Liberty, change O2 sensors all together or not?

[rjack321]

The paper says the data was got from an actual batch of 1,000 cells. As this is a study done by rEvo they would only have used their own cells as they are the only cells that they supply, so the brand dataset would not be relevant. But can you explain what exactly do you mean by 'cell condition'?

Revo (now Mares) has never been an O2 cell manufacturer.

At the time that paper was written and posted on the internet, there were three main manufacturers: Teledyne, vandergraph, and analytical industries. Teledyne is out of the CCR business, AI has changed and seemingly improved their manufacturing significantly, and in 2022 there are multiple newer cell manufacturers making cells for CCRs (e.g. Maxtec and AST). If there ever was substantive data behind this paper (which is dubious) what makes you think it's relevant today?

As other threads here have demonstrated, storing cells in a humid environment, or hot, or frozen, or diving them daily, or letting them sit on a shelf. There are a myriad of ways that a 6 month old cell can both be pampered or abused. We don't know any of those use or storage conditions.

 

[mdwalter]

Is that me or Paul or Rjack? Extreme Value Theory is not crap it is just very infrequent so its hard to measure by anyone

I just want to add one thing about the tail being left skewed in the above chart/histogram, it comes about typically from a bad batch, so if you replace 2 cells at the same time the tail underestimates the probability both cells are bad because they come from the same batch. I.E. the probability both cells are bad is not 10% * 10% its greater.

How much greater the paper doesn't try to calculate, but its does show its a risk by separately showin a similar distribution from a bad batch, so my take from the paper is swapping all your cells out of your rebreather at the same time appears to be a poor idea.

10% * 10% is 1%. So the probability of 2 cells being bad (even from a bad batch) will always be lower, not higher, than the probability of only one of them being bad, right?

 

[mdwalter]

Until now I used the 4 sensors for about 18 month, checked them at 6, 12, 13, 14, 15, 16, 17 month up to 3.5 bar. No problems so far.

Now I was told this is dangerous because all the 4 sensors come out of one serie. It is better to change 2 sensors after 9 month and the other 2 sensors after 9 more month. So I have different series of sensors. In case of faulty serie I always have 2 good sensors.

I am not sure if this makes sense:
- are there bad series?
- will I detect a bad serie by overpressure testing to 3.5 bar?
- I have 2 young sensors, 2 older sensors, they behave differently. The unit believes the young sensors are wrong and kicks them out. Until now it is OK. Now let's assume one of the older is dammaged, now I have a serious problem. In case of 4 equal sensors I have no problem, if one is dammaged.

I know about an accident on a AP Inspiration. The diver changed one ot the 3 sensors every 6 month, so the age was:
No 1: 6 month
No 2: 12 month
No 3: 18 month
No 3 died during the dive because of its age, it gave wrong measurement.
No 2 was dammaged for any reason. Unfortunately it behaved exactly as No 3.
The unit kicked out No 1, the only working sensor. The diver died.
Not exactly the same situation as on a Liberty, it has 4 sensors. But the problem is similar.

How do you change sensors on a Liberty?
How do you change sensors on any other unit?

So, having read every single post in this thread and in several other threads (one of which spawned this one), here are my thoughts.

My current rebreather is on its first set of (4) sensors, which all appear to be from the same batch. I have been wondering about what sort of replacement protocol I should follow (e.g. all at once or some sort of staggered rotation as described by several contributors here) since the day I got my unit. I don't fear the mythical "bad batch" per se, but I do buy into the concept that these 4 cells from the same batch will, having spent their lives in pretty much the same environment, will likely reach their end of life and start failing very close to the same time. Therefore, to minimize the likelihood of multiple cells failing all at the same time, I'm thinking of one of the following methodologies for staggering the replacement of my cells:

1) I wait until about 6 or 9 months into the life of my original cells, then buy 1 cell as a "hot standby" (i.e. I have it to use immediately if one fails) spare, but do not install it right away. I then wait until about 12 months in, and if none has failed I buy 2 more and replace 2 of my 4 cells with the spare and one of the new ones and the other new one is the new spare (and maybe I keep the 2 old ones that have not failed as additional spares for another 6 months or so). Now at 18 months I buy 2 more and install the 6 month old spare and one new one, keep the other new one as my new spare, and discard all 4 original 18 month old cells. Now this puts me in a state where I have one cell that's 9 months old and has been in the RB 6 months, one cell that's 6 months old and in the RB 6 months, one cell 6 months old and just went in RB, one new cell in RB and one new cell spare.
1)a) From this point forward I can just buy 2 new cells every 6 or 9 months and rotate as described and I'll always have 4 cells from 3 different batches running and a spare that is less than 50% of it's shelf life.
1)b) From this point forward I can just replace cells as they fail, go non-linear, or age out at like 18 to 24 months with the spare I have and then buy a new spare.

2) At the 9 or 12 month mark, I buy 2 cells and proactively replace 2 of my 4 original cells, keeping the old cells that have not failed as spares. At 18 months, I buy 3 cells and replace the other 2 originals, keeping the third new one as a spare and discard all 4 originals which are now 18 months old. Now every 9 months I can buy 2 more, swap in the spare and one new one and have a new spare. This is simpler, and I will always have 4 running cells from 2 batches.

I feel like any of these scenarios spreads the risk (of multiple failed cells close together) and cost (of new cells) out across time and allows for easy instant replacement of failed cells without the waste of spares that age out and never get used.

The next question is if replacing 4 cells 2 at a time, does it make more sense to replace 1 & 3 together and 2 & 4 together or always put new cells in 1 & 2, then move the "middle aged" cells 3 & 4 and throw out the oldest cells from 3 & 4, given that 1 & 2 are shared between controller and HUD, 3 is ctrlr only, and 4 is HUD only. The first method ensures either the controller or HUD always has 2 new cells, and cells never change positions. The second ensures that both Controller and HUD always see the 2 newest cells, but requires moving some cells during each rotation.

Thoughts?

 

[Big-swede]

10% * 10% is 1%. So the probability of 2 cells being bad (even from a bad batch) will always be lower, not higher, than the probability of only one of them being bad, right?

Not always, if the cause of the problem is in the manufacturing of the cells the probability is higher (just look at those new cells that came out a coupleof years ago).