Please register or login

Welcome to ScubaBoard, the world's largest scuba diving community. Registration is not required to read the forums, but we encourage you to join. Joining has its benefits and enables you to participate in the discussions.

Benefits of registering include

  • Ability to post and comment on topics and discussions.
  • A Free photo gallery to share your dive photos with the world.
  • You can make this box go away

Joining is quick and easy. Log in or Register now!

From a pure physics perspective "tank rolling" to avoid gas stratification is clearly total bunkum!

Gases are driven by Brownian motion, the vibration and and random movement of their molecules in that gas. This movement is what gives us "pressure" because that pressure is actually those molecules hitting and bouncing off the walls of the tank. The more you compress something, the higher the pressure, because in any given volume there are more molecules hitting the confining walls at any given moment.

So what would be the mechanism for stratification of gases in a tank? Clearly the main one, assuming there is no direct chemical reaction, is gravity seperation due to the differences in gas density.

Here are the densities for these gases at 1bar 0degC

Oxygen: 1.43 g/l
Nitrogen: 1.25 g/l

Therefore oxygen is just 14% more dense, ie it experiences a 14% greater force in the direction of gravity.

This is clearly a small difference and driven only by the force of gravity and means that brownian motion is many time more dominant even at 1bar, and certainly way too small an effect to overcome the forces of brownian motion at hundreds of bars of pressure.

Industrial gas centrifuges that are expressely designed to seperate different densities of gas require high speed rotation that acts with tens of thousands of times the force of gravity in order to practically seperate gases with such small density differences.

And of course, rolling a tank itself doesn't mean the gas inside actually moves, as that gas has an inertia and only the small amount of boundary layer friction between the gas and the moving tank walls would drive the gas around with the tank.

Finally, the hole that routes the gas into a cylinder is pretty small, a cross section in the order of tens of square millimeters at most, which means the incomming gasstream velocity is pretty high, (which is also why you fill tanks slowly to avoid dynamic high pressure, and hence high temperature, zones) and that high velocity stream causes massive swirling and mixing of the gases inside the cylinder, once again rendering "rolling the tank" as a total waste of time.

So, we are left with the conclusion:

1) tank rolling doesn't cause the contents to mix
2) The contents mix themselves via brownian motion
3) the force of gravity is insufficient over any time scale to cause stratification due to different constituent densities




Finally, with regard to breathing pure nitrogen, ie at a PPo2 of zero bar. The old airforce tables give an expected time to loss of conciousness of as little as 5 seconds at pressures below 0.05 bar. At a true zero PPo2, your lungs will actually REMOVE the existing o2 from your bloodstream. Chances are, i'm going to suggest that a single full breath of pure N2, is likely to result in pretty much instaneous LoC, and assuming the next breath wasn't suitably hyperoxic death within a minute.
Sounds like plasma separation from blood when placed in a centrifuge.
 
Ok, simple question:

Demonstrate that tank rolling makes a difference.

Should be easy right?

But, how are you going to do this? How are you going to demonstrate that the gas inside your high pressure cylinder is stratified to a statistically valid level of certainty?

Someone just said "i have seen it with my own eyes"? er, what have you seen? Both Oxygen and Nitrogen are transparent gases, of similar density and kinematic viscousity, and are contained within a solid metalic high pressure container? So how have you "seen it"?

and don't say "my gas analyser showed me" because the variables in that sampling are too numerous to determine. In fact, to determine the true distribution of a gas sample in such situations you would need an expensively modified cylinder that allowed you to draw off, without contamination from the ambient air, samples from different parts of the cylinder, and a precision analyser that had an inert damper medium, or operated in a near vacuum to avoid cross contamination. Simply sticking a typical commercial gas analyser on the tank valve and saying look the composition is different or whatever is NOT proof of any kind.


And finally, if in fact O2 and N2 did stratify, it would happen with our atmosphere, which is made up of mostly Nitrogen and oxygen. But our atmosphere, although reducing in pressure with altitude, does not in fact change it's broad composition as you gain height (until you get high enough that solar formation of ozone etc occur) And it would of course happen in normal scuba tanks containing air under pressure! Imagine that massive bank of cylinders at your dive shop, that sit fixed to the wall, probably over 1,000 litres (at 1atm) of gas volume and with a single pipe comming of the neck of each (usually at the top) they too would stratify and the first person on monday morning to get a fill would get all the Nitrogen and no oxygen.........

A fair question. Many, many times I have gotten a partial pressure fill of nitrox. That means they add around 500 psi of oxygen - slowly into an empty tank. Then the tank is placed in a cold water bath and air is added, maybe slightly faster - I don't know. When the fill pressure is reached 2800, 3000 or maybe 3500, the tank is removed and I check with an analyzer.

The attendants used to tell me "I just filled that tank - you need to roll it and slosh it around before analyzing. I used to ignore them, because (to he honest) so many of them have no clue what they are talking about. But sure enough my mixes would be off. I am embarrassed to say I do not remember which way; I think they may come out too rich?

Anyway after SEVERAL times where the mix came out wrong, I disconnected the analyzer, rolled the tanks around for a minute or so and re-tested. The mix always came out right afterward - 2 minutes later and without adjusting the analyzer. I swear I would NEVER believe it, if I had not seen it.

Incidentally, nobody is claiming that the gases won't eventually mix (by brownian movement) or what ever other phenomenon is occurring, but it takes some time after the fill to occur. If the tank sits for an hour or so after the fill, there has never been a discrepancy that I witnessed. Possibly the natural mixing will occur in just 10 minutes, I don't know, but when they come immediately off the whip, it is best to roll the tanks around before analyzing.
 
  • Like
Reactions: OTF
From a pure physics perspective "tank rolling" to avoid gas stratification is clearly total bunkum!

Gases are driven by Brownian motion, the vibration and and random movement of their molecules in that gas. This movement is what gives us "pressure" because that pressure is actually those molecules hitting and bouncing off the walls of the tank. The more you compress something, the higher the pressure, because in any given volume there are more molecules hitting the confining walls at any given moment.

So what would be the mechanism for stratification of gases in a tank? Clearly the main one, assuming there is no direct chemical reaction, is gravity seperation due to the differences in gas density.

Here are the densities for these gases at 1bar 0degC

Oxygen: 1.43 g/l
Nitrogen: 1.25 g/l

Therefore oxygen is just 14% more dense, ie it experiences a 14% greater force in the direction of gravity.

This is clearly a small difference and driven only by the force of gravity and means that brownian motion is many time more dominant even at 1bar, and certainly way too small an effect to overcome the forces of brownian motion at hundreds of bars of pressure.

Industrial gas centrifuges that are expressely designed to seperate different densities of gas require high speed rotation that acts with tens of thousands of times the force of gravity in order to practically seperate gases with such small density differences.

And of course, rolling a tank itself doesn't mean the gas inside actually moves, as that gas has an inertia and only the small amount of boundary layer friction between the gas and the moving tank walls would drive the gas around with the tank.

Finally, the hole that routes the gas into a cylinder is pretty small, a cross section in the order of tens of square millimeters at most, which means the incomming gasstream velocity is pretty high, (which is also why you fill tanks slowly to avoid dynamic high pressure, and hence high temperature, zones) and that high velocity stream causes massive swirling and mixing of the gases inside the cylinder, once again rendering "rolling the tank" as a total waste of time.

So, we are left with the conclusion:

1) tank rolling doesn't cause the contents to mix
2) The contents mix themselves via brownian motion
3) the force of gravity is insufficient over any time scale to cause stratification due to different constituent densities




Finally, with regard to breathing pure nitrogen, ie at a PPo2 of zero bar. The old airforce tables give an expected time to loss of conciousness of as little as 5 seconds at pressures below 0.05 bar. At a true zero PPo2, your lungs will actually REMOVE the existing o2 from your bloodstream. Chances are, i'm going to suggest that a single full breath of pure N2, is likely to result in pretty much instaneous LoC, and assuming the next breath wasn't suitably hyperoxic death within a minute.

LOL,, you are wrong... I completely agree with you about why you should be right, but you are wrong. I would never have believed tank rolling makes a difference.. but it does. I have seen it way too many times. Don't believe me? I can certainly understand that and don't blame you.

Ok, simple question:

Demonstrate that tank rolling makes a difference.

Should be easy right?

But, how are you going to do this? How are you going to demonstrate that the gas inside your high pressure cylinder is stratified to a statistically valid level of certainty?

Someone just said "i have seen it with my own eyes"? er, what have you seen? Both Oxygen and Nitrogen are transparent gases, of similar density and kinematic viscousity, and are contained within a solid metalic high pressure container? So how have you "seen it"?

and don't say "my gas analyser showed me" because the variables in that sampling are too numerous to determine. In fact, to determine the true distribution of a gas sample in such situations you would need an expensively modified cylinder that allowed you to draw off, without contamination from the ambient air, samples from different parts of the cylinder, and a precision analyser that had an inert damper medium, or operated in a near vacuum to avoid cross contamination. Simply sticking a typical commercial gas analyser on the tank valve and saying look the composition is different or whatever is NOT proof of any kind.


And finally, if in fact O2 and N2 did stratify, it would happen with our atmosphere, which is made up of mostly Nitrogen and oxygen. But our atmosphere, although reducing in pressure with altitude, does not in fact change it's broad composition as you gain height (until you get high enough that solar formation of ozone etc occur) And it would of course happen in normal scuba tanks containing air under pressure! Imagine that massive bank of cylinders at your dive shop, that sit fixed to the wall, probably over 1,000 litres (at 1atm) of gas volume and with a single pipe comming of the neck of each (usually at the top) they too would stratify and the first person on monday morning to get a fill would get all the Nitrogen and no oxygen.........


I too totally agree with what your saying, and im not doubting the science and befor I started blending I would have totally called BS on the rolling the tank thing, but...

Please explain this to me and im sure others have experience this as well hence where the whole rolling tank thing originated from.

lets say I want to make an 80% mix. First I would boost in the o2 then if im to lazy to fire up the compressor and top off with air from a tank and boost the last little bit. If I do this really slow I can 100% guarantee you that if you were to analyze the tank it would show about 70%. Now if roll the tank around a couple times then it will read 80% bang on.

Please explain what is happening?
 
A fair question. Many, many times I have gotten a partial pressure fill of nitrox. That means they add around 500 psi of oxygen - slowly into an empty tank. Then the tank is placed in a cold water bath and air is added, maybe slightly faster - I don't know. When the fill pressure is reached 2800, 3000 or maybe 3500, the tank is removed and I check with an analyzer.

The attendants used to tell me "I just filled that tank - you need to roll it and slosh it around before analyzing. I used to ignore them, because (to he honest) so many of them have no clue what they are talking about. But sure enough my mixes would be off. I am embarrassed to say I do not remember which way; I think they may come out too rich?

Anyway after SEVERAL times where the mix came out wrong, I disconnected the analyzer, rolled the tanks around for a minute or so and re-tested. The mix always came out right afterward - 2 minutes later and without adjusting the analyzer. I swear I would NEVER believe it, if I had not seen it.

Incidentally, nobody is claiming that the gases won't eventually mix (by brownian movement) or what ever other phenomenon is occurring, but it takes some time after the fill to occur. If the tank sits for an hour or so after the fill, there has never been a discrepancy that I witnessed. Possibly the natural mixing will occur in just 10 minutes, I don't know, but when they come immediately off the whip, it is best to roll the tanks around before analyzing.

My guess is that rolling == waiting. If you did nothing for a minute vs spending a minute rolling you'd get the same result.
 
Remember your gas analyser is a reaction cell. It sits mostly at 21% O2 because it sits in air when you aren't using it and before you push your mix through it. So you push a rich mix through it, and it takes TIME for the o2 to both physically propagate into the galvanic cell, but also for the internal cell concentrations to balance. Go look at the reaction times on the datasheet for you particularly cell, for a 10-90% swing at say 21degC, those time can be in the order of minutes, so for high concentration o2 mixes i'm not surprised the second measurement shows a higher concentration because the cell is already saturated at a high partial pressure of oxygen from the first measurement!

It's also worth thinking about what happens when you fill a high pressure vessel with a gas. It's NOT Like pouring water into a cup, where the water is 1,000 times more dense and has 10 times the kinematic viscousity of the air that currently fills that cup. Both of those mean the water as you pour it in stays together in a "clump" (also thanks to massive surface tension, water tends to pul together into droplets) and sits at the bottom of the cup in a pool, pushing the much less dense air out of the way

However, when you pump say O2 into a cylinder with 1atm of air in it, that o2 and air mix, and fill the entire volume of the tank, the o2 does not sit at the bottom or top of the tank, but expands to an even pressure and fills the entire tank. As mentioned at normal fill rates say 10 bar per minute (120 litres a minute for a 12l tank) the velocity of the gas stream exiting the dip tube and into the tank is significant, and more than enough to cause the entire contents (which only has a mass of a few kg) to be entrained and to spin and to mix (more advanced readers can amuse themselves by calculating the reynolds number of that free stream as it exits the dip tube..... ;-)

But the biggest and most obvious problem with tank rolling is that you'd have to put a HUGE physical effort into shaking and spinning the tank to actually cause the gases inside to even move, let alone mix (as their densities are so close). Take a tank at 200 bar, which is 20 Newtons per square millimeter. That means the brownian motion of that gas is excerting a force of about 1,3 tonnes on each square inch of the tank walls! How hard do you think you would have to shake or roll a tank to in anyway get anywhere near such a force?

Now i'm not saying that rolling the tank is not changing your measured concentrations, but i am saying that rolling the tank is not the thing that is changing those concentrations measurements!
 
Oh, and BTW, if tank rolling did do anything, it would actually cause the gases to seperate more, because the centripetal acceleration would drive the gas with the highest density to the outside and the low density gas to the inside of the tank!


Remember your gas analyser is a reaction cell. It sits mostly at 21% O2 because it sits in air when you aren't using it and before you push your mix through it. So you push a rich mix through it, and it takes TIME for the o2 to both physically propagate into the galvanic cell, but also for the internal cell concentrations to balance. Go look at the reaction times on the datasheet for you particularly cell, for a 10-90% swing at say 21degC, those time can be in the order of minutes, so for high concentration o2 mixes i'm not surprised the second measurement shows a higher concentration because the cell is already saturated at a high partial pressure of oxygen from the first measurement!

It's also worth thinking about what happens when you fill a high pressure vessel with a gas. It's NOT Like pouring water into a cup, where the water is 1,000 times more dense and has 10 times the kinematic viscousity of the air that currently fills that cup. Both of those mean the water as you pour it in stays together in a "clump" (also thanks to massive surface tension, water tends to pul together into droplets) and sits at the bottom of the cup in a pool, pushing the much less dense air out of the way

However, when you pump say O2 into a cylinder with 1atm of air in it, that o2 and air mix, and fill the entire volume of the tank, the o2 does not sit at the bottom or top of the tank, but expands to an even pressure and fills the entire tank. As mentioned at normal fill rates say 10 bar per minute (120 litres a minute for a 12l tank) the velocity of the gas stream exiting the dip tube and into the tank is significant, and more than enough to cause the entire contents (which only has a mass of a few kg) to be entrained and to spin and to mix (more advanced readers can amuse themselves by calculating the reynolds number of that free stream as it exits the dip tube..... ;-)

But the biggest and most obvious problem with tank rolling is that you'd have to put a HUGE physical effort into shaking and spinning the tank to actually cause the gases inside to even move, let alone mix (as their densities are so close). Take a tank at 200 bar, which is 20 Newtons per square millimeter. That means the brownian motion of that gas is excerting a force of about 1,3 tonnes on each square inch of the tank walls! How hard do you think you would have to shake or roll a tank to in anyway get anywhere near such a force?

Now i'm not saying that rolling the tank is not changing your measured concentrations, but i am saying that rolling the tank is not the thing that is changing those concentrations measurements!

You’re wrong, long time ago I worked at lab for gasses. We had equipment for rolling tanks.
 
From a pure physics perspective "tank rolling" to avoid gas stratification is clearly total bunkum!

My experience is different.

Circa 2000 I started partial pressure blending a lot of nitrox tanks, mostly AL 80's. I owned a little dive shop in the Keys at Cudjoe Gardens Marina. I had a maximum of 12 divers and I would try and blend the nitrox the night before so it could homogenize. Sometimes that did not happen.

I would learn that I needed 2, 4 or whatever more tanks about one hour prior to the boat leaving the dock. When I first had to do this I had heard that rolling the tanks would make it homogenize faster, I was skeptical.

So...I blended 2 tanks and stood them up, blended 2 more and rolled them around for 5-7 minutes. Lo and behold those I rolled analyzed out more stable than those that were standing up.

Was this a controlled experiment? No..

Is it possible I am wrong? Yes...

But my experience is otherwise.
 

Back
Top Bottom