DIY Nitrox Mixing
Steve Plotkin
cylinders and then topping of with air.
Has anyone got any experience of this ?
Also, where do you get the tables from to show the amount of O2
required to reach a desired mix.?
Cheers
Steve
Sent via Deja.com http://www.deja.com/
Before you buy.
Nigel Hewitt
> I am interested in blending my own Nitrox by cascade filling from O2
> cylinders and then topping of with air.
>
> Has anyone got any experience of this ?
It's irresponsible and you shouldn't do it but I
can't stop you. It's not really cost effective but
I like it. You need a supplier of gas and an
oxygen filling whip. You also need a good
idea about concepts like slow, clean, and cool.
Web search for the Oxygen Hacker's Handbook
or similar.
> Also, where do you get the tables from to show the amount of O2
> required to reach a desired mix.?
You can have mine if you run windows....
http://homepages.pavilion.co.uk/nigelvh/blender.zip
usual disclaimers apply.
--
nigelH
I.M. Smith
> cylinders and then topping of with air.
>
> Has anyone got any experience of this ?
Thought about it, but not done it yet. DPO won't let me rack up tanks in
the house. :-(
> Also, where do you get the tables from to show the amount of O2
> required to reach a desired mix.?
You should be able to work it out. (And if you can't, should you really
be trying to do this?) Probably easiest to write a spreadsheet to do the
job. Don't forget that air is 21% O2!
Iain
Pete Young
>Also, where do you get the tables from to show the amount of O2
>required to reach a desired mix.?
http://www.cisatlantic.com/trimix/tools.htm has several. Bjorn
Forshei's is quite neat and easy to use.
There's also some useful information at
http://www.cisatlantic.com/trimix/mixing.htm
You'll need a filling whip and an accurate pressure gauge (an old
digital air-integrated computer works quite well for this, much better
than using one for diving ;-)
Pete
--
____________________________________________________________________
Pete Young pe...@antipope.org
"Just another crouton, floating on the bouillabaisse of life"
Lee Bell
> I am interested in blending my own Nitrox by cascade filling from O2
> cylinders and then topping of with air.
> Has anyone got any experience of this ?
> Also, where do you get the tables from to show the amount of O2
> required to reach a desired mix.?
Without intending any disrecpect, it's my opinion that someone who's ready
to do partial pressure nitrox fills should, at a minimum, be able to the
match it takes to fill to a desired O2 percentage. The best place to learn
how, get experience and get tables if your really need them is a formal
course.
Lee
Mig
news:8vj9hc$edf$1...@nnrp1.deja.com...
>
> Also, where do you get the tables from to show the amount of O2
> required to reach a desired mix.?
>
Isn't this a matter of very basic arithmetic, or am I being naive ?
Mig
John Baker
> cylinders and then topping of with air.
>
> Has anyone got any experience of this ?
>
> required to reach a desired mix.?
>
Mig
ought to put my money where my mouth is and try it.
I reckon that letting R = percentage O2 required in final mix
% of pure O2 required = 100 - ( 5 (100 - R) / 4 )
Thus for a 32% mix in a 300 bar cylinder
% pure O2 Required = 100 - (5 (100 - 32 ) / 4 ) = 37.5 %
37.5% x 300 bar = 112.5 bar of pure oxygen
Then top up with air
I have assumed that air is 4/5 O2 but if you needed more accuracy you could
just put 21/79 into the equation instead. I can't think of any reason why
this should not work - the filling should be linear so long as temp remains
reasonably steady.
However I must admit it wasn't as trivial a problem to work out as I thought
at first glance (and I could have made a silly mistake somewhere).
Mig
Mig @iname.com> <mkmitchell<no-spam> wrote in message
news:3a1e...@news.telinco.net...
> Steve Plotkin <stev...@my-deja.com> wrote in message
> news:8vj9hc$edf$1...@nnrp1.deja.com...
> >
> > Also, where do you get the tables from to show the amount of O2
> > required to reach a desired mix.?
> >
>
Nigel Hewitt
> It occured to me afterward that this was very easy to say so I thought I
> ought to put my money where my mouth is and try it.
>
> I reckon that letting R = percentage O2 required in final mix
>
> % of pure O2 required = 100 - ( 5 (100 - R) / 4 )
>
> Thus for a 32% mix in a 300 bar cylinder
>
> % pure O2 Required = 100 - (5 (100 - 32 ) / 4 ) = 37.5 %
>
> 37.5% x 300 bar = 112.5 bar of pure oxygen
>
> Then top up with air
>
> I have assumed that air is 4/5 O2 but if you needed more accuracy you could
> just put 21/79 into the equation instead. I can't think of any reason why
> this should not work - the filling should be linear so long as temp remains
> reasonably steady.
Sorry.
112 bar prefill is a formula for about a 50% mix.
Up to 230bar you can get away with simple maths
but at 300 you're way out. Alistair and I have a
perpetual sniping battle over the precise mechanism
but over 200bar the Van der Waal forces are significant
contributors.
Up to that point though the maths is simple but
surprisingly it is better done with nitrogen.
Existing tank has P1 bar of F1 fraction Oxygen 120 0.3
You want P2 and F2 final results 200 0.35
So initial tank has PN1 = P1*(1-F1) nitrogen 84
You want PN2 = P2*(1-F2) 130
PN3 = PN2-PN1 of Nitrogen in the air fill 54
so that carries PIO3 = PP3*21/79 of Oxygen with it 14.35 ish
So the oxygen sum is:
You started with PO1=P1*F1 36
You end with PO2=P2*F2 70
You added PO3 in the air fill 14.35
So you added PO4 = PO2-PO1-PO3 19.65
So you oxygen fill from 120 to 140 but if your supply tank is
below about 150 bar you will have to let the tank to be filled
down a bit which means you use more oxygen as you are throwing
away an already rich mix.
If you always dump the tank you can ignore the P1 F1 stuff but
that is where the economies of home mixing come. Shops tend not
to allow for that in their pricing.
Easy on a spread sheet. If you use the VdW calculation you only
get 1 bar less on the oxygen fill at 200 but at 300 bar final
pressure the oxygen fill is 8 bar different and the final mix
is about 37% if you use the simple maths above.
My blender program for windows is a freebie on the VdW page
http://homepages.pavilion.co.uk/nigelvh/nv-vdw.html
and you can switch ideal gas law/VdW on the system menu.
--
nigelH
Nigel Hewitt
> so that carries PIO3 = PP3*21/79 of Oxygen with it 14.35 ish
When he meant
so that carries PO3 = PN3*21/79 of Oxygen with it 14.35 ish
sorry and all
I checked the sums but missed the typos
--
nigelH
Nigel Hewitt
> It occured to me afterward that this was very easy to say so I thought I
> ought to put my money where my mouth is and try it.
>
> I reckon that letting R = percentage O2 required in final mix
>
> % of pure O2 required = 100 - ( 5 (100 - R) / 4 )
>
> Thus for a 32% mix in a 300 bar cylinder
>
> % pure O2 Required = 100 - (5 (100 - 32 ) / 4 ) = 37.5 %
>
> 37.5% x 300 bar = 112.5 bar of pure oxygen
>
> Then top up with air
>
> I have assumed that air is 4/5 O2 but if you needed more accuracy you could
> just put 21/79 into the equation instead. I can't think of any reason why
> this should not work - the filling should be linear so long as temp remains
> reasonably steady.
<egg on face>
Well I have canceled my two previous posts but usenet being
usenet they'll probably carry my typos round the world...
This is what I ought to have said
</egg>
Sorry.
112 bar prefill is a formula for about a 50% mix.
Up to 230bar you can get away with simple maths
but at 300 you're way out. Alistair and I have a
perpetual sniping battle over the precise mechanism
but over 200bar the Van der Waal forces are significant
contributors.
Up to that point though the maths is simple but
surprisingly it is better done with nitrogen.
Existing tank has P1 bar of F1 fraction Oxygen 120 0.3
You want P2 and F2 final results 200 0.35
So initial tank has PN1 = P1*(1-F1) nitrogen 84
You want PN2 = P2*(1-F2) 130
PN3 = PN2-PN1 of Nitrogen in the air fill 46
so that carries PO3 = PN3*21/79 of Oxygen with it 12.13 ish
So the oxygen sum is:
You started with PO1=P1*F1 36
You end with PO2=P2*F2 70
You added PO3 in the air fill 12.13
So you added PO4 = PO2-PO1-PO3 21.87
So you oxygen fill from 120 to 142 but if your supply tank is
below about 150 bar you will have to let the tank to be filled
down a bit which means you use more oxygen as you are throwing
away an already rich mix.
If you always dump the tank you can ignore the P1 F1 stuff but
that is where the economies of home mixing come. Shops tend not
to allow for that in their pricing.
Easy on a spread sheet. If you use the VdW calculation you only
get 1 bar less on the oxygen fill at 200 but at 300 bar final
pressure the oxygen fill is 8 bar different and the final mix
is about 37% if you use the simple maths above.
My blender program for windows is a freebie on the VdW page
http://homepages.pavilion.co.uk/nigelvh/nv-vdw.html
and you can switch ideal gas law/VdW on the system menu.
--
nigelH
Now I go and hide...
Lee Bell
> > Also, where do you get the tables from to show the amount of O2
> > required to reach a desired mix.?
>
> Isn't this a matter of very basic arithmetic, or am I being naive ?
It's pretty basic provided you account for the gas already in the tank when
filled. In most cases, tanks are drained to or very near empty. What you
have to account for is the fact that an empty tank isn't really empty. It's
got one atmosphere (bar) of gas that has to be accounted for to get the
final mix desired. This complicates the arithmetic a bit, but not beyond
the ability of most to handle it. Mix tables take this into account for
you.
Another factor that will effect the final mix is temperature change during
the filling process. As a tank is filled, some increase in temperature is
inevitable, no matter how slowly you fill it and that changes the pressure
of the O2 introduced first and the air used to dilute it to the desired mix.
Accordingly, even filling based on a mix table, if a specific mix is
desired, can be at least partly art rather than entirely science.
Lee
Lee Bell
> It occured to me afterward that this was very easy to say so I thought I
> ought to put my money where my mouth is and try it.
> I reckon that letting R = percentage O2 required in final mix
> % of pure O2 required = 100 - ( 5 (100 - R) / 4 )
> Thus for a 32% mix in a 300 bar cylinder
> % pure O2 Required = 100 - (5 (100 - 32 ) / 4 ) = 37.5 %
> 37.5% x 300 bar = 112.5 bar of pure oxygen
> Then top up with air
> I have assumed that air is 4/5 O2 but if you needed more accuracy you
could
> just put 21/79 into the equation instead. I can't think of any reason why
> this should not work - the filling should be linear so long as temp
remains
> reasonably steady.
> However I must admit it wasn't as trivial a problem to work out as I
thought
> at first glance (and I could have made a silly mistake somewhere).
I think you did make a mistake, but I don't think in bar (U.S. bias, sorry),
making it a bit more difficult for me to be sure. You can check more easily
than I. Did you account for the gas in the tank at the beginning of the
process? BTW, temperature doesn't remain steady, it increases during the
filling process, even for a very slow filling process.
As a purely intellectual exercise, I put together an Excell spreadsheet to
calculate how much oxygen and air to add to a tank that was not "empty" at
the start of the process. While it's relatively easy to account for the 1
bar in an "empty" tank, it gets a bit more complex when the tank is
partially full. This is not really a practical approach in most cases
because O2 is not normally available at pressures required to add the gas
required to bring a partially full tank to the desired concentration, at
least it's not normally available here in the States. I did not account for
temperature change. That would have required more research than I was ready
to put into the project.
Lee
Lee Bell
> My blender program for windows is a freebie on the VdW page
> http://homepages.pavilion.co.uk/nigelvh/nv-vdw.html
> and you can switch ideal gas law/VdW on the system menu.
Curse you Nigel Hewitt. You provided more information than I wanted to
know, but, unfortunately, not more information than I should know if I'm to
understand what I'm doing . . . so bless you too. I downloaded your sheet
and bookmarked your site. I'll no doubt spend considerable time with it
before fully understanding what you've done to me, if I ever do. I too had
university level physics, and thermodynamics, and both calculus and
differential equations, but it's been 30 years since I used any of them.
Wish me luck.
>Now I go and hide...
As you should. 8^).
Lee
Lee Bell
Please, do it all again, this time in Farenheit, psi, and pounds. 8^).
Just kidding. If I can sort through your information, I can convert the
results.
Lee
David harrison
reliable analyser & an O2 decanting hose. I have knocked an Excel
spreadsheet together that covers most of the basics. Please contact me
if you want me to email it to you. I would post it here but I would
probably start a flame war & I can't be bothered with all that. (polite
way of saying what I really mean) Some people should get out more!
Regards
Dave Harrison
Steve Plotkin wrote:
> I am interested in blending my own Nitrox by cascade filling from O2
> cylinders and then topping of with air.
>
> Has anyone got any experience of this ?
>
> Also, where do you get the tables from to show the amount of O2
> required to reach a desired mix.?
>
Edward Watson
from "Lee Bell" <lee...@ix.netcom.com> contains these words:
> Another factor that will effect the final mix is temperature change during
> the filling process. As a tank is filled, some increase in temperature is
> inevitable, no matter how slowly you fill it and that changes the pressure
> of the O2 introduced first and the air used to dilute it to the desired mix.
> Accordingly, even filling based on a mix table, if a specific mix is
> desired, can be at least partly art rather than entirely science.
In practise I've never had a problem with this, and neither have the
others locally, although our O2 cylinders will never get as hot in
our less tropical climate than they would in Florida - I just put in
the required O2, reasonably slowly, then top off, the first bit
slowly, then faster as the perncetage of O2 decreases, allow about
10% over on the compressor gauge, and by the time its cooled down and
analysed its rarely more than 1-1.5% either way off the mark
(according the the alpha-one analyser anyway)
F
Nigel Hewitt
> "Lee Bell" wrote
>
> > Another factor that will effect the final mix is temperature change during
> > the filling process. As a tank is filled, some increase in temperature is
> > inevitable, no matter how slowly you fill it and that changes the pressure
> > of the O2 introduced first and the air used to dilute it to the desired mix.
> > Accordingly, even filling based on a mix table, if a specific mix is
> > desired, can be at least partly art rather than entirely science.
>
> In practise I've never had a problem with this, and neither have the
> others locally, although our O2 cylinders will never get as hot in
> our less tropical climate than they would in Florida - I just put in
> the required O2, reasonably slowly, then top off, the first bit
> slowly, then faster as the perncetage of O2 decreases, allow about
> 10% over on the compressor gauge, and by the time its cooled down and
> analysed its rarely more than 1-1.5% either way off the mark
> (according the the alpha-one analyser anyway)
Argh!
I went to home filling because I was tired of way out
fills which I defined as a whole 1% out.
If the oxygen is right to get 1% over you have to short
fill by 10% or more which implies 20C over temperature
if the shops pressure gauge is right.
Provided I am careful to pull the usual tricks (like
always reading the whip gauge on a rising pressure and
sloooow decant and leave the tank and get them to top
it off when cold I expect to get only decimals of error
(normally down to the shop's idea of final pressure being
different from mine).
I definitely don't like electronic pressure gauges as
they are not even consistent so I don't bother to ask
questions like accurate.
I have 33% in the tanks today. I asked for 30 but I was
away from home and they were hurrying so I was pleased
to get any fill. If I had a 30% plan that would put my
margin down considerably but I'll just have to factor
it into the next dive choice.
--
nigelH
Edward Watson
from "Nigel Hewitt" <no.spam.pl...@combro.co.uk> contains
these words: >
> Argh!
> I went to home filling because I was tired of way out
> fills which I defined as a whole 1% out.
How accurate do you think your analyser is?
Nigel Hewitt
> "Nigel Hewitt" wrote
> > Argh!
> > I went to home filling because I was tired of way out
> > fills which I defined as a whole 1% out.
>
>
> How accurate do you think your analyser is?
OK so it's rated at 1% but I do check it at 100%
and when that does not give me air accurately I
change the sensor. Actually the Galvanic sensors
should be a lot better than 1% but they do suffer
from temperature sensitivity so the manufacturers
don't rate them better.
--
nigelH
Lee Bell
> OK so it's rated at 1% but I do check it at 100%
> and when that does not give me air accurately I
> change the sensor. Actually the Galvanic sensors
> should be a lot better than 1% but they do suffer
> from temperature sensitivity so the manufacturers
> don't rate them better.
True, but understated. They are sensitive to temperature, humidity and
pressure of the gas being measured at the sensor. To be at their full
accuracy, you would have to:
1. Adjust for humidity and temperature. Mine comes with a chart . . . which
I ignore since all my measurements are made under similar conditions (sea
level w/high humidity).
2. Ensure the pressure of the gas used to calibrate the unit is the same as
the pressure of the gas you're measuring.
Lee
Nigel Hewitt
> Nigel Hewitt wrote
>
> > Actually the Galvanic sensors
> > should be a lot better than 1% but they do suffer
> > from temperature sensitivity so the manufacturers
> > don't rate them better.
>
> True, but understated. They are sensitive to temperature, humidity and
> pressure of the gas being measured at the sensor. To be at their full
> accuracy, you would have to:
> 1. Adjust for humidity and temperature. Mine comes with a chart . . . which
> I ignore since all my measurements are made under similar conditions (sea
> level w/high humidity).
> 2. Ensure the pressure of the gas used to calibrate the unit is the same as
> the pressure of the gas you're measuring.
The humidity one surprises me. I would have thought that
the humidity of a decompressing gas was vanishingly small.
As to pressure I thought they measured partial pressure
(well actually oxygen density but it's effectively the
same thing) so as I can't get a significant pressure
difference on the pipe thing on my Vandergraph unless I
open the tank valve stupidly that was not a problem either
as atmospheric pressure compensates out when I dial up
20.7% to calibrate it.
Am I missing a trick here? I have no chart so I don't know
what order of compensations you are talking about.
--
nigelH
Lee Bell
> > True, but understated. They are sensitive to temperature, humidity and
> > pressure of the gas being measured at the sensor. To be at their full
> > accuracy, you would have to:
> > 1. Adjust for humidity and temperature. Mine comes with a chart . . .
which
> > I ignore since all my measurements are made under similar conditions
(sea
> > level w/high humidity).
> > 2. Ensure the pressure of the gas used to calibrate the unit is the same
as
> > the pressure of the gas you're measuring.
> The humidity one surprises me. I would have thought that
> the humidity of a decompressing gas was vanishingly small.
You misunderstood, or rather, I failed to communicate adequately. It's the
humidity of the gas used to calibrate the unit that is most variable unless,
of course, the calibration gas has been filtered for tank filling too. If
you calibrate in open air, as most divers here do, you have to account for
the humidity of the air to be most precise.
> As to pressure I thought they measured partial pressure
> (well actually oxygen density but it's effectively the
> same thing) so as I can't get a significant pressure
> difference on the pipe thing on my Vandergraph unless I
> open the tank valve stupidly that was not a problem either
> as atmospheric pressure compensates out when I dial up
> 20.7% to calibrate it.
You are right that it is partial pressure that is being analyzed, but it's
the same only at the same pressure. If you calibrate in open air and test
under a pressure significantly different from 1 ata, you will get a false
reading. Tank valves are not made for precision control of flow and it does
not take much of a twist to change the pressure significantly. Of course,
the design of the test chamber has a lot to do with how much of an increase
in pressure results from how much of a change in flow. While this sounds
easy to control, it's not as easy as it seems. My O2 stick has a flow
restrictor, but not a very good one. It's only effective for very low flow
rates. I can easily over pressurize it. The analyzer I learned on had an
open T chamber with the gas flowing through the straight section and the
analyzer attached to the T portion. This type of analyzer does not
overpressurize quite as easily, but is still not all that hard to get a
false reading from. I've found it quite easy to change the reading on both
types significantly with only a slight increase in flow rate.
The best situation is to calibrate and test at the same humidity,
temperature and flow rate. The most common way to do this is with a unit
that includes a flow restrictor and takes the gas from the intermediate
inflator hose of a regulator. Calibrate on a tank of air at stabilized
temperature, then test a tank of nitrox, also at stabilized temperature,
using the same regulator (to ensure the same intermediate pressure.
Please note that we're talking maximum precision here, not necessarily
something required for real life testing, where + or - 1% is generally
considered adequate. O2 toxicity, N2 narcosis and DCS are not precise
concepts. Since these are the primary considerations in O2 analysis, there
is a point where precision of measurement becomes an intellectual issue
only. In fact, some O2 analyzers don't even read to 10th of a percent, the
older model Handi's being a good example (I think I read that the new ones
read in 10ths).
> Am I missing a trick here? I have no chart so I don't know
> what order of compensations you are talking about.
I'll look around and see if I can still find the chart that came with my O2
stick. Just to be sure I don't miscommunicate again, the adjustment is made
to the calibration. I'm not going to guess at the range or direction,
mostly because I'd probably guess wrong.
Lee
Nigel Hewitt
> You misunderstood, or rather, I failed to communicate adequately. It's the
> humidity of the gas used to calibrate the unit that is most variable unless,
> of course, the calibration gas has been filtered for tank filling too. If
> you calibrate in open air, as most divers here do, you have to account for
> the humidity of the air to be most precise.
YES! I missed that entirely!
> You are right that it is partial pressure that is being analyzed, but it's
> the same only at the same pressure. If you calibrate in open air and test
> under a pressure significantly different from 1 ata, you will get a false
> reading. Tank valves are not made for precision control of flow and it does
> not take much of a twist to change the pressure significantly. Of course,
> the design of the test chamber has a lot to do with how much of an increase
> in pressure results from how much of a change in flow. While this sounds
> easy to control, it's not as easy as it seems. My O2 stick has a flow
> restrictor, but not a very good one. It's only effective for very low flow
> rates. I can easily over pressurize it. The analyzer I learned on had an
> open T chamber with the gas flowing through the straight section and the
> analyzer attached to the T portion. This type of analyzer does not
> overpressurize quite as easily, but is still not all that hard to get a
> false reading from. I've found it quite easy to change the reading on both
> types significantly with only a slight increase in flow rate.
So I spent ten minutes playing and get virtually no change
on the Vandergraph so perhaps that big T-piece and pipe kit
on it works well.
> Please note that we're talking maximum precision here, not necessarily
> something required for real life testing, where + or - 1% is generally
> considered adequate.
Combro is an instrumentation company. I care about accuracy.
Look at it as my hobby (or maybe that should be obsession). 8-)
> I'll look around and see if I can still find the chart that came with my O2
> stick.
I would appreciate seeing some example numbers if
it shows up but don't turn the house upside down.
--
nigelH
Lee Bell
> So I spent ten minutes playing and get virtually no change
> on the Vandergraph so perhaps that big T-piece and pipe kit
> on it works well.
It worked pretty well when I trained on it, but there were people in the
class who had to be coached to take it easy on the valve. They seemed to
think that a tank valve only has two positions, closed and fully open.
> > Please note that we're talking maximum precision here, not necessarily
> > something required for real life testing, where + or - 1% is generally
> > considered adequate.
>
> Combro is an instrumentation company. I care about accuracy.
> Look at it as my hobby (or maybe that should be obsession). 8-)
Well, then, perhaps the known pressure/flow restricted system is just right
for you. One of the fill stations I use has one. The tester is a bit bulky
for travel, but about as precise as anything I've seen. It consists of a
regulator first stage with the analyzer hooked to an intermediate pressure
port. The line has a flow restrictor and, downstream of the the analyzer,
venting for the gas flow. It also has a pressure release to allow the first
stage to be removed from the tank easily. The station has a tank of air
available for calibration. The process is, hook the regulator to the air
tank and calibrate it (I use 20.9). Remove the unit from the air tank and
attach it to the nitrox tank and take a reading. Because humidity is very
similar (both dry breathing gas) and the intermediate pressure and,
therefore, the limited flow rate are as close as possible, the readings are
as accurate as I think you'll find without a gas spectrometer. Only the
temperature is a problem. A tank that has just been filled is not at the
same temperature as one that was filled earlier and has stabilized at
ambient temperature. Generally, I handle this by giving my tanks as much
time as possible to stabilize before recording the contents. I normally
keep my tanks full, which gives me lots of time between the fill and
recording contents. The added advantage to this is that a tank that has
been filled only to its service pressure, will not remain at service
preassure when stabilized. I have the opportunity to have it topped off
if/when the volume of my gas supply is likely to be important. Usually,
it's not. I'm what you might call easy on gas.
Handi makes on optional low pressure inflator connector with a flow
restricter built in. This unit is compact enough to travel with easily, but
unless they have changed (I have heard they have), the unit only reads to
the nearest percentage. The newer ones may read to the nearest .10 ata.
Lee
