ pipe diameters gives you a massive velocity increase depending on the pressure differential.

Now for most of other common uses, standards (like CGA G-4.4) dictate velocities based on the pressure range, and whether the piping is an impingement site or not. At 6 barg the CGA G-4.4 states a velocity of 30 m/s at a working pressure of 0.6 Mpa, or 6 bar, at impingement sites. For non-impingement sites the speed is 60 m/s at that pressure rating.

Now when you say "client's hose", is it a metal hose or a flexible hose with reinforced fluoroplastic tubing? PTFE is quite common, as well as PFA. Metal hoses are more prone to ignition. Properly cleaned 316L tubing is safe to use with oxygen, even at sonic velocities at 200 bar.

"So do I have adiabatic compression or exceeding of the impingement velocity with the presence of impurities in the hose (normally we use our own hose but this time it was too short so we took one from the client - probably mistake number 1)" ayayay. Yes that was a mistake. Never use a non-certified hose for oxygen systems.

Now when it comes to oxygen you have many potential ignition mechanisms:

Particle impact - most likely. A minor contaminant is more than enough.
Heat of compression - not likely, but a small contributing factor.
Flow friction - not likely.
Mechanical impact - not likely unless a nonmetal got ignited.
Fresh metal exposure - not likely unless there is aluminium or titanium present in the system.
Static discharge - not enough info to base an opinion, but not that common.
Chemical reaction - not enough info to base an opinion, but not that common.
Thermal runway - not enough info to base an opinion.
Resonance - rare but possible.
External heat - Possible.

There are more but these are the ones often listed. In all of my years working the most common causes is: particle impact and heat of compression acting in unison. A small contaminant is enough to start an ignition, and there you go; a kindling chain. Now, I must state that particle impact is a lot more likely in high velocity systems, and where the impact point is 45 degrees to perpendicular to the path of the particle.

I would look at the materials in the system, possible contaminants and their flammability ratings. Consider the total heat that can accumulate in the system (see ignition mechanisms) and relate that to autoignition temperatures and heat of combustion of the materials (metals and nonmetals) in the system. Note that many ignition mechanisms can exist in the same system, and a kindling chain assessment is therefore needed (the ability of ignition to propagate with a component or system).

Now there could be other reasons as well, depending on factors like oxygen concentration, the temperature of the oxygen, the possibility of LPG backflow into the oxygen distribution system (A client had that once, ugly business.) Not sure if this helps you, but there is, like you say, a certain difficulty in figuring out the reasons for an explosion after it happens.

 

[Antonia74]

Thank you very much! Indeed your text is very helpful. I think you are right about adiabatic compression and impingement velocity unison. These two phenomena are very similar and can happen simultaneously. The hose was made from rubber, but corrugated, so traces of oil have been carried over into the skid where the ignition happed and spread bidirectionally. Then the hose burned down. We learned the lesson of never using non-approved hoses.

 

[Prometheus21]

Depending on the rubber it would most likely have ignited anyway, even if the oil was not there. The corrugation would have created flow friction, several impingement zones, turbulence scenarios that would easily ignite a non-metal that is not compatible with oxygen service. Oil residues on top of all that is a disaster. Hopefully no one was permanently injured by the accident.

 

[Antonia74]

Yes, true, the corrugations can cause flow friction. Normally we use our hose that is certified for O2 and smooth on the inside. Three people were injured, luckily they are all recovered.