How accurate is an analyser at high % of oxygen?

[Johanan]

I purchased a new deco cylinder with oxygen, analysed it and the device indicated 87%. The vendor swore, their oxygen was 99,9%. In fact, a reputable firm over here. My technician said, he did not trust the analyser, as the readings may be incorrect at this high content of oxygen.

What should I enter in my computer, what do you think? if I trust the analyser and start to breathe this gas at a greater depth, I may get hit. If I trust the vendor but the mix really is 87%, I may do not enough deco. The safer option would be to use the 87% value and switch to the deco bottle at 6m, yet, somewhat annoying. In your experience, how accurate are the analysers at the high end of the scale?

 

[rjack321]

The cell in the analyzer is probably old. They last 2 to 3 years max.
Try a different analyzer

Don't breathe the gas unless you are confident what it is. O2 from the gas supplier is not 87% and if it was what's the other 13%??!!??

 

[kelemvor]

I have to agree with @rjack321 I would not breathe that gas until I got a confirmation of the result one way or another. You could get bent, or worse, have a seizure.

Analyzers do need to be calibrated regularly, and they need to be calibrated in the range you're testing for. i.e. calibrating with 21% is NFG for testing 100%. Some analyzers like my @DiveNav cootoo support "multi point calibration". I can calibrate using a known 21% source and then a known 100% source and will get reasonably accurate results near either of those.

Also, as previously stated - the sensors have a limited lifespan. They will go bad even if they're sitting on a shelf unused. Some sensors are rated for 1yr lifespan, some for several years. Due to the limited shelf life, you can't stockpile the things.

 

[tbone1004]

sensors are quite accurate, but they are only considered accurate when close to their calibration gas. I.e. if you calibrate in air and try to analyze O2, then the analysis is not considered accurate due to the lack of linearity in the cells. If you calibrate with pure O2 and try to analyze air, then it's going to be just as wrong. Cells can be off by 5-10% easily, but can also be current limited and not be able to calibrate properly in pure O2.

If you are trying to use your analyzer to analyze a tank with O2 in it, you need to calibrate that analyzer against one of the O2 bottles that they are getting from the gass supplier which will be 99.9%+.

 

[tridacna]

We have 4 different reference gasses that we use to calibrate an O2 sensor before measurement. The variance when using one calibrated for air to measure 100% can be significant indeed.

 

[rjack321]

If you are trying to use your analyzer to analyze a tank with O2 in it, you need to calibrate that analyzer against one of the O2 bottles that they are getting from the gass supplier which will be 99.9%+.

I'm going to bet that this sensor will never read 100% because its old and worn out.

Although there is a small chance that there was another gas left in the booster's supply lines etc (like pure helium) that ended up being 13% of the fill in his tank. Very very small chance of that but more than zero.

 

[broncobowsher]

Another tester please.

Talk to anybody with a rebreather about a cell that gets old and is current limited. They read fine in the bottom of the range, but on the top end they choke up and flat line. Bad thing in a rebreather as they say you have too little O2 so it adds more and more and more and you are screwed. That's why they typically run 3 cells, hopefully from different batches. When one doesn't read right compared to the other two, you quit diving and figure what is wrong.

 

[BRT]

My testers, calibrated on air, always read in the 96-98% range for O2.

 

[Johanan]

Thank you for your comments and advice! I'll try to use another tester and calibrate it with O2. It is better to know exactly, what is in the cylinder.

 

[happy-diver]

I wonder what difference purging a new deco cylinder with oxygen first makes