After a tank cools

I am sure this could be covered in class, but if others are anything like me, when you get onsite, and are chatting with friends, it's easy to take your eye off the ball so to speak and not really pay attention. Doing the mix check pre-dive also shows that the mix may change slightly over time. Especially in PP fills where air is used to top off and the mix is not done hard, the ppO2 may creep up.

This quote got me thinking. I am not new to diving with enriched air (which for the purposes of this thread I will call Nitrox) - recreational and technical/advanced. I also teach recreational divers to dive with Nitrox. I do not, however, yet fill tanks with Nitrox (haven't taken a gas blending course just yet). In all my experiences, the difference in the Nitrox % from checking it at the shop after the tank has been filled and then again on the site, the O2% reading has always decreased slightly - no more than 0.5%. Our shop uses the partial pressure mixing method (referenced to as PP fills in the quote above), where they add O2 into the tank first, then top off the tank compressed air. They have a gas blending program so they know how much psi of each to add - in case there were others out there that don't know how this is done.

So what kind of danger could a diver be in if they planned a proper dive to the Maximum Operating Depth (MOD) of a gas, using a recreational 1.4 ATA max, knowing the % of Nitrox can change over time due to cooling and settling?

Referencing my handy Nitrox depth table showing the MODs of 1.4 and 1.6 for 26% - 40%, let's first run through an example.

A diver planning a dive to 99' at the maximum recreational 1.4 ATA limit could dive a 35% mix. That mix would have to increase 5% to a 40% for 99' to reach a 1.6 ATA. An increase of 5% is also the constant difference for a blend with a 1.4 ATA at "depth X" and a blend with a 1.6 ATA at "depth X". I am not aware of a fill settling or cooling in a manner that would cause an increase in O2 by 5%.

Again, our shop mixes using the partial pressure method, and when there is a change it's a slight decrease in the O2% after settling/cooling. So, barring there being an analyzer error (which can lead to more problems than the one we are discussing here), can a tank filled with Nitrox tank settle or cool in a manner that would lead to an increase in O2 by 5%, which could make that mix potential harmful* for that diver and their dive, if planned to the max MOD at a 1.4 ATA? *From DAN If You Dive Nitrox You Should Know About OXTOX, "The possibility of oxygen toxicity at 1.6 ata is low, but the margin of error is very slim compared to 1.4 ata."

Since I have never taken a gas blending course, I am curious to hear from those who may be able to shed some light on this issue. It would also be excellent information to add to my Nitrox course!

NOTE: I understand that technical diving needs fills to be specific and accurate to the 1/10% due to the level of increased planning down to the foot and minute. The quote above and subsequent thought is in regards to recreational diving, dive planning, and MODS for recreational mixes ONLY.

I rarely use Nitrox but do air fills often (my tanks only) so I would think that it would be more likely for the O2% to increase rather than decrease with cooling.

You aren't putting in as much O2 percentage wise as you are air so the air cooling should have the greater effect thereby making it a richer mix.

I guess it depends on how full the tank was in the first place, what kind of tank (HP) and how fast the tank was filled.

I'll note again that this is just my guess and is not based on experience actually doing Nitrox fills.

I'll be interested to hear the "real answer" from someone with more practical knowledge.

I have seen tanks filled with air change by more than 5% after cooling.

I have seen tanks filled with air change by more than 5% after cooling.

The 5% I am referencing in my post, is in the O2 content - for example the difference between a 35% mix and a 40%.

The difference in % I have experienced in the settling/cooling has been minimal (within 0.5%) and it has been a decrease in % from what is checked at the shop, catalogged, and then marked on the Nitrox sticker on the tank. I wonder what the mechanics are as well. But more importantly, I am curious if a method of filling Nitrox tanks could effect an increase in the O2 % in the mix, not the tank pressure, by 5% or more (35% vs. 40%), which as I noted above would be the scenario that could lead to a dangerous dive situation in a recreational setting if a diver has planned to dive to the acceptable trained limits of a 1.4 ATA.

Just goes to show, even engineers don't know diddly.

The decrease in F02 is exactly backwards from what I'd expect, so I'm at a total loss to describe it. I'd figured the O2 going in cool from a tank, then the air going in slightly warmer from a bank, or significantly warmer 'right out of the compressor' (thus the warmer bank/cascade). In that case, as the warmer air cooled, it's pressure would drop which would end up with a slightly higher F02. From everything I've read on real vs. ideal gas laws, there's too many fudge factors to come up with a solid number.

Can someone explain the correct (as observed by Nicolle) gas dynamics that'd account for an FO2 drop? Write slowly, don't forget I are an engineer. I'm *really* curious to know just how messed up I am! About the gas, that is.

Can someone explain the correct (as observed by Nicolle) gas dynamics that'd account for an FO2 drop?

So I'm not nuts? The FO2 can drop after cooling/settling? I called a buddy of mine to see if he recalls that to be the case between shop and site as well, and he was like "yep - the O2% can be a tad bit less when you check it at the site - don't know why though." Maybe our brains freeze up here due to the cold and we are all nuts. DD was nuts, he got in the water both days while here this weekend!

This is curious. I noticed when I took my Nitrox course last month that the tank which supposedly was filled with EAN32 the day before only tested to have 31.5% O2.

One possible explanation (I guessing now) is that if the tank heats up during the fill, the partial pressure of the Oxygen in the mix drops slightly faster with temperature than the partial pressure of Nitrogen. This is logical considering that, for all pressures in the range of a tank fill, Oxygen boils at a higher temperature than Nitrogen. At extremely low temperature, like minus 250 deg f, all the Oxygen will liquify and there will be 100% Nitrogen in the gas phase. (Sea Water is a little stiff at this temperature too.)

For more modest temperature drops, the effect would be much less dramatic, but could possibly result in a small percentage change in relative partial pressure of the two gasses. I don't think it would be as high as 5% though.

Edited by BubbleBoy, 15 May 2006 - 06:22 PM.

I have seen tanks filled with air change by more than 5% after cooling.

The 5% I am referencing in my post, is in the O2 content - for example the difference between a 35% mix and a 40%.

The difference in % I have experienced in the settling/cooling has been minimal (within 0.5%) and it has been a decrease in % from what is checked at the shop, catalogged, and then marked on the Nitrox sticker on the tank. I wonder what the mechanics are as well. But more importantly, I am curious if a method of filling Nitrox tanks could effect an increase in the O2 % in the mix, not the tank pressure, by 5% or more (35% vs. 40%), which as I noted above would be the scenario that could lead to a dangerous dive situation in a recreational setting if a diver has planned to dive to the acceptable trained limits of a 1.4 ATA.

I understood the question and maybe I'm not expressing my guess clearly (and it may not be right or make sense) but...

If you slowly fill t HP tank to 3,500 psi but aren't letting it cool overnight it can cool to 3,000 psi easily. That is a decrease in psi of 14%. Convert this to volume if you want to (taking into account tank size). This would result in a smaller percentage of air vs oxygen than when you planned the ratios at fill time. If you have less air than you planned then you should have more oxygen as a percentage.

If any of this is true then you should have a higher blend than you planned on.

I'm probably missing something and am probably not correct but I'd like someone who blends gases to explain why my scenario couldn't happen.

A couple points...
1)Blending formulas are based on 'ideal gas laws' not 'real gas laws'
2)Gases do not compress in a linear fashion
3)Bubble Boy's great explanation
but saving the most common factor for last....the gauges used
4a)the pressure gauge is less accurate at the begining and at the end.Example a 5000psi gauge is most accurate in the middle of it's range 2500psi.Your adding O2 at the least accurate part of the gauge.Every fill station is slightly different and you tend to develop 'fudge' factors for a particular set up that will not work on another set up.Also gauges come in different levels of tolarance which is defined as a % of error over the working range.
4b)the O2 anylyzer it self is effected by temp.,humidity,age,battery condition

4a is the one that has the greatest effect but if each one creates a slight error when combined then it becomes an noticable issue

I think that the change has more to do with the homogenizing of the gas than anything else. This becomes much more dramatic with Helium mixes than with Oxygen-Nitrogen blending. However, the same principle applies.

If you mix Trimix by blending Helium and banked Nitrox, it will mix pretty quickly. If you fill by partial pressure of Oxygen, Air and Helium, it can take several hours for the gas to mix. If you mix by adding oxygen and then topping with air to create Nitrox, it also takes a little while, albeit much less time than when partial pressure filling of Helium-based mixes, for the gas to blend. Lots of things can influence how long it will take to reach the final mix including how fast the air is added to the oxygen and the temperature change of the tanks caused by the fills (Amonton's Law which most people think of as Charles' Law).

It can get pretty complicated. For example, I partially fill my tanks at home since I can get a significant savings especially on the Helium. (Also, not every shop near me carries it for filling.) I always add Helium at home, very slowly, keeping the tanks at room temperature. If I have to add Oxygen, I do the same. Then, I take the dive shop for either an air or Nitrox top-off.

When I first started, my mix would be off by up to 1.5 percent on the Helium side after the shop top-off. I had to learn to compensate for the tanks heating up during the fill at the shop, causing the pressure to drop as they would cool off. After a while, you get to know the habits of your tanks and the compressors used to fill them. (Some compressors really heat up the tanks during the fill. Some don't.) So, you start to build in your own personal fudge factors of which gas to add or reduce beyond what the textbook equation says.

With my double 120s, I got to the point where I could get both the Helium and Oxygen to within a half a percent of the target from either shop. Now, I will have to start all over again with my little 27s for my CCR. Maybe it will be easier since the tanks are smaller.

It's not as easy as you might think. Mixing is mostly science but has some art blended in as well.

Sorry,

I was just relating my experiences with my own Nitrox fills. We tend to do VERY slow fills, oxygen first if we are doing PP fills that day and not banked. In that case, what I have seen is that the air and the o2 are not thoroughly mixed and it seems that the air sits "on top" of the o2. So we roll the tanks and take a measure.

If we don't have to run out to dive right away, we mix the air in VERY quickly on top of the o2 apparently getting better mixing between the two, but this raises the tank temperature (and pressure). So we let it cool, then top off. We then analyze the mix.

In Nitrox class, when we did just slow fill's I notices as much as .5% between intended mix and final mix after a 2 hour ride in the car.

Again, I can offer no explanations. I am not a gas blender yet. Just personal experience, for whatever that's worth.

Unless someone puts something else in the tank, the mix does not change. It canNOT change. Your measurement of it can change. You are not actually measuring the % of O2 in the tank. You are measuring electrical current. The amount of O2 flowing through the sensor will determine the amount of current produced. Flow rate and temperature can change the results, but the % doesn't change.

It does not take time for the gases to mix, they are being shot into the tank, that mixes them immediately.

There is another long term storage issue that I learned about in Nitrox class. That is when rust or some other oxidation reaction consumes the oxygen in a tank. Our Nitrox class instructor told us of a particular situation where the O2 percent in a tank originally filled with air dropped to 11% after nine months of storage, due to rust formation inside the tank.

This is a long term process that takes months, not days. So obviously it isn't the explanation for the observations in this thread. Still, I found it interesting.

Edited by BubbleBoy, 15 May 2006 - 07:48 PM.

It does not take time for the gases to mix, they are being shot into the tank, that mixes them immediately.

You ever fill doubles with the isolator closed? Then open it? Does it mix immediately?

I like walters explanation.....

It does not take time for the gases to mix, they are being shot into the tank, that mixes them immediately.

You ever fill doubles with the isolator closed? Then open it? Does it mix immediately?

I also agree. If you mix Trimix and then check the mix, see what happens in twelve hours to the mix. It's not just because of tank cooling. I use a flow meter on my analyzer. So, it's not the flow rate. It takes several hours for Trimix to fully mix from partial pressure filling.