Could you enlighten me as to what COTS stands for?

Seems you don't have much driving force for separation here, with the stream being at such a low pressure.  Whatever you do is going to consume a fair bit of energy relative to the quantity of hydrogen recovered.

COTS = Commercial off the shelf

The H2 in this application is actually a isotope of H2 and is of high value hence the desire to separate it for recycling. The value of the material changes the economic equation significantly.

what is pipe delivery diameter? how fast is the hydrogen moving through the pipe? how much room have you got to play with? If a lot then why not allow you stream to slow down by passing through a large sphere top of sphere has a small pipe, hydrogen has a very large boyancy force in air and will rise up the pipe, a valve normally shut can open when the hydrogen concentration builds up?? I don't see why you need a fancy membrane?

HydroScope- were you joking?  Separation of gas or miscible liquid mixtures by density difference alone just plain doesn't work.  Density difference can slow down the admixture of gases/vapours, but it won't re-separate them once they've been mixed!

OWG007:  you're not going to find the device you need on anybody's shelf IMHO.  If you want to use a membrane separation method to do a decent job of concentrating your (hopefully deuterium, not tritium!) out of your air stream, you'll need to compress it to at least 100 psig to provide driving force to get it across the membrane- vacuum on the hydrogen side of the membrane would almost certainly not be enough driving force for a membrane of an economical size- but who knows- maybe somebody has a really expensive but hyper-efficient membrane for this job.  There are flammability concerns if you compress this mixture...at 2% H2 in air, you're near the LEL even at atmospheric pressure, and an increase in pressure will reduce the LEL further.  Compressing the mixture won't be hazard-free.

If the deuterium is really valuable to you, you could consider cryogenic temperature-swing adsorption of the air components, letting the deuterium pass through- but at 30 cfm you'd need a LOT of cooling to accomplish that...

Another option is to combust the deuterium and collect the heavy water by condensation- you can re-generate the deuterium again by electrolysis.  Not much use to you if the air stream also contains ordinary water vapour, because you'd be into a very complex separation process to upgrade the isotopic purity again. Perhaps you could scrupulously dry the air stream prior to combusting it- that would help give you the isotopic purity which makes the deuterium valuable.

Sorry, but I don't have any vendors for you to consider.

Genesis Fueltech in Spokane Washington utilizes a selective membrane to separate H2 from the reformate stream in their alcohol fuel reformer.  You might want to contact them.
http://www.genesisfueltech.com/tech.html

First issue,  some gas membrane companies for H2

Air Liquide/Medal (my company)
UOP
Air Products (I'm pretty sure)

This sounds like a small volume specialty application, so it might not be interesting on a size basis for a gas company.  (in other words, not enough $$ to do the the design work, unless you're willing to pay $$$)

The important design criteria for a membrane is pressure ratio, not absolute pressure gradient, so a vacuum pump may work.   The explosion issue and residual vacuum oil, I will leave aside for a design engineer. (I'm in membrane R&D)

The real issue to me is how pure do you require your product?  Polymeric membranes will not get really high purities from a 1-2% feed, say 50%, with a moderate recovery.  (purity and recovery are a tradeoff)  If you require high purities, VSA or cryo would be better bets.  I believe that someone sells (maybe one of the GC companies??) a Pt/Pd membrane for super pure H2, but the flux is very low.

Dave Hasse
Air Liquide
dave.hasse@airliquide.com

Oops,

talked to my boss.  He pointed out for a high value stream, such as this I assume, multistage membranes with recompression can give much higher purities and better recoveries.  At which point I went "doh"...

Dave

If you're still after this one, one clarification to my previous note.  You'll want to scrupulously dry the stream to increase the isotopic purity of the deuterated water you ultimately collect, then CATALYTICALLY combust the deuterium over a supported platinum catalyst, then condense/dry out the deuterated water- no fire nor any added fuel as these will add non-deuterated water to the mix and screw everything up.  Deuterium should combust at pretty low temperatures over a platinum catalyst, and you shouldn't need a gigantic amount of catalyst to combust over 99% of the deuterium in the stream.  You can electrically heat the catalyst to get the reaction going.  At 2% fuel in air, the heat of combustion will be significant but manageable- you may need to moderate catalyst temperatures by diluting with a bit of (pre-dried) additional air.  Of course, you'll need to be a little worried about flame propagation back into your source of this material, but this can be handled by proper combustor design.

You can then regenerate concentrated deuterium wherever/whenever it's needed by electrolysis of the collected heavy water.  The idea here is that the deuterium will be far easier to handle and separate from the air stream as heavy water than as a gas.