Average Gas Consumption

[scubadada]

I haven't participated in the poll yet. I was wondering if for the calculation we (incorrectly) assume that air/Nitrox compresses linearly or take into account the actual variation in compression with pressure (as Subsurface does).

For example, on a recent dive my RMV using linear compression was 11.7L/min but 10.3L/min in Subsurface. This was using a 300bar tank where the differences between linear and actual compression are a lot higher. The day before I used 2 x 230bar tanks and had RMV of 12.0 and 11.3 respectively.

Hi @ATJ

Thanks for your post. I would imagine that everyone has posted their average RMV using whatever method they generally use for making the calculation. I was unaware that Subsurface corrects for the nonlinear compression of gas. Do you know of other programs that make this correction in the calculation?

Your examples are interesting. With the 300 bar cylinder your RMVs were 10.3 L/min/0.36 cu ft/min and 11.7/0.41. With the 230 bar cylinder, the RMVs were 11.3/0.40 and 12.0/0.42. The lower values were calculated by Subsurface.

My 10 year old Oceanic VT3/OceanLog software and my Shearwater Teric give me the same RMV, I would imagine with linear compression. I most commonly dive AL80s with a working pressure of 3000 psi/207 bar and occasionally dive HP100s with a working pressure of 3442 psi/237 bar. There is an interesting discussion of this topic on the Dive Gear Express website Calculating SCUBA Cylinder Capacity | Dive Gear Express®

What RMV value do go by, Subsurface?

 

[Storker]

I haven't participated in the poll yet. I was wondering if for the calculation we (incorrectly) assume that air/Nitrox compresses linearly or take into account the actual variation in compression with pressure (as Subsurface does).

For example, on a recent dive my RMV using linear compression was 11.7L/min but 10.3L/min in Subsurface. This was using a 300bar tank where the differences between linear and actual compression are a lot higher. The day before I used 2 x 230bar tanks and had RMV of 12.0 and 11.3 respectively.

Do you know how Subsurface accounts for compressibility? vdW equation?

I've played around with different RMV calculations, and it seems as if Suunto's Dive Manager uses the vdW equation with standard, published constants. The results still show some anomalies, although a lot fewer than what simple ideal gas calculations - like those you find in Diving Log - do (my RMV is 50% higher for a very short dive from a 300 bar tank than if I breathe it down? Come ON!). In my own spreadsheet I use the vdW equation, but I've adjusted the constants to fit published compressibility data. Lo and behold, no anomalies.

 

[dirkhh]

Try this article from our resident scientist:
When real gas corrections matter – The Theoretical Diver

 

[jborg]

Subsurface uses this gas compressibility model: subsurface/subsurface

I'm not sure what the constants actually mean, though.

 

[Storker]

I'm not sure what the constants actually mean, though

They mean nada. They're just chosen to make the predictions fit the observations.

The ideal gas law (PV = nRT, of which the other gas laws, e.g. Boyle's law, are subsets) is only valid at very low pressures, because it assumes that the gas molecules don't take up any place and that there are no attraction forces between them. As pressure increases, we see the effect of interactions between gas molecules (excluded volume plus attraction forces), so the gas becomes non-ideal. That means that we can't predict its behavior with the ideal gas law. The most used non-ideal gas law is van der Waals' equation. It has some constants which have to be fitted to what we actually see, to properly calculate how a non-ideal gas behaves.

Those constants don't mean anything, they're no more than fudge factors that we throw in to be able to make better predictions.

 

[atdotde]

We had experimented with van der Waals, but the results were not really good compared to tabulated data (on the range relevant to diving cylinders which is not really close to the tried-critical point. Do we decided to do a least square fit to a quintic in the relevant pressure range. That’s described in the blog post mentioned by dirkhh.

 

[Storker]

We had experimented with van der Waals, but the results were not really good compared to tabulated data

I'm not very surprised. If you check published values for N2-O2 or N2-He mix compressibilities, they don't fit with tabulated vdW constants.

But it's no big deal to adjust the vdW constants to fit published data.

 

[Miyaru]

Example, 12 liter doubles:

(yes, it's metric)

 

[ATJ]

What RMV value do go by, Subsurface?

When I was almost exclusively diving with 230bar tanks, and my starting pressures and ending pressures were in the same ballpark (starting on 228-233 bar and ending 20-40 bar), using linear compression was good enough for comparing my RMV from one dive to the next. It was easy to see when my RMV was higher due to being cold, more surge, etc..

Now that I am diving with 300 bar tanks more often but not always, using linear compression makes the RMV for dives with the 300 bar tanks look a lot worse than the 230 bar tanks. I'm now leaning more on the Subsurface compression.

If I look at my average RMV for the year so far, using Subsurface I'm 10.9L/min as opposed to 12.0L/min using linear compression. Subsurface would have me in the second option in the poll whereas the straight volume x pressure would have me in the third.

 

[Storker]

Calculating and then doing partial pressure blending of trimix is a good exercise in humility. Particularly if the end pressure is above 232 bar. He has a distinctly non-ideal behavior, especially beyond 232 bar. Even if you add it at low pressure. And even disregarding adiabatic heating effects.