Analyzing multiple nitrox tanks

[BRT]

Interesting question, which seems pretty simple, and interesting responses. I'll share a recent story that will provide a different perspective.

I was in Cozumel a couple months ago. My party of four had asked for 8 nitrox tanks (2 each) for our dives on that particular day. I have my own O2 analyzer and while I usually don't bother bringing it with me when I travel, this time I did.

When we arrived at the dive shop, the tanks were there and ready to be loaded on the boat. The shop had already analyzed them and had labeled them all 32%. I pulled out my analyzer and started checking them for myself.

32... 32... 32... 21

Yup. One of the 8 tanks was air. How could this have possibly happened!?!? That's always the question immediately following an accident. Fortunately it didn't get to that point.

The shop guy was indignant and insisted something was wrong with my analyzer. I was equally indignant and told him something was wrong with him and his analyzer. And so began the pissing match.

He pulled his analyzer out (yes, his actual analyzer, not his...) and re-analyzed the tanks:

32... 32... 32... 32

He was ready to declare victory until I told him "recalibrate your analyzer and then measure that tank again."

He did. And his smug look disappeared. It read 21.

So what happened? I didn't tell him this, but just the week before I was in a shop at home and the same thing had happened to me: analyzing several nitrox tanks in a row, the readings were similar... just as I expected them to be. But some of my readings were slightly different than what the shop had measured. I pointed this out, and the shop guy told me that when analyzing multiple tanks in quick succession, residual gas from the previous tank (i.e. gas that gets trapped in the analyzer) can cause an inaccurate reading. He had me re-calibrate the analyzer between tanks (i.e. clear the residual gas from the previous tank) and remeasure. My readings then matched his.

The potential for mistake compounded by expectations: in the case of my Cozumel tanks, the shop guy expected every tank to be 32. So when he quickly moved his analyzer from one tank to the next, and he got 32 for each, he simply confirmed his expectations.

Two lessons here:

1. Be mindful of what your analyzer is really reading, i.e. the gas that is in contact with the sensor, and where it came from. If you have any doubt, re-calibrate.

2. Analyze your own gas. It's a standard rule that's hammered on in every nitrox class (or at least should be), but some places (like the shop I went to in Coz) just assume they are always right and encourage you to trust them (because, after all, you're on vacation!) But if you get bent from diving on 21 when you thought you had 32... the most you'll get from them is an apology.

Recalibrating your analyzer between tanks of nitrox will cause your reading to be lower than actual. The analyzer takes awhile to get the higher O2 air flushed out and if you calibrate it at that point it accepts that high O2 air as 21% lowering subsequent readings.

 

[mi000ke]

Well this gets interesting. So per Yle, you want to recalibrate between tanks to override the gas from previous reading. But according to BRT if you recalibrate you might calibrate to the previous gas rather than air. Which kind of leads back to my original post. So what would be the best way to flush out the previous gas? I assume a stream of air under pressure from a tank. But.... (and now we are back to the beginning).

 

[PfcAJ]

So what would be the best way to flush out the previous gas? I assume a stream of air under pressure from a tank.

Patience is the best way. If you're super concerned suck the gas out till it reads 21%.

 

[Joneill]

Well this gets interesting. So per Yle, you want to recalibrate between tanks to override the gas from previous reading. But according to BRT if you recalibrate you might calibrate to the previous gas rather than air. Which kind of leads back to my original post. So what would be the best way to flush out the previous gas? I assume a stream of air under pressure from a tank. But.... (and now we are back to the beginning).

If you are flowing a stream of gas through the sensor (as with the BCD adapter on my Cootwo), then I do not believe there is a problem measuring successive tanks as long as you ensure the reading stabilizes bevore you stop the analysis.

However, if you did want to recalibrate with atmospheric O2 after analyzing higher O2 gas sources, then you would want to let the unit stabilize back to atmospheric O2 levels first (takes up to 5 minutes, I believe).

 

[BRT]

Here is an old thread where an analyzer manufacturer agreed with me.

Calibrating Analox O2EII - How often do you need to calibrate the analyzer?

 

[oldschoolto]

And don't forget to mark your tank.. You guy worry to much about this stuff...

Jim...

 

[NAM001]

I cal at the start and then at the end I verify the cal.
I also have a O2 sensor that is open to the atmosphere and not embedded in the device so it is exposed to atmosphere then ever the cover is not on it.. I cal and sample a 32 tank remove the sensor and it immediately drops back towards 21, once it is in the mid 20s I samlple the second tank ect. I never recall.

Now I use an OXYCHEQ elcheapo. (below) the sensor in not part of the readout device. there is no cavity of gas the sensor sits in. The face of the sensor is flushed nearly immediately when you remove it from the tank valve.

Here you see the actual white sensor in a blue T., I don't use the T so the sensing element is instantly exposed to the 20.9 air when removed from the valve output. I hold the sensor face in the tanks expelled gas slow.

Now look at this the sensor is embedded in the readout device. This takes more time to self flush the old gas out. Purging normally happens with the new gas being applied and it takes a little bit for the reading to become stable as the old gas is replaced by new gas. With these you can cal to 21 take a reading remove the device and it drops to say 24. If you recal the device again you are making 24% give a readout of 21. It will go back to 21 if you force surrounding aire through it. That is tough to do so you use a tank of known air to do that..

The sensor as it functions will kill the O2 that remains in it and drop the cc or so of gas to nearly 0%. Again that is why you use a test air tank to insure you have flushed the dead gas away from the sensor prior to calibrating it.

So the advice as to how to do it depends on the device you are using. You also do not have to use air to cal it. as long as you know what the O2 content of the sample gas is you can use it and you adjust the cal to the value of the known test gas. I have used a cal tank of known 28% and when I cal I use the test gas and adjust the cal knob to read 28.

One last thing is you shouldn't use the same meter to verify your tank that the shop uses to check it. use your own meter to do that. if you have a tank of 28 and it reads 32 on the shop meter,,,, what good does it do to use the same faulty checker to verify the contents. Many shops have multiple testers. one at the fill sight and one at the front counter. That is a safe process. still check it again with your meter at the dive site just as you put on your regs.

 

[tbone1004]

Well this gets interesting. So per Yle, you want to recalibrate between tanks to override the gas from previous reading. But according to BRT if you recalibrate you might calibrate to the previous gas rather than air. Which kind of leads back to my original post. So what would be the best way to flush out the previous gas? I assume a stream of air under pressure from a tank. But.... (and now we are back to the beginning).

you do NOT recalibrate between tanks, certainly not based on what they were saying. You calibrate once, and you wait until the reading stabilizes, that may very well take 15+ seconds depending on the o2 sensor, but you have to be patient.

Recalibrating between tanks, based on what @yle said is 100% guaranteed to cause a serious issue at some point down the road. If you think about it, you are calibrating the sensor to "air" based on what is inside the housing, and based on what you said, that isn't actually air. So now you are telling the sensor that "air" is whatever is left inside the housing and every time you do that, the readings are going to be wrong. Real world issue? You'll mess up a deco calculation or MoD at some point and someone will either be bent, or quite possibly go into oxtox. That is a very dangerous statement and that holds no scientific validity, do NOT recalibrate unless you KNOW that the only gas inside that housing is air, or some sort of calibration gas

 

[mi000ke]

Here is an old thread where an analyzer manufacturer agreed with me.

Calibrating Analox O2EII - How often do you need to calibrate the analyzer?

Cool

you do NOT recalibrate between tanks

Settled then. No recalibration.

 

[yle]

you do NOT recalibrate between tanks, certainly not based on what they were saying. You calibrate once, and you wait until the reading stabilizes, that may very well take 15+ seconds depending on the o2 sensor, but you have to be patient.

Recalibrating between tanks, based on what @yle said is 100% guaranteed to cause a serious issue at some point down the road. If you think about it, you are calibrating the sensor to "air" based on what is inside the housing, and based on what you said, that isn't actually air. So now you are telling the sensor that "air" is whatever is left inside the housing and every time you do that, the readings are going to be wrong. Real world issue? You'll mess up a deco calculation or MoD at some point and someone will either be bent, or quite possibly go into oxtox. That is a very dangerous statement and that holds no scientific validity, do NOT recalibrate unless you KNOW that the only gas inside that housing is air, or some sort of calibration gas

I think you totally misunderstood what I said. When I suggested recalibrating between tanks, I didn't mean "turn the little dial until it reads 21%." I realize that's what some people think "calibrating" is; unfortunately, I come from a background of lab work and I know what it means to calibrate instruments, to verify that they are reading a known source correctly. Some people don't know this... and to those people I apologize if my previous post caused confusion.

Ironically, you said the same thing I did: "do NOT recalibrate unless you KNOW that the only gas inside that housing is air, or some sort of calibration gas." Exactly. Because that is what CALIBRATION IS.

I explained what happened in my story, and I'll stand by my statements that (a) preconceived expectations of what is in a tank can cause problems (if a faulty reading appears to confirm those expectations... this is what happened to the shop guy in Coz) and (b) be mindful of what your analyzer is really reading (i.e. what gas is actually in contact with the sensor when the reading is displayed.)