Compressing Hydrogen and Oxygen

[giovyledzep]

Hello,

Newbie here. Just one question: what would it take to safely blend a mixture of 96% Hydrogen and 4% Oxygen to 3000 psi to avoid any hazard of explosion? The mixture itself is not considered explosive so is it possible to just run the two into a compressor without any more attention than required when handling high oxygen blends?

Thank you

 

[rotw]

Put the compressor inside a deeply burried bunker that has very very thick walls...

Life is not about waiting for the storm to pass. It's about learning dance in the rain.

 

[georgeverghese]

UFL for H2 in pure O2 is 95% v/v, so you are very close to this point, too close I would suggest.

 

[giovyledzep]

I know of people who do this successfully without incidents. I'd just like to know the additional safety steps required and where the weak points of a system would be.

 

[georgeverghese]

The autoignition temp for H2 is stated to be very high at 530degC, and I am no expert on whether such temperatures can be reached by accidental knocking or rubbing metal surfaces in a compressor. At the very least, suggest working on 97.5 % v/v H2 provided some one can give some assurance that (a) the compressor can be safely electrically grounded and (b) in case of loss of lube oil or some damage to compressor internals, that any rubbing surface temp cannot momentarily reach 430degC or so.

 

[rotw]

A journal bearing experiencing shaft rotation in reverse (unlikely but can occur) can warm up at temperatures orders of magnitude higher than what is considered to be the normal temperature (about 110 deg C). The journal bearing is not in contact with process gas as long as the end-shaft seals are working properly. Tight clearance rotating elements (labyrinth, impellers, etc.) can rub also inside the casing. What if a bolt breaks while the machine is in operation?
Just saying...
- What is the point, to prevent explosion or flammability?
- In regard to the later, is it true that a mixture of H2 and O2 is always flammable irrespective of the mixture ratio?


Life is not about waiting for the storm to pass. It's about learning dance in the rain.

 

[georgeverghese]

@rotw,

"Tight clearance rotating elements (labyrinth, impellers, etc.) can rub also inside the casing." - this would be the main concern.

We typically try to keep well away from flammability limits. For LFL, we maintain operating operating conditions at not more than 50% LFL. I have extrapolated the approach to UFL also in this case.

With both these safeguards in place (2 barriers before loss of containment), we would have a safe configuration in principle. A design - operating concept with just one barrier in place is beyond my work experience / knowledge.

 

[rotw]

All this is also beyond my knowledge.I am just thinking out loud.

But considering the combustibles are already inside the casing and ready to ignite, why then would loss of containment be the issue here?

Life is not about waiting for the storm to pass. It's about learning dance in the rain.

 

[georgeverghese]

@rotw,
Agreed, in this case we already have a fire or explosion risk to begin with. But here, consequential loss of containment ( if the compressor cannot withstand the resulting internal overpressure from internal detonation, which is likely) can lead to escalation of the fire to include other neighbouring equipment.

A safer approach would be to compress H2 and O2 individually, cool and then mix to get to the desired blend. With this approach, you could produce blends closer to the UFL.

 

[giovyledzep]

@georgeverghese

Thanks for the reply. My concern is actually more about how to blend the two gases after individual compression as you said, to reach a final pressure of 3000-3500 psi. How would you go about blending them?

 

[georgeverghese]

Okay, that sounds a lot safer. There may be more than 1 way to blend these streams at 3500psig. Offhand, I would setup a flow ratio control loop to do this, and use an inline static mixer downstream of the mixing tee. If you are not familiar with setting up a flow ratio control loop, talk to a plant process controls engineer also.

 

[rotw]

By the way if you go for separate streams compression, make sure you find a compressor manufacturer that is capable to offer a metallurgy that can handle Oxygen at 100% purity or so...

Life is not about waiting for the storm to pass. It's about learning dance in the rain.

 

[rotw]

More about oxygen compression here

https://www.eiga.eu/index.php?eID=dumpFile&t=f&f=2252&token=b2ed707cbf38e0057950d253914b1e5d0a032d62

Your compressor can burn, if it does it can burn violently...

Life is not about waiting for the storm to pass. It's about learning dance in the rain.

 

[georgeverghese]

I have direct operating experience with running these Sulzer oil free compressors in 99.8%v/v O2 (remainder is probably mostly argon) service at 500psig, and the attachment, on page 228, fig 12 says they can make models operating at up to 3500psig operating pressure or so. Very reliable machines, never gave us any trouble, unlike some of the other compressors in the plant.

https://pdfs.semanticscholar.org/baee/d16f490677ddc00fa2a85aa5383881d5f4fb.pdf