Trimix blending question

[Gareth J]

To answer the original Post.

Blending is generally taught around 'ideal gas laws', however, gases do not truly behave in this way, they behave as the 'real gas laws', dictate.
The problem, is that as the pressure increases, the volume occupied by each gas molecule is reduced. Which distorts the calculation.

Various softwares attempt to compensate for the compression.

Fairly obviously, as blending pressures increase, the distortion increases. One of the reasons that most blenders prefer to blending to a maximum of 220bar. 300bar mixing is problematic due to this distortion.

I don't mix with other blenders, but I personally find it far more convenient to apply "Kentucky windage," let the mix sit overnight or throw it in the car and drive around for 10 minutes or so so it will mix more quickly, test, and top. This gets my O2 within a percent reliably and the He within a percent or two as a rule.

Certainly, when I blend Trimix, I like to leave the cylinders between each gas addition, to allow the temperature to return to ambient, and thus have an accurate pressure. Doing this ensures an accurate mix. It is less of a problem now than when I started with affordable Helium Analysers. (Although I so seldom mix my own trimix thesis days that I don't own a Helium Analyser.)

Your comment about driving the cylinder around is true despite browning motion saying the gas will have mixed.
I was taught to add gas slowly, reducing heat, improving accuracy. Despite what logic and Browning Motion would indicate, I have experienced 'layering' . Where the gas doesn't appear to mix.
Either leaving the cylinder for some time, or rolling the cylinder, seems to resolve the issue.

 

[Kevrumbo]

To account for compressibility compensation at high pressure in home blending mixes, try a simple practical "fudge" factor like, for example, starting with 10% less the ideal gas calculation for the amount of pure O2 a particular blend calls for (refer bottom p.35 of Vance Harlow's Oxygen Hacker's Companion as it applies to adding O2 first in mixing a Trimix Blend and topping-off with Air) and see what you get when you top off with clean compatible Air through your local dive shop's boosted fill station. Adjust fudge factor as needed via trial & error for your particular cylinder(s) as you gain blending experience over subsequent fills.

So for 3442psi/230bar in a 12L/HP100 cylinder for example, the partial pressure ideal gas blending recipe for Nitrox32 requires 13.92 psi or 13.92 bar of Oxygen for every 100 psi or 100 bar of Nitrox32: subtract 10% instead to yield 12.52 psi or 12.52 bar of O2 for every 100 psi or 100 bar.

Hence for every 100 psi X 34.42 = 3442psi full fill, you need 12.52 psi X 34.42 = 430psi of O2, with the remainder a top off of hyperfiltered oxygen compatible clean air; or 100 bar X 2.30 = 230bar full fill, you need 12.52 bar X 2.30 = 29 bar of O2, with the remainder a top off of hyperfiltered oxygen compatible clean air.

My LDS here is Pacific Wilderness, and they fill & maintain all the working HP100 single tanks for the Port Police, County Sheriff and Fire Department dive teams of the Port of Los Angeles and Long Beach. They've been filling with 420psi of Oxygen and then topping off to get a final full fill analyzed blend of 31.8% for Nitrox32.