Sensors: which gas shall I have in the loop when not in use?

[Agro]

Understanding_oxygen_sensors.pdf

Mate google the sentence and see what you can see

https://www.google.com/search?q=Understanding_oxygen_sensors.pdf&rlz=1C1ONGR_en-GBAU943AU943&sourceid=chrome&ie=UTF-8

What do you do underwater, write a letter

Thank you. This paper is from 2010 and I do not draw the same conclusion.

 

[Wibble]

Thank you. This paper is from 2010 and I do not draw the same conclusion.

Most people diving rebreathers do not change all their cells at one time and from the same batch of sensors.

In the past there’s been many examples of bad batches. People learn from that.

 

[Agro]

This is a conclusion you may take.

My conclusion is another one which I took from an Inspiration accident: 3 cells, I don't remember the ages but let's say 0/6/12 month. The 12 month old cell gave wrong reading because of age. The 6 month old cell had a failure and unfortunately gave the same wrong reading as cell 12. The rebreather kicked out the new sensor (the only working sensor), worked on sensor 6 and 12, the diver died.

I prefere having all the cells new, I ckeck them to 3.5 bar pO2 and exchange them after 12 month. Until now I changed after 18 month or in case of faillure, in future I will change after 12 month.

We can talk about probability. In which case can happen what ever?
We can compare Liberty's 4 cells to Inspiration's 3 cells.
When I have a new cell and some older cells I always have the risk of this Inspiration accident, no matter how clever rebreather is dealing with that, no matter how many sensors I have.

 

[Wibble]

This is a conclusion you may take.

My conclusion is another one which I took from an Inspiration accident: 3 cells, I don't remember the ages but let's say 0/6/12 month. The 12 month old cell gave wrong reading because of age. The 6 month old cell had a failure and unfortunately gave the same wrong reading as cell 12. The rebreather kicked out the new sensor (the only working sensor), worked on sensor 6 and 12, the diver died.

I prefere having all the cells new, I ckeck them to 3.5 bar pO2 and exchange them after 12 month. Until now I changed after 18 month or in case of faillure, in future I will change after 12 month.

We can talk about probability. In which case can happen what ever?
We can compare Liberty's 4 cells to Inspiration's 3 cells.
When I have a new cell and some older cells I always have the risk of this Inspiration accident, no matter how clever rebreather is dealing with that, no matter how many sensors I have.

One of the problems of only having 3 cells - what happens when two fail. This happened to me on my MOD1 but having 5 cells on a Revo meant even an inexperienced CCR diver could clearly see the failure.

 

[tbone1004]

Thank you. This paper is from 2010 and I do not draw the same conclusion.

We are using the exact same technology now as we were when this paper was written. If anything had changed between then and now that's one thing, but you can't use the age of the paper against it in this case.

This is a conclusion you may take.

My conclusion is another one which I took from an Inspiration accident: 3 cells, I don't remember the ages but let's say 0/6/12 month. The 12 month old cell gave wrong reading because of age. The 6 month old cell had a failure and unfortunately gave the same wrong reading as cell 12. The rebreather kicked out the new sensor (the only working sensor), worked on sensor 6 and 12, the diver died.

I prefere having all the cells new, I ckeck them to 3.5 bar pO2 and exchange them after 12 month. Until now I changed after 18 month or in case of faillure, in future I will change after 12 month.

We can talk about probability. In which case can happen what ever?
We can compare Liberty's 4 cells to Inspiration's 3 cells.
When I have a new cell and some older cells I always have the risk of this Inspiration accident, no matter how clever rebreather is dealing with that, no matter how many sensors I have.

Your argument has no bearing on changing them all at once though, but more importantly you have the breather on the market where you can choose to manually disable a cell so if you saw that both of those cells were bad you could have disabled them and essentially overridden the voting logic. It has nothing to do with 4 vs 3, everything to do with your ability to manually disable them after you perform a dil flush and confirm which cells are behaving.
Incidentally I think that argues for using pretty rich dil mixes that are in line with what you are using for a setpoint. I.e. EAN32 at 30m/100ft instead of air where your dil flush validates function at 1.3 instead of 0.8.

Step back and think about your Inspiration though. Risk analysis is all about risk severity and risk probability. You are not decreasing the probability of failure by having all cells be the exact same age from the exact same lot when they are the same age. Listen to all of the points that we made in your other thread *which I admittedly forgot about, though I'm sure there are many others on this board* but I am genuinely not aware of anyone who advocates for changing all cells at the same time.

 

[Agro]

We are using the exact same technology now as we were when this paper was written. If anything had changed between then and now that's one thing, but you can't use the age of the paper against it in this case.

This paper sais, as far as I understand, cells are not linear anymore at pO2>2 bar, there is a graph showing this. My actuall cells are perfectly linear up to 3.5 bar, in case they are in good shape. Did cells change in this behaviour? Or did I missunderstand it?

 

[Agro]

We are using the exact same technology now as we were when this paper was written. If anything had changed between then and now that's one thing, but you can't use the age of the paper against it in this case.



Your argument has no bearing on changing them all at once though, but more importantly you have the breather on the market where you can choose to manually disable a cell so if you saw that both of those cells were bad you could have disabled them and essentially overridden the voting logic. It has nothing to do with 4 vs 3, everything to do with your ability to manually disable them after you perform a dil flush and confirm which cells are behaving.
Incidentally I think that argues for using pretty rich dil mixes that are in line with what you are using for a setpoint. I.e. EAN32 at 30m/100ft instead of air where your dil flush validates function at 1.3 instead of 0.8.

Step back and think about your Inspiration though. Risk analysis is all about risk severity and risk probability. You are not decreasing the probability of failure by having all cells be the exact same age from the exact same lot when they are the same age. Listen to all of the points that we made in your other thread *which I admittedly forgot about, though I'm sure there are many others on this board* but I am genuinely not aware of anyone who advocates for changing all cells at the same time.

Still I draw other conclusions. Having new and old cells can mean they behave differently, this can cause problems. Solving this problem by manually disabling is already one step further. I prefere not even having this problem.

 

[Wibble]

Still I draw other conclusions. Having new and old cells can mean they behave differently, this can cause problems. Solving this problem by manually disabling is already one step further. I prefere not even having this problem.

1. Cells fail in all sorts of ways
2. Cells have a finite lifespan
3. It has been known that cells from the same batch had manufacturing defects
4. Cells from the same batch will tend to fail together.

Therefore it's accepted good practice to spread cells across batches and avoid where possible changing them all at the same time. Agreed this isn't always possible.

Some people switch cells on a 4 monthly or 6 monthly cycle (e.g. Revo)

 

[tbone1004]

Still I draw other conclusions. Having new and old cells can mean they behave differently, this can cause problems. Solving this problem by manually disabling is already one step further. I prefere not even having this problem.

having them the same age does not preclude them from behaving differently

 

[Agro]

having them the same age does not preclude them from behaving differently

Right. And if one is behaving differently then it probably is because of a faulty cell, not because of the age.