Net benefits and true cost of owning a compressor.

[rddvet]

Its a very simple concept. you put in x% O2 and you top off. No bunch of fancy regulators and flow meters, continuous sampling that you see in shops with lots of cash to invest in a quality CB system. Its not set it and forget it. Its not so portable either. Its great for banking nitrox. Its great for those that have constant supply presure of O2, but for a tank by tank fill PP is the fastest and least expensive way to do it. Yes you can pump the O2 tank dry and not waste cash on refilling partial filled O2 tanks. Even the small shop that used to function in my area attempted CB filling. NO its not just a stick and a hose coming form an O2 tank. The output had a regulator on it to provide air to a flow meter to provide gas to the O2 sampler. Seldom was the fo2 measured on the rig the same as in the tank an hour later after it cooled. Of course yo can get the automatic regulators that adjust the O2 feed. And that is pretty well set and forget it, but only after lots of money is invested. If you are filling for your self. Its really just too expensive to do it with out constant monitoring of the process on the cheap. If one had 2k cuft of storage then it would be worth the CB filling method. And then o have to have an O2 shut off to insure you dont dump your O2 after the compressor stops sucking. No really expensive to put in but another one of many expences required to set up an automatic CB fill system. but then again in a small operation you would sit there all day waiting for the cascade to fill. Can it be done , YES, can it be done easily NO. Can it be done cheaply? NO Is it easy once you get it working YES. Is it worth it to fill your own 2-4 tanks every other weekend? NO. What is fail safe is that you know if you put in 25% O2 and top off you get 40% nitrox. I dont have to adjust the system to make it, cause im not making anything but air adn using it to top off with.

You can make a chart of O2 for a mix per hundred # of gas or as i do i do it in my head. 2500 # of nitrox 32 thats 800 O2 adn 1700 N2. .2/10 of 1700 is the O2 that is with the N2 from theh air. Then 800 - 340 is the Psi of O2 tp put in the tank and then top off. Or you just know that you need 18-19% O2 to mis nitrox 32. You can cut is a bit easier and say 20% of 2400 psi (2500-100) and find out the O2 amount to fill with O2 prior to top off. No matter how you do it,,,,,, it is easier than CB for a small operation wehn it comes to cost of building the set up that does not require constant monitoring. This is not to say that there is not definate benefits to CB blending, There is. Think about the amount of waste of O2 you have when PP blending compared to the cost of bullding the system.

I guess we'll have to agree to disagree. I have a home built stick that I then had my local compressor guy tap and fit the same oxygen line he puts in his sticks he builds professionally. My setup includes the oxygen reg, a high pressure pop off in case oxygen pressure builds too high(which never happens), and a flow meter into the stick, and an oxygen shut off solenoid for when compressor stops. It wasn't that expensive (around $300). It's easy to use. Set the % and it doesn't vary much through fills. Maybe 0.5% max so I just watch it. You say your experience if final % doesn't match input %? I have no idea why. I see that all the time with pp blending, never with cb.
I just don't really see what you're saying as reality, but I don't own a shop. I do fill a ton of tanks via cb with nothing but ease and accuracy. I think the majority here that have pp and cb filled would say cb makes the most sense in any size situation.
It sounds like you haven't experienced a properly setup cb system. How is paying attention to oxygen flow rates into a tank and monitoring pressures in fill bottles easier?

 

[NAM001]

I guess we'll have to agree to disagree. I have a home built stick that I then had my local compressor guy tap and fit the same oxygen line he puts in his sticks he builds professionally. My setup includes the oxygen reg, a high pressure pop off in case oxygen pressure builds too high(which never happens), and a flow meter into the stick, and an oxygen shut off solenoid for when compressor stops. It wasn't that expensive (around $300). It's easy to use. Set the % and it doesn't vary much through fills. Maybe 0.5% max so I just watch it. You say your experience if final % doesn't match input %? I have no idea why. I see that all the time with pp blending, never with cb.
I just don't really see what you're saying as reality, but I don't own a shop. I do fill a ton of tanks via cb with nothing but ease and accuracy. I think the majority here that have pp and cb filled would say cb makes the most sense in any size situation.
It sounds like you haven't experienced a properly setup cb system. How is paying attention to oxygen flow rates into a tank and monitoring pressures in fill bottles easier?

In your scenario it will work. but not for the normal schmuck filling 1-2 tanks a week. Its going to take the first 1/2 tank to get the O2 output stable at a say 32%. by then you have already filled [part of the the tank with less than 32%. . A 6 cu ft compressor takes about 10-12 minutes to top off a used tank. say from 500 psi to full. if it takes 3-4 minutes to finally read 32% at the input of the compressor , you have already pumped about 20 cuft or more with less than 32%. Then depending where you are sampling the filtered air you add even more time lag. say 15 minutes or so depending on how large the filter system is and if the filter system was already at 3000 psi and 32%. to absorb the starting variation of the FO2 till it gets stable. Your post just said tat you fill tons of tanks. As such you get the system stableized and you keep it there because it does not have a week to bleed down, My filter can loose 1000 psi by sitting a week. in which case your premis of CB is valid. Unfortunately when you fill 1 or 2 tanks a week you dont have that full system charges at filling pressure at the desired FO2. You often have to recharge the filter system, that takes perhaps 5 minutes to get it up tho the psi to overcome the back pressure reg and allow flow to the fill whip. Then you sample the gas at the whip and you find it low,,, now yo have to increase the O2 flow to make up for it. Filling 50 tanks at a time allows for ou to CB fill but not a couple. We found nearly a 10 minute lag in the filter output fo2 change when changing the flow rate at the stick. Its the size of the operation that determines whether you can accurately CB. Tht is why if filling only a COUPLE OF TANKS its easier to PP.

 

[Brandon]

I dunno... I'm a normal schmuck filling just a few tanks a week, and continuous blending works fine for me. My typical fill session is anywhere between 1-3 hours, usually on Thursday or Friday evenings. 200-600 CF of gas at a time - doubles, singles, stages, nitrox/trimix... whatever I'm using that weekend.

I don't think it's necessarily the size of the operation that matters - it's what your main demand is. I rarely dive air and prefer 32%, so I'm set up to blend 32% as easily and efficiently as possible. I also PP blend trimix (21/35 and 18/45) and deco gas (50%), but for all three of those I add a single gas (HE or O2) and top with 32%.

The main benefit of CB for me is that I don't have to worry about the pressure of my supply cylinders. I can take the O2 bottles down to about 20 PSI. I'm not at all adverse to math, but it is nice that there isn't any math or table reference required for my typical fills.

Here's a portable CB system I built when I was living on a remote island in the Bering Sea. Total cost was about $200. Was filling for myself and one other person, two sets of doubles and a few singles at a time, every other week or so. This is pretty much as simple as it gets for CB.



This system had constant monitoring of the O2 input to the compressor via the O2 analyzer, which was within .1-.3% of the output. It was easy to set up, extremely simple to operate, and totally worth it to fill 2-4 tanks every other weekend. The only added complication vs filling air was connecting the input hose, calibrating the O2 sensor and setting the O2 flowrate. Took all of a minute. Yes, you had to fiddle with the flowrate every couple minutes or so, but you were always nearby monitoring the compressor anyway, so that wasn't a big deal.

My current CB system is similar, but mounted on a panel in my garage. Total cost for the CB components was still about $200.



You can make it as complicated as you want. Eventually I got bored and added in an automatic system for controlling the O2 feed, which is an Arduino that monitors the input O2 sensor and controls the O2 flow via a servo connected to the flowmeter. This also serves as the O2 cutoff if the O2 levels get too high or too low. I also added continuous monitoring for HE, O2 and CO on the output side by plumbing my helium analyzer and CO analyzer into the fill whip with a 2 LPM jeweled orifice. Total cost for those additions was about $250, not including the analyzers. I'm always tinkering with it... wait until you see the next iteration.



When I want to fill 32%, I flip the switch on the compressor, and press the on/off switch on the mixer. The mixer automatically calibrates, and about 15 seconds later I open the valve on the O2 supply cylinder and press the green button on the mixer to start blending 32%. From then on, the blending process is automated and it's just like filling air. 30 seconds of setup for all the nitrox you care to blend in that session. Can do custom blends too, just takes an extra few seconds to change the setpoint.

Yes... there are little idiosyncracies along the way. I have most of the leaks taken care of, and my system reliably stays pressurized to 2000 PSI. The startup process is fast - it stabilizes at the selected input within 30 seconds. That's all of perhaps 2CF of gas - not enough to affect the final mix by more than a fraction.

If I do need to switch from 32% to 21% I just run the compressor until I start to see the oxygen content falling on the output side. Then I swap to my shop air supply cylinder and fill that until the residual O2 in the filter/lines is back to 20.9%. Then I fill air. I reverse that process when swapping from 21% to 32%, and the supply cylinder stays full. My shop air has an O2 content of around 23%. No big deal.

Alternatively, it's easy enough to learn to fudge it once you understand how your particular system operates. My 3 CFM compressor and filters take about 8-11 minutes to stabilize with the new input. Let's assume I'm starting with air. With 500 PSI of 32% in HP100s to begin with, I've added about 500 PSI of air by time the compressor output stabilizes. My fill cylinders are now at roughly 26%. If I had 1000 PSI of 26%, and I wanted 3500 PSI of 32%, I'd need to add 2500 PSI of 34%. So I set the input to 34% to start with. When they get to about 3000 PSI, I'll switch the input to 32% and by time they hit 3500 PSI, the fill cylinders and the compressor will all be at about 32%. Then I move on to the next fill.

I'm always within +/- .5% on the final fill, but usually closer to +/- .2%. That's as least as accurate as the analyzers are, and far more precise than is actually needed. I think the TDI blending standard is what... 1%?

I think it's easy, and I have better fills than the local shops have ever been able to do for me... but I'm just a schmuck filling out of his garage.

 

[rddvet]

In your scenario it will work. but not for the normal schmuck filling 1-2 tanks a week. Its going to take the first 1/2 tank to get the O2 output stable at a say 32%. by then you have already filled [part of the the tank with less than 32%. . A 6 cu ft compressor takes about 10-12 minutes to top off a used tank. say from 500 psi to full. if it takes 3-4 minutes to finally read 32% at the input of the compressor , you have already pumped about 20 cuft or more with less than 32%. Then depending where you are sampling the filtered air you add even more time lag. say 15 minutes or so depending on how large the filter system is and if the filter system was already at 3000 psi and 32%. to absorb the starting variation of the FO2 till it gets stable. Your post just said tat you fill tons of tanks. As such you get the system stableized and you keep it there because it does not have a week to bleed down, My filter can loose 1000 psi by sitting a week. in which case your premis of CB is valid. Unfortunately when you fill 1 or 2 tanks a week you dont have that full system charges at filling pressure at the desired FO2. You often have to recharge the filter system, that takes perhaps 5 minutes to get it up tho the psi to overcome the back pressure reg and allow flow to the fill whip. Then you sample the gas at the whip and you find it low,,, now yo have to increase the O2 flow to make up for it. Filling 50 tanks at a time allows for ou to CB fill but not a couple. We found nearly a 10 minute lag in the filter output fo2 change when changing the flow rate at the stick. Its the size of the operation that determines whether you can accurately CB. Tht is why if filling only a COUPLE OF TANKS its easier to PP.

We're still,going to have to agree to disagree. Yes, I fill a ton of tanks, but I also regularly fill one or two tanks at a time. I also have had a leak that bleeds down my filter within 48 hours, and I've been too lazy to fix it. So every new set of fills I need to allow for pressurization and dwell time of my filtration. Even with that I get minimal % drift. I'm not sure how you get that it will take 1/2 of the first tank to hit equilibrium.
I really think your experiences with cb filling are very very atypical.

 

[rddvet]

Brandon, I'm jealous of that arduino servo. I have 0 knowledge of that type of stuff so would never be able to implement it. I love my setup and get compliments on it all the time, but man that controller makes me think my setup is crappy

 

[NAM001]

Brandon, You have a nice set up, and by the post you have overcome the problems with CB when you shift from air to ean pumping. Your shop tank is the trick that makes it work. With out it you will tanks for ever getting the output of the filter stack to be the desired mix for the tank. I dont have a shop tank and i have to bleed down the filter stack and then start the CB mixing process. When doing that,,, the only lag in time that exists is the time to refill the filter as the FO2 equilazation process is no linger in play.. The filter output is very close to the stick output when doing this. I never have any concern with going back to air pumping cause if i want air and get 24% in the final tank mix it is a moot thing for the dives being done. It isnt so easy the other way around. Accuracy is much more critical. The equalizsation issue is moot if all you ever pump is one mix. for hose that do both air and ean. It is too difficult to make the CB give the desired product when the mix demands change from tank to tank.

When our LDS attempted this. they tried to change from 32 to 36% the stick said 36 but the tank output said closer to 32 and gradually crept up towards 36. By the time it got full we often ended up with mid 33 instead of 36. That problem was even worse going from air to 32%. we often ended up with 25-26%. We had to mix 40% or greater at the stick, to speed up the equalizing output till the compressor/tank FO2 reached 32% and then alter the O2 bleed for 32% stick output. we eventually just bled down the filter and started the pumping from there. We never thought of using a shop tank to keep the filter psi high till it equalized. Draining the filter used the least O2, for the shop it was not easy to get O2 as we needed a perscription to get it. Of course the larger the filter system the worse the problem becomes.

 

[Brandon]

I understand the point you were making KWS. Filling 21%, 32% and 36% all at the same time would not be fun. I could pull it off plenty accurately, it'd just take an extra 8 minutes or so every time I switched to a different gas. Could also just bleed the whip for 8 minutes too - which would be a better solution than bleeding the filter towers. If I didn't have an extra tank to catch the gas, I'd probably rig up a silencer on a female DIN adapter so I wasn't driving myself batty with all the gas hissing.

Fortunately, 95% of the time I blend 32%. That's really the only nitrox blend people dive around here anyway. Extra bonus is that I can make everything I'm trained to use with CB'd 32% and trans-filled helium and oxygen.

 

[NAM001]

I understand the point you were making KWS. Filling 21%, 32% and 36% all at the same time would not be fun. I could pull it off plenty accurately, it'd just take an extra 8 minutes or so every time I switched to a different gas. Could also just bleed the whip for 8 minutes too - which would be a better solution than bleeding the filter towers. If I didn't have an extra tank to catch the gas, I'd probably rig up a silencer on a female DIN adapter so I wasn't driving myself batty with all the gas hissing.

Fortunately, 95% of the time I blend 32%. That's really the only nitrox blend people dive around here anyway. Extra bonus is that I can make everything I'm trained to use with CB'd 32% and trans-filled helium and oxygen.

The problem with just bleeding the whip and not the tower is that the tower has a bulk of gas at the non desired mix. Im sure there is an algorithm that you could develope to say what mix to CB fro the stick what when combined with a filter tower and filling one tank would provide the right mix. The LDS that used to serve that area had a piping system that would hold 1/3 of a al80 gas fill. That line filed with 3500 psi would easaily raise al al80 8-900 psi with out opening a bank valve. The back pressure control on the filter system was provided by the cascade tanks connected to it. so with banks shut off and opening a whip you drained the filter stack. What eventually happened is that one compressor system was used only for ean mixing. The large compressor pumped air and the bauer 6 cuft made nitrox only. Even after that we did PP blending for the gross mix and trimmed it so to seak with CB topping off. That made the least amount of changing in O2 bleed to the stick. Also any innaccuracies in the bleed to the stick was dampened by the prior PP blending.

Fortunate for me i don't do diving tha is mix critical. if i want 32 the 30-34 is ok. I dont use best mix for depth which allows me to accept less than precise mixes.

 

[clownfishsydney]

I built an oxygen stick and the total cost was about AUS$200 including oxygen regulator. After doing some tests, all I need to do now is decide what % I want, look up a table I made, then set the PSI on the regulator and I get within 0.2% of the target. Simple, easy. If I need to change a tank from one % to another (say 26% to 32%), then I made an Excel spreadsheet which makes it simple to work out what to set the incoming mix to. Again, very accurate, probably 0.5%. I only actually check the mix after I have finished.

 

[BRT]

Brandon, You have a nice set up, and by the post you have overcome the problems with CB when you shift from air to ean pumping. Your shop tank is the trick that makes it work. With out it you will tanks for ever getting the output of the filter stack to be the desired mix for the tank. I dont have a shop tank and i have to bleed down the filter stack and then start the CB mixing process. When doing that,,, the only lag in time that exists is the time to refill the filter as the FO2 equilazation process is no linger in play.. The filter output is very close to the stick output when doing this. I never have any concern with going back to air pumping cause if i want air and get 24% in the final tank mix it is a moot thing for the dives being done. It isnt so easy the other way around. Accuracy is much more critical. The equalizsation issue is moot if all you ever pump is one mix. for hose that do both air and ean. It is too difficult to make the CB give the desired product when the mix demands change from tank to tank.

When our LDS attempted this. they tried to change from 32 to 36% the stick said 36 but the tank output said closer to 32 and gradually crept up towards 36. By the time it got full we often ended up with mid 33 instead of 36. That problem was even worse going from air to 32%. we often ended up with 25-26%. We had to mix 40% or greater at the stick, to speed up the equalizing output till the compressor/tank FO2 reached 32% and then alter the O2 bleed for 32% stick output. we eventually just bled down the filter and started the pumping from there. We never thought of using a shop tank to keep the filter psi high till it equalized. Draining the filter used the least O2, for the shop it was not easy to get O2 as we needed a perscription to get it. Of course the larger the filter system the worse the problem becomes.

Use a flowmeter. You know what your compressor puts out and you can get the O2% very close from the start with the flowmeter.