The latest Eagle's Nest mishap involved a doubles fill with the isolator closed. It raised some questions as to what would happen if a mix were attempted with a closed isolator, and that isolator were later opened.
What would the analysis of such a set of doubles look like?
Would the two tanks mix through the now open manifold?
What are the potential problems?
So, I did a little experiment.
I drained a set of (nitrox) doubles and then, with the isolator closed I put 500 psi oxygen in one tank and 500 psi air in the other.
Then I analysed each tank.
Right post - 93.3%
Left post - 21.8%
I then opened the isolator and let it sit for two hours.
with the isolator still open,
Right post analysis was wildly erratic, values ranged from 40% to 70%. I let this go for five minutes and the swings did not steady out at all.
Left post initially swung from 56% to 60% a couple of times, then settled down at 56.8% steady.
Closed isolator
Right post - 93.3%
Left post - 22.0%
Opened isolator...
After 10 hours with isolator open...
Right post - after an initial spike to 62% and dip to 54%, settled at 59.0%
Left post - normal looking analysis at 58.8%
Closed isolator...
Right post - 93.1%
Left post - 22.4%
so...
After 10 hours with an open isolator, no appreciable mixing between the bottles has occured.
Considering that, I am at a loss to explain why the open-isolator analyses are less erratic than they were at the two hour mark.
Isolator reopened...
After 26½ hours with the isolator open...
Left post - 57.7% after a little bit of erratic movement
Right post - 60.0% after quite a bit of erratic movement
Isolator closed:
Left post - 23.1%
Right post - 93.1%
Reopened isolator and tried to reanalyse - very erratic both posts...
After 44 hours with the isolator open...
At this point in the experiment it was brought to my attention that I may not be using a procedure that would guarantee an accurate analysis of each bottle when I had the isolator closed, because there may be residual gas in the crossbar, so...
Opened valve on right post for a few seconds. This filled the crossbar with the contents of the left tank.
Closed Isolator, analysed left post... 22.7%
Opened Isolator.
Opened valve on left post for a few seconds. This filled the crossbar with the contents of the right tank.
Close isolator, analysed right tank... 93.3%
Therefore... with the isolator open for 44 hours there has been zero measurable mixing between the contents of the two tanks!
Now I didn't think there'd be much, but I did think there'd be at least enough to measure after nearly two days.
Interesting...
So, I'm ready to draw the following conclusions
If you were to fill doubles with different gasses:
(1) It would take weeks (if not years) for them to thoroughly mix through the manifold
(2) The situation would probably be apparent during analysis by the erratic nature of the analysis, but it is possible to get a normal looking analysis.
(3) Whether there would be adequate mixing during breathing such a set of bottles to avoid dangerously high or low O2 levels even with the isolator open is still uncertain.
(4) If you did dive such a set thinking the gasses were mixed and you had to isolate you could be in deeeeep kimchee.
Rick
What would the analysis of such a set of doubles look like?
Would the two tanks mix through the now open manifold?
What are the potential problems?
So, I did a little experiment.
I drained a set of (nitrox) doubles and then, with the isolator closed I put 500 psi oxygen in one tank and 500 psi air in the other.
Then I analysed each tank.
Right post - 93.3%
Left post - 21.8%
I then opened the isolator and let it sit for two hours.
with the isolator still open,
Right post analysis was wildly erratic, values ranged from 40% to 70%. I let this go for five minutes and the swings did not steady out at all.
Left post initially swung from 56% to 60% a couple of times, then settled down at 56.8% steady.
Closed isolator
Right post - 93.3%
Left post - 22.0%
Opened isolator...
After 10 hours with isolator open...
Right post - after an initial spike to 62% and dip to 54%, settled at 59.0%
Left post - normal looking analysis at 58.8%
Closed isolator...
Right post - 93.1%
Left post - 22.4%
so...
After 10 hours with an open isolator, no appreciable mixing between the bottles has occured.
Considering that, I am at a loss to explain why the open-isolator analyses are less erratic than they were at the two hour mark.
Isolator reopened...
After 26½ hours with the isolator open...
Left post - 57.7% after a little bit of erratic movement
Right post - 60.0% after quite a bit of erratic movement
Isolator closed:
Left post - 23.1%
Right post - 93.1%
Reopened isolator and tried to reanalyse - very erratic both posts...
After 44 hours with the isolator open...
At this point in the experiment it was brought to my attention that I may not be using a procedure that would guarantee an accurate analysis of each bottle when I had the isolator closed, because there may be residual gas in the crossbar, so...
Opened valve on right post for a few seconds. This filled the crossbar with the contents of the left tank.
Closed Isolator, analysed left post... 22.7%
Opened Isolator.
Opened valve on left post for a few seconds. This filled the crossbar with the contents of the right tank.
Close isolator, analysed right tank... 93.3%
Therefore... with the isolator open for 44 hours there has been zero measurable mixing between the contents of the two tanks!
Now I didn't think there'd be much, but I did think there'd be at least enough to measure after nearly two days.
Interesting...
So, I'm ready to draw the following conclusions
If you were to fill doubles with different gasses:
(1) It would take weeks (if not years) for them to thoroughly mix through the manifold
(2) The situation would probably be apparent during analysis by the erratic nature of the analysis, but it is possible to get a normal looking analysis.
(3) Whether there would be adequate mixing during breathing such a set of bottles to avoid dangerously high or low O2 levels even with the isolator open is still uncertain.
(4) If you did dive such a set thinking the gasses were mixed and you had to isolate you could be in deeeeep kimchee.
Rick
I'll hit compulsive when I wake up at night and run down to the garage to check my valves..... 
