Lost Oxygen in blending? (Nitrox stick)

[Fl_Key_Guy]

Everyone has been great with offering suggestions on why my output nitrox is not matching the input!

I have been calibrating by shutting off the oxygen flow and calibrating to 20.9 on the stick with the compressor running. I reduce initial humidity by positioning the air intake at the outlet of a window air conditioner. For sure it's wet ambient air in the Keys. For sure it is dry air in the tank. As far as humidity goes, I have not noticed much difference between readings if calibrating to 20.9 in the open air and then testing a known straight air fill- both are 20.9, maybe a tad higher off the air fill but never a whole percent. ( I was actually checking for LOW oxygen on tanks not used in a while.)

I'm filling banks as well as tanks, and filling 4 tanks at a time when on direct tank filling, so the little bit of unenriched air in the compressor and hose would not affect the percentage noticeably. The compressor runs for hours.

There is very minimal restriction in the stick- much less than any scuba regulator that isn't free flowing. I have a manometer coming to confirm vacuum level under running conditions.

Although the hose between the stick and the intake filter seems secure enough at both ends, I'm going to add hose clamps in case I am getting some leakage due to vibration.
If nobody else has had this problem, it MUST be air sneaking in SOMEWHERE. I'll also check the air tightness of the intake filter housing.
But not tomorrow- it's DIVE DAY! Gotta breathe up some of that weak nitrox....

 

[Web Monkey]

If nobody else has had this problem, it MUST be air sneaking in SOMEWHERE. I'll also check the air tightness of the intake filter housing.
But not tomorrow- it's DIVE DAY! Gotta breathe up some of that weak nitrox....

I'm absolutely not a compressor expert, but it's my understanding that under certain conditions, the compressor oil can burn, which would account for lower O2 and higher CO.

I have no idea if this is what you have, but you might want to check for Carbon Monoxide.

Terry

 

[Fl_Key_Guy]

That maybe my oil or activated charcoal was consuming oxygen were my big concerns along with the byproducts thereof. However, my oil is clear and my CO & moisture eye show no sign of distress.

Either I have a very unreliable sensor (scary thought) or I gotta be leakng air in somewhere.

A quick and reckless calculation is I would only have to leak 0.15 LPM in to drop my nitrox blend 3%. (That seems too low to be right so i'll recalc when I have had my second cup of java.)

Anyone with a clear head, the actual question is: 5CFM compressor production, Stick at 34%, output mix 28%- How much air had to sneak in to dilute the flow that much?

 

[rmc77]

Maybe I messed up converting the units, but I got 2.3CFM or 65LPM...

 

[CaptBill]

Had the same problem at a shop I worked at. We got the compressor tech to come in and go over the system. He fixed a few leaks in the compressor and that corrected it to about a 1% difference. So if I wanted a 32% fill I set the system at 33%. IMHO the Stick is just a leaky system and you have to fill by your experience on your system, and not just by the O2 sensor.

 

[Fl_Key_Guy]

Our friend in Brazil rmc77's calcs matched my cleared head numbers.

The hose clamps on what I thought were adequately snug connections seems to have solved most of the loss. It is amazing how much leakage can result when things get shaking I guess. Almost half of my final mix was leak after the stick! By the way, it was raining during part of my tank filling today and humidity does seem to play a part. The oxygen is bone dry, the air moist. The output never exactly matched the input. Fortunately, it is not critical for me unless it came out too high as I am using it on an air computer under 115 FSW for the safety margin and the nitrox benefits that the "experts" claim don't really exist. "Old fart's gas"

The calc, in case anyone needs it:
where
N=nitrox blend in the stick
A=air sneaking in
M=final mix at total output=142 liters=5CFM

0.34N + 0.21A = 0.28M
N + A = 142 liters = 5cfm
Solve for N or A and substitute for the other, as in:
0.34(142-A)+.21A=39.8 lpm
48.3-0.34A+0.21A=39.8 lpm
-0.13A=-8.48 lpm
A=65 lpm and N=77 lpm

Thanks to all for the help.

 

[Fl_Key_Guy]

Also, I'll be trying "Scuba Duck" Derek's calibration correction spreadsheet (link offered in this thread) next mix to see how much closer the results are. It is simple enough to use and certainly the temperature and humidity are accounting for some apparent loss.