Stratification of Nitrox Mixes

[Jackal]

Having just completed the DSAT Blenders course, thinking about this and applying some aeronautical engineering knowledge, and seeing a PADI Master Instructor shaking a tank prior to analysing the O2 content of a tank, I have a short question regarding the stratification of Nitrox mixes. I thought the shaking of the tank was a pointless exercise prior to analysing it. For those non-technical people out there, stratification is the seperation of two gases or fluids, like water on oil.

Does stratification within a tank occur ? I argue that it doesn't with the reasons as follows.

Air is simply a nitrox mix of 21%. If it stratified over time, resulting in a layer of Nitrogen on top of Oxygen ( since Nitrogen has a molecular weight of 14.0067, and Oxygen 15.994), then if you attached a regulator on top of a tank, you wouldn't be able to breathe it until you shook it and stired it up (since all you'd be breating in would be Nitrogen ).

The homogenous nature of a gas means that it's constituent molecules are evenly distributed within the gas. Unlike a fluid, the gases have a continual motion, so are constantly mixing, but over time, this motion slows due to molecular friction with the tank walls. It would be interesting to leave a EANx 40 tank standing up for about a week, analyse it ( without moving the tank ), then shake the tank and re-analyse to note the different. Would it continue to be a 40 mix ? Thoughts ?

Fortunately in diving, the regular movement of a tank on and off the boat etc, would be sufficient to ensure a complete mix of the gases if they did stratify. So the only real reason for this question is when analysing, which is quite serious. If a EANx mix does stratify, and you analyse it, you might get majorly incorrect O2 contents, resulting in a significantly dangerous situation for the diver.

Now the really interesting question is Trimix. Since the molecular weight of Helium is only 4.0020 (ie: 1/4 the weight of oxygen ) does Helium Stratify from the rest of the air mix making up the Trimix. I've heard that if a tank has a slow leak around the valve, and it the tank is standing up, the Helium can bleed off first, since it is pooled around the upper end of the tank. This would result in a significant change in the tanks contents, and obvious danger to the diver. Anyone with any trimix knowledge, please comment on whether there is a requirement to re-analyse a tank within a certain time period of a dive.

[cyklon_300]

http://www.scubaboard.com/showthread.php?s=&threadid=17982&highlight=physics

[MechDiver]

There was a thread on this recently and your answer is

maybe :rolleyes:

Actually, I do not believe a consensus was ever arrived at. Roll the tank, wait for awhile, do it now, whatever pops your cork.

As far as the trimix leaking thing, helium will leak from a connection that "air" will not, or at least I've been told that. I analyze anything before I dive it.

Phil

[SPEEDSTER]

my guess is no on the stratification, I test the mix right after filling, then mark the tank. They don't sit around long enough for anything to seperate, but who knows. I don't think I would re-check the tank again anyways even if it did sit around, just the pressure as I geared up.

But what I'd really like to know is how was the DSAT program?

mike

[ianr33]

Just to be pedantic ('cos I used to teach High School Chemistry a long time ago) Nitrogen goes around as N2 molecules so has a relative Atomic Mass of 28.01. Similarly Oxygen is O2 and has a mass of 31.99.

I dont know of the top of my head what the velocity of molecules in a gas is at room temperature but I do know it is incredibaly high (100's of meters/second???? ) which means that gases mix very rapidly.

I dont dive trimix,but if I did the least of my concerns would be shaking the tank

As a final thought,if gases stratified the CO2 in the atmosphere (mass 44) would all fall to the bottom,making breathing somewhat difficult

[SPEEDSTER]

ianr33 once bubbled...

As a final thought,if gases stratified the CO2 in the atmosphere (mass 44) would all fall to the bottom,making breathing somewhat difficult

I thought about that as well, but I think that natural factors, ie weather, wind, earth's rotation, etc, would keep the atmosphere sufficiently mixed up.

[ianr33]

Ah,but what about those very still,temperature inversion type days you have in Southern California :)

[WreckWriter]

based on theory it does not occur. Based on experience it does occur.

I recommend rolling single tanks, standing/laying down doubles a few times prior to analyzing.

Tom

[SPEEDSTER]

ianr33 once bubbled...
Ah,but what about those very still,temperature inversion type days you have in Southern California :)

ahh yes, the reason I left Florida for California, well, we've still got the earth's rotation to keep our air breathable, plus, sometimes the earth shakes just for good measure.:D

[SPEEDSTER]

WreckWriter once bubbled...
based on theory it does not occur. Based on experience it does occur.

I recommend rolling single tanks, standing/laying down doubles a few times prior to analyzing.

Tom

Tom,

how much of a difference have you seen and what method do you use to do your fills?

mike

[WreckWriter]

SPEEDSTER once bubbled...


Tom,

how much of a difference have you seen and what method do you use to do your fills?

mike

We do partial pressure fills. Any other method likely would not see stratification...

I see an average difference of 1-3% O2 overnight, later analysis being higher in O2 percentage. Not much but enough to prove O2 stratification to me. I have no doubt that helium stratifies to an even greater degree but I don't have an He analyser to prove it.

Tom

[SPEEDSTER]

WreckWriter once bubbled...


I see an average difference of 1-3% O2 overnight, later analysis being higher in O2 percentage. Not much but enough to prove O2 stratification to me. I have no doubt that helium stratifies to an even greater degree but I don't have an He analyser to prove it.

Tom

hmm, perhaps this is more of a slow mixing of the gases than stratification? If the oxygen and nitrogen were seperating, the o2% would become lower over time, nitrogen is lighter than o2, so you would get a higher % of nitrogen in the reading. I'm assuming you store your tanks upright, by the way, otherwise this would be the opposite.

regardless, your experience supports the other thread cyklon posted.

-mike

[WreckWriter]

SPEEDSTER once bubbled...
hmm, perhaps this is more of a slow mixing of the gases than stratification?

Agreed, not permanent stratification, just slow mixing which creates a temporary stratification.

Tom

[spaz]

ummmm........no.


No combination of temperatures or pressures acheived by mixing SCUBA gasses, can cause the gasses used to achieved any physical state other than 'gas'. Gasses disperse infinitly into each other, completely irrespective of individual molecular weights, and depending on temperature they do it real fast (real a** fast).

Thank you, that is all I have.

Peace

[SPEEDSTER]

spaz once bubbled...
ummmm........no.


No combination of temperatures or pressures acheived by mixing SCUBA gasses, can cause the gasses used to achieved any physical state other than 'gas'. Gasses disperse infinitly into each other, completely irrespective of individual molecular weights, and depending on temperature they do it real fast (real a** fast).


spaz,

I see you are a chemist by day, and as such, I would have a hard time disagreeing with you, especially as I DO agree with you...however, what would lead to a change in o2% in Tom's tanks over time?

I initially thought that there would be an immediate mixing of gasses w/in the tank and would not change over time either, however I am not about to call Tom a liar as I believe him to be an all around good source for unbiased information. (scuba wise at least;) )

So, what do you think caused the variation in o2% readings?

-mike

[Friggincold]

Helium locks for your tanks valves:) get em wow you still have HE in your tanks don't let it drift away:bonk:

[ianr33]

Does the readout from an O2 analyser vary with temperature?. I would imagine that it does. If you calibrate the analyser at room temperature ,then pass hot gas from a just filled tank thru it,could that explain the difference.

[Walter]

Rolling tanks is effective, but it does not mix the gasses. Injection of gases at high pressure mixes them more than rolling ever could. Additionally, the gases would mix almost instantly even if they weren't mixed under pressure. Rolling cools a tank which will result in a more accurate reading.

[Genesis]

are indeed temperature sensitive.

They are, after all, just a fuel cell. And chemical reactions (which is how they work) are temperature dependant.

Most sensors contain internal temperature components (typically a calibrated thermistor) which attempts to compensate for ambient temperature differences. The problem is that the ambient temperature around the sensor is not the temperature of the gas impinging on the membrane, and immediately after a fill it could be QUITE HOT.

For example, its pretty common around here to get a fill of Nitrox in an AL80 that reads 3200 psi when the guy filling the tank is complete. A few hours later, with the valve closed, in 80F ambient temperatures, the gauge reads 3000 psi.

Simple application of Boyle's and Charle's law tells us how how the gas was - since we know how hot it IS.

To go from 3200 to 3000 psi there was a 6.6% reduction in temperature in KELVIN.

Since the endpoint was 299.66 K (~300K), the gas was at 318K, or 113F - at the completion of the tank fill.

Yes, that's almost 30F hotter.

This could easily account for an "off" reading if the sensor is not at gas temperature (and it won't be unless its immersed in the gas - the temperature compensating resistor is typically in the back of the shell, behind the membrane)

[WreckWriter]

so many fancy answers....

How many of you actually mix gases rather than just "know" gas "laws"?

Tom

[ianr33]

so many fancy answers....

How many of you actually mix gases rather than just "know" gas "laws"?


Hey Wreckwriter, I dont mix gases but I absoluteley believe your observation of the O2 % changing with time,question is why? I find it really hard to believe it is due to incomplete mixing. The change in temperature seems a likely answer to me.

[Genesis]

don't play favorites.

Nor do they get suspended because you think they are.

When you find an experimental result that conflicts with the laws of physics, you need to investigate why before just shooting off and declaring the law invalid.

Next time you have such a "suspect" tank, make one change to your analyzer setup - run the gas through a good length of wide (2" or so) copper pipe so that the gas comes to ambient temperature before it gets to your sensor. Do the same when you analyze it later. THEN compare notes.

I'd be willing to bet your "anomaly" disappears - what you're measuring is a sensor non-linearity caused by the gas temperature, rather than an O2 percentage difference.

[roakey]

Problem is, hox mix makes the O2 read high, not low.

I've seen this when my Nitrox cylinder got put out in the sun and my reference cylinder of air was kept in the shade.

Roak

[Genesis]

But in practice my assumption would also not be that the gas laws suddenly decided not to apply to that tank :)

I'd go looking for the cause of the deviation from expected results - there are a number of ways to do this, including transfilling a few hundred PSI from the "reads low" tank into an empty tank, then measuring THAT after both have reached equilibrium.

(Since the amount drawn off would by definition be from that first exhausted, the difference should be "captured" by this procedure. If its not.....)

[spaz]

Do the experiment!!!...Do many...I want answers...I want answers now!!!!
Thanks

Sorry, it is the manager in me talking.

[TheDecoStop]

in response to "Problem is, hot mix makes the O2 read high, not low."

I just took a quick look at some older sensor calibration charts that give the numbers to use as constants while calibrating a sensor at a specific temperature and humidity. The charts show the O2 reading decreasing with an increase in temperature at the same humidity. They state the main reason for this as a gas's ability to hold moisture at different temperatures. They then proceed to state the fact that humidity greatly affects the sensors readings.

This would lean towards the cooling effect of rolling or sitting and then analyzing as being the reason for the varied readings vs. gas mixing or stratification. I have done tests after mixing that have shown me that my readings are more along the lines of what I expect if I roll a tank a little and then analyze, or just let it sit for a bit and come back and analyze it later. I haven't spent too much time trying to figure out why, because I only view the act of analyzing as a confirmation of what I already know I put in the tank... as long as I did the math right! :-)

Sorry for the long winded post...

Rick

[MikeFerrara]

I mix alot of gas by PP blending. As far as nitrox goes I mix more 50% than anything. I won't pretend to be a physics expert but her is what I see several times a week. I can duplicate it at will and in fact will do it tonight for students.

Whjen I mix something like 32% I can mix and analyze and over time there will be very little change. When I can I add the O2 let cool, check the presure and top as needed then top with air.

If I am mixing 50% I use the same procedure but I am adding less air. The whip from the hyper filter is flow restricted and so is the hyperfilter. This gives a super slow fill rate and little heating. If I analyze immediately I can read as low as thirty something percent. If I let the mix sit for a couple of hours it will read 50%. If I roll the tank it will read 50%. As slow as I fill this much of a variation is not due to temp.

BTW if I don't use the hyper filter (flow restricted) and add air much faster I can analyze immediately and hit it right on and it will be on 24 hours later as well. This is totally repeatable.

Alot of air fast gives good numbers that stay good vs a little air slow slow requires time or further agitation.

In order to do a mix like this and analyze right away we must add the last bit of air fast, let the mix sit or roll the tank.

Once mixed (stable reading) the gas stays mixed.

I don't know what gas laws to apply but I don't believe you will get a homogenous mix instantly without some agitation. Anybody who says different hasen't done much PP mixing.

[MikeFerrara]

I did a little research and there seems to be much written on just how to get gasses to mix. There seems to be some products on the market also. If gasses mix so fast by themselves I couldn't really see the point. Last night I was finishing up a blender class where two of my three students work at Purdue. One is faculty and the other is staff in one of the engineering departments. BTW the third student is a Lawyer who started life as a biologist.

As I stated in my last post the class witnessed the fact that the initial reading after topping with air VERY SLOWLY was way way low (like 20% low).

We talked this over and none of them was surprised. None could present equations or specific laws that explained it though. One student did point out what happens when a smoker exhales. The idea was to use a visable gas as an example. Anyway the student who is on staff in the engineering department says the people who pretty much wrote the book on gas mixing are in his department. He has promised to try to get us a quote.

[ianr33]

MikeFerrara once bubbled...

We talked this over and none of them was surprised. None could present equations or specific laws that explained it though. One student did point out what happens when a smoker exhales. The idea was to use a visable gas as an example. Anyway the student who is on staff in the engineering department says the people who pretty much wrote the book on gas mixing are in his department. He has promised to try to get us a quote. [/B]

I look forward to the answers to this,it really does seem just wrong to me,yet it is obviously observed. Might have to change my belief system yet

:)

And just to be awkward,the smoke you see from a cigarette is a finely divided solid,so will be MUCH heavier than a gas and so is not a fair analogy

[Bottomfeeder]

MikeFerrara once bubbled...


Alot of air fast gives good numbers that stay good vs a little air slow slow requires time or further agitation.
<snip>
I don't know what gas laws to apply but I don't believe you will get a homogenous mix instantly without some agitation. Anybody who says different hasen't done much PP mixing.

I mix nitrox and trimix every week, and that is my observation also.

I often use a Haskel booster that is very slow. The most dramatic differences come when I adjust a mix by boosting O2 or He on top of already mixed gas. My setup is particularly slow, as I have to boost, wait for the drive compressor to catch up, boost, wait, etc. If I boost 200 psi of gas on top of say 3400 PSI, the tanks do not get even a little warmer. If I boosted He and expected 14% O2, the immediate reading will be something like 2 - 4%. When I check it the next morning, it will be maybe 10%. I have had it take up to 2 days to settle out. When I boost O2, I get ridiculously high readings for a day afterward.

If I boost and then put the tanks in the truck and drive 5 hours to dive, the readings are what I expect and stable when I get there.

If I mix normally and blow air on top of O2 or heliox the analysis is correct immediately.

All of this is both repeatable and reproducible