Creating the Proper Mix Using Partial Pressure Blending

[Oleg G.]

True, one of the biggest sources of error is warming. The calculations with different gas models (Real and Ideal Gas) show how far the deviations are apart. As a tip, with this program, the different calculations can be easily compared: Gas blender toolkit

 

[CptTightPants21]

When I PP fill, which I will continue to do until I get my new membrane up and running, I measure what is still in the cylinder. I use the Trimix tools app to calculate my new mix and pressure. I tell the app that I’m filling 3442 cylinders to 3700, and actually fill them to 4000 smoking hot. I tell it I’m filling 3000 cylinders to 3300 and then fill them to 3700 smoking hot. I end up within .2 to .4% every time at exactly the correct fill pressure (3500 or 3100).

Do you do the helium slowly and then jack up the O2 and air or do you do everything pretty much the same speed?

 

[Manatee Diver]

Love me some Trimix tools.

I wish they made it for Android, I've seen it used at PDC and I like it a lot.

@morejanus

 

[colint]

The calculations are easy. The problem is the gas heats up when you fill and the pressure drops when it cools. I have found two ways:
Allow the gas to cool and top up to the correct pressure
Add a bit more gas to allow for cooling.
The second method requires some guesswork based on how much gas is added and how quickly. It comes with experience of the equipment and practice but can be quite accurate with some luck.

 

[rjack321]

Do you do the helium slowly and then jack up the O2 and air or do you do everything pretty much the same speed?

Personally I don't do that many 3 gas PP fills at all. Most of mine are O2+air for EAN50 or helium+32% for trimix.

I try to do O2 slow to avoid a fire
helium, air, or 32% additions I just squirt in there however fast they go - BUT I just try and get close to the needed final pressure. I let them cool at least overnight before adjusting to get final pressure. Net this means 50% fill takes at least overnight - because I can get close enough on the air top. A trimix fill usually takes 2 nights.
50% fill
Day 1 - Add o2
Day 2 - check O2 and top off, then add air but overfilling - for a 3,000psi al80 I usually pump the air fraction to 3300 and it cools to 3000.

trimix fill
Day 1 - add helium
Day 2 - check helium and top off, then add 32%
Day 3 - check pressure and top off 32% to final

 

[Billy Northrup]

When I PP fill, which I will continue to do until I get my new membrane up and running, I measure what is still in the cylinder. I use the Trimix tools app to calculate my new mix and pressure. I tell the app that I’m filling 3442 cylinders to 3700, and actually fill them to 4000 smoking hot. I tell it I’m filling 3000 cylinders to 3300 and then fill them to 3700 smoking hot. I end up within .2 to .4% every time at exactly the correct fill pressure (3500 or 3100).

Yup.

cant make any money fiddling with a tank more than once, usually done in batches of 6 to 8. Every time I touch a tank I lose money.

Trimix I do in a tank with a needle valve, but you need to pick it up tomorrow because I need some down time to do it.

 

[Wookie]

Do you do the helium slowly and then jack up the O2 and air or do you do everything pretty much the same speed?

The only gas I don't move fast is oxygen, and then it goes the speed of the booster.

 

[Storker]

Do you do the helium slowly and then jack up the O2 and air or do you do everything pretty much the same speed?

If you take a blender class, one of the things they'll stress is that 100% O2 should never be rushed nor jacked up. With pure O2, you really want to take it slow.

 

[Dan_P]

Helium first.
Whichever fraction I want in the mix, I whack in the empty tank.
Nx32 top-up.
Done.

If I don't have Nx32, I mix that directly on top of the He.
He first.
14% O2.
Air top.
Done.

Ultra easy.

If you´re filling the same gas on a partially filled tank, you're just filling to a lower pressure (the missing pressure in the tank/s).
But same approac
Easy.

When I'm mixing down to a thicker mix at the end of a sortie, I usually just add Nx32 and disregard that I'm wasting some of the remaining He in my reserve from the previous dive.
Easy
The only time there's a bit of actual legwork involved, is when you want to go from one gas to a leaner one, without draining the tank. But even that isn't THAT hard at all. I like doing the calculations with pen and paper, but the mobile apps work a charm. Sometimes the target mix isn't possible but that's the name of the game.

Other than that, I just keep the speed down especially He and O2 (Nx32 and air fills can be done and re-done a hundred times till the tank is cold, but I can't yank helium back out of the mix).

Oh, and I'd advice to try not to blow up if boosting oxygen
Go slow, keep things clean and make sure the whip has the correct diametre spec'ed for the booster (once worked a booster where the whip had been changed to a slimmer diametre - boom and fire ensued).

 

[Chris H]

I've had good results using Baltic Blender which corrects for temperature during the fill. If you are guessing how much to overshoot to compensate for the warmer tank and contents, you are going to have issues. I use a IR thermometer and adjust the temp in the app as I fill.

I've seen some really strange results immediately after boosting helium to top a mix. Laying the cylinders on their side seems to help with mixing, if it is a single cylinder, you could roll it. Specialty Gas manufacturers often roll their cylinders to promote mixing. I worked in a specialty gas lab for 5 years, and the Gas Chromatagraph results pre and post roll were sometimes very surprising. Obviously it is not easy to roll a set of doubles.