To Build an Oxygen Concentrator or Not?

Interesting project. If you decide to proceed I hope you will let us watch the progress with regular updates.

.

.

I have a couple comments and questions:

.

.

1. I think your estimate of 5ml/min concentrated O2 output by the typical medical unit is off by several orders of magnitude. Output is probably far greater.

I'm basing this mainly on the idea that for an O2 concentrator to benefit respiration, the flow rate would need to be at least a significant portion of a patient's probable maximum instantaneous breath flow rate.....otherwise it would just be concentrating it to have it diluted.

In a minute sitting here relaxed I make roughly 10 breaths of around a pint each. I would guess I double that when winded. This provides a very rough estimated range of 1.25 gal./min to 2.5 gal/min. Assuming the machine should make at least 40% of the max (to give a nice round figure) suggests the machines are probably producing close to 1 gal/min rather than 5 ml/min.

.

.

2. 1 gal/min of 95% O2 @stp still isn't a huge amount, but it certainly makes the prospect of operating a few units in parallel or constructing one with sufficient flow so that operation is on demand only. My point here is that on-demand avoids the hazards involved with creating one off system to accumulate and store pressurized 95% O2 in your home shop for later use.

.

.

3. If you construct your own system, I would avoid using PVC to construct your own canister. A PVC canister filled with 95% O2 and exposed to cycling pressures, possibly occasional intense ultraviolet radiation, and occasional shower of glowing metal sparks is eventually going to end badly.

The cyclic pressure and volume of 95% O2 under pressure within make selecting an adequate container (and associated control/delivery systems) for the environment an important safety consideration, beyond just the reverse threaded connections.

That applies for any system whether demand or accumulation based. The probability and severity of mishaps accumulates right along with amount of O2 you accumulate.

.

.

4.I looked into something similar a while back and while i don't remember where to tell you to look for the reference at the moment, one of the critical issues for economical production is maintaining very clean air into the zeolite. Oil, certain gasses and other contaminants can quickly foul the zeolite... You wouldn't want to have a fine mist of oil in concentrated o2 anyway.

The point being, air contaminants around welding are not typically what a medical o2 concentrator is exposed to. Whether you construct you own or use another, steps to insure only very clean dry air reaches the concentrator will probably make operation much less expensive.

.

.

5. You mention possibly separating off Argon. What method were you considering for separation?

.

.

Good luck. Screw luck, actively steer yourself to the good outcome.

I looked back over my blog, and all the mL should be just Liters so you are correct I am off by an order of magnitude, I will go back and make the changes to the blog.

Good point about PVC degrading over time and becoming a flammability issue in the presence of 95% O2, it wouldn't be that difficult to use a steel pipe, although still doesn't need to be medical grade.

I do already have an filter dryer on the output of my compressor, although perhaps some sort additional filter on the unit might not be a bad idea. When I weld I am either just inside the door of my garage with the door open, or just outside the door so the air quality shouldn't be that bad.

With the system that uses two different zeolites the first zeolite takes out the nitrogen and the second type takes out the remaining argon. If I separate the zeolites in two different canisters in series, and put a solenoid between them that shuts on the low pressure swing, the output should be mostly argon.

i have a medical oxygen concentrator , the figures are right at 5 litres / min , that is a high rate in fact

typical is

1.5 litres min @ 90 % O2

3 litres min @ 60 % O2

5 litres min @ 40 % O2

Nice project but be very careful because any oil fumes from the compressor will likely accumulate somewhere and self ignite in the high oxygen content. It is not guaranty that the sieve will remove everything. Try to use an oil-less compressor and clean the piping thoroughly before use.

We use a lot of pure oxygen in our ozone systems and all the O2 (and O3) pipes must be oxygen clean.

Make sure that the room is well vented. O2 accumulation is extremely dangerous. Bringing the ambient air, O2 level to 22-24% makes anything ignite easily and burn very fast.

By the way, I have an old industrial unit that will be given away for the metal. Needs the controls though. I don't know if the solenoid valves work. Stop by our shop to look at it if you drive by Montreal.

In my research making sure the incoming compressed air is clean and dry is something I know is important. Adding a filter and desiccator is something I had considered.

Not planing any trips to Montreal anytime soon, but I have done the drive before, although not in a while, but getting a broken unit for cheap or free would be intriguing.

Frankd20:

.

I've been thinking more about your project (and about a possible similar project of my own).

.

I haven't had any luck digging up specific information on pressure and O2 concentration responses to demand for various models.

.

I am assuming that at zero flow conditions, delivery pressure would match feed air pressure, which I understand is typically 20 psig. Any ideas on typical delivery pressure at various percentages or rated flow, say 10% 20% 50%? It would be great to have some graphs relating pressure/flow/O2 concentration.

.

Also, information on the amplitude of the pressure pulses and any changes in such at various flow rates would be useful....or at a minimum knowing an upper limit for pressure pulses and at what flow this is likely to peak.

.

Strong pulses will cause excessive turbulence in the flame as well as a continual oscillation between oxidizing and reducing flame conditions. If strong enough the flame might be too unstable to reliably stay lit, or perhaps fuel might be sucked a short distance back into the oxygen supply...

.

As an aside, observations of a pressure gauge installed as original equipment on a medical unit could provide evidence that pulse severity is sufficient to warrant addressing, but unknowns like the gauge response time and any dampening in the gauge (which would be desirable for ease of reading given intended original use) mean that significant pressure spikes might exist that are not indicated on the gauge.

.

If the pulses are strong enough that they need to be reduced/smoothed out, typical solutions will create flow restrictions and/or new stored volumes. Minimizing the storage of pressurized O2 is a high priority in my perspective, which favors solutions which will likely result in added flow restrictions.... which brings us back to delivery pressures...

.

In my estimation, if you use medical O2 concentrator, as-is, ability to meaningfully adjust O2 delivery through a torch (beyond 'on' or 'off') will be almost non existent. When attempting to adjust O2 flow with a regulator or throttle valve, after initiating flow, a point will probably very quickly be reached where adjusting further results in no increase.

.

I'm basing this on relatively low delivery pressure (under 20 psig) with on demand replenishment and significant reductions in % O2 with increasing flow. Unlike when adjusting the torch hooked up to an O2 tank, attempts to increase flow will cause at least slight supply pressure drop combined with meaningfully lower O2 concentration in what is being delivered. The increase in flow will be largely offset with the decrease in concentration.

.

I'm also uncertain what flame effects might be noticeable as additional nitrogen is delivered at increased flow rates.

.

While the flame could still adjust without changes to O2 flow; from say a carburizing flame to sharp white-blue oxidizing flame, by reducing the MAAP or acetylene flow, there won't be much ability to increase the size of an oxidixing or neutral flame.

.

A couple ideas for counteracting probable limitation on the ability to adjust O2 delivery and potential problems with pressure transients:

.

- Feed Air pressures should be made as high as the process and safety considerations allow.

.

- Designed maximum flow rates (and therefore the amount of zeolite used) should be far in excess of the flow rates at maximum intended use. I'm guessing designed max should at least 10 times greater than max intended use, to keep concentrations somewhat consistent.

.

-A system that controls cycling of multiple independent zeolyte chambers sensing demand via changes in delivery pressure could be used to provide high, relatively constant concentration O2, over a broad range of flow rates. Multiple independent chambers cycled in succession could also reduce pressure pulses.

-

By the way, thank you for enlightening me on the potential collecting argon in systems like these.

In an electronics facility I was associated with, we investigated Nitrogen concentrators. The bypass gas was the concentrated Oxygen atmosphere that remained. Same concept with molecular seive.

You might look at second hand units from that industry instead and they would already have teh necessary plumbing and pre-filters in place.

Have fun.

My concentrator, I have emphysema, delivers up to 5 litres.

I have encountered oxygen concentrators used with ozone water sterilization systems- I am not sure what technology they use, but we are likely talking pretty small quantities. Anyway, these units REQUIRE and air drier to work properly. Apparently, atmospheric moisture can seriously degrade your media...

PS. Build it. Why? Because you can...

You could build yourself an acetylene generator while you are at it.

Any ideas how to do that in a way that is more economical than just getting the tank filled?

I used to have some calcium carbide when I was younger, but I wouldn't want to have to stock up on that either, but it was sort of fun to play with.

It is handy to know how much gas will be required for the welding operation. For my Smiths Airliner torch I use the AW203 tip the most for 1/16 to 3/32 steel.

The AW203 has a tip size of a #67 drill & with 10psi oxygen & fuel gas uses 2.3 SCFH

(standard cubic feet per hour) of gas.

The AW205 for 1/8 inch steel with a tip size of a #57 drill & with 10psi oxygen & fuel gas uses 6.0 SCFH of gas.

My largest tip AW209 for 3/8 steel with a tip size of a #49 drill & with 10psi oxygen &

fuel gas uses 23.0 SCFH of gas.

These numbers are from the Smiths instruction manual.

Are those just for welding tips, what about cutting tips? 23 SCFH is a little over 10 L/min. For the medical ones as others have stated the % of O2 drops off with higher rates. I think my estimates of how much O2 I would need are about correct.

Yes the comments were for Smith airliner torch welding tips.

The MC12 series cutting tips use more oxygen.

MC12-00 for 3/16 cutting uses about 30psi oxygen at 7SCFH preheat & 30SCFH cutting.

MC12-0 for 3/8 cutting uses about 50psi oxygen at 7SCFH preheat & 46SCFH cutting.

MC12-1 for 5/8 cutting uses about 60psi oxygen at 9SCFH preheat & 81SCFH cutting.

I currently use 4 5lpm medical units for a glass torch. The idea u present is one I have been investicating as well. I have modified my units to increase flow and pressure, by upgrading compreSsor and removing internal regulators. The 4 units are connected together and than storage tanks are used to elimintate the breathing of machines. I plan to use an oileSs compressor to boost pressure into a holding tank. I feel limited by the medical units though because they are designed to breath off of and not intended for large volume or pressure. I have investicated industrai psa equipment but the cost has me investigating other routes.The simpiicty of these machines leads me to believe a home built unit could be built cost effectivly. I also have done some test on oil and O2 mixing and can't get any spectacular events to happen, ie explosions. Not that I recomend it in any way but I feel like the mixing of the 2 bad thing happen when you have high pressure o2 +500psi. These systems never come near that level.

'.... I also have done some test on oil and O2 mixing and can't get any spectacular events to happen, ie explosions. Not that I recomend it in any way but I feel like the mixing of the 2 bad thing happen when you have high pressure o2 +500psi. These systems never come near that level.....'

.

Pressure in a compressor can be many times greater than the gauge pressure at the output. A mechanical problem with a compressor can exacerbate the problem.

Please take necessary precautions to keep the feed air and the compressed oxygen oil free. Not only does oil create an explosion risk, it also fouls the media.

Were you able to build it?

I'm trying to build one and struck up with calculations over how much Compressed Air do I need to get an amount of Oxygem at 60 psi.

^ generally you need double the psi going into sieve bed. So 120psi in will yield 60psi. Also to consider is the flow, so if you need 10lpm you will need to provide 40+lpm of compressed air. It would be beneficial to boost pressure after it is purified to get 60psi.

Just for information... I am another glassblower that uses medical concentrators for oxygen. one 5 lpm machine wont run a very big torch, even after it has been altered like the glassblower above described. But, you can use a compressed air tank as a holding tank and daisy chain multiple machines together and get more oxy that way. now about how to go about some of this on the cheap... I now have 16 5 lpm airsep concentrators sitting on my carport that I got for FREE!!! Contact your local medical supply places. they tend to dumpster machines that are out of warranty and having issues.its usually a matter of needing sieve beds re poured rather than the compressor etc going out. My cousin is on oxygen,so I asked his supplier and I got a call and picked up 16, thats all that would fit in my van. They cut the cords off for liability reasons. The often have 15 lpm machines and 10 lpm machines. I paid the worker off in glass work and jewelry for his wife.. It pays to bribe/ be nice lolololol.So, you may be able to get a few free machines that need minimal work to be operational again and useful for your purposes.

I stumbled across this thread by way of another glass person/scientific glassblower/ electrical engineer. I have managed to find the sieve material now, and am researching on how to safely re pour the sieve beds. I am also reading that you can bake the material and rejuvinate it, but am concerned about possibly contaminating my glass kiln, releasing toxic stuff into my shop etc. Any one have any input about that?

thanks in advance, Candice.. The Estrogen Engineer

Any body can suggest where to get the A5 zeolite from?

Many thx

I'm very interested in the paper you found about how to make an oxygen concentrator for hospital use, could you please send me the link to it?

i am in desperate need of a oxygen constrator that will work on dc. must deliver 6 liters continous. no poatable goes over 3. can you help me? harry french ;540-345-9115. thanks

As a quick solution find a 120v 10lpm machine, and use a off the shelf 12v invertor. Also you could look into changing out the compressor in a medical concentrator to a DC compressor, only problem would be powering the electronics that controlled the seive beds from the DC source. The electronics could be easily hacked to accept DC current since they run off DC anyway.

Hello Sir,

I read about you making your own oxygen concentrator I am also interested in making one for my use I just have some queries about the instrument like:

What are the dimensions of the tank required by you to build a concentrator giving a flow rate of 20LPM.

How much amount of zeolite was required by you.

I would be glad if you would be able to answer my queries by giving some of your time.

Awaiting a reply.

Thank you.

Hello sir,

I read about you making your own oxygen concentrator. I am also thinking to make one for my use but have some queries regarding the instrument which I think you would be able to guide me through my queries are:

1. What are the dimensions of the cylinder required by a concentrator giving a flow rate of 20 LPM.

2. What is the amount of zeolite required by the same.

I would be greatly thankful if you would give some of your time and help me through.

Awaiting a reply.

Thank you.