Tank Gas Management ... a new approach

pat w

Member
Nov 4, 2009
462
0
16
Daphne, AL (east Mobile Bay)
Edit: Design updated and more detailed writeup in Post #18

I've been following two topics of late; the discussion begun by Tom on the gas concentration differences between canister system and wet/dry systems and their relative characteristics regarding O2:CO2 ratios, and the regrettable news about Gerry gassing a large number of his fish due to an understandable mistake. These and general nagging concerns of my own over the “gas” balance in my own tank prompted this exercise.

First let me state that this is by no means a condemnation of the methods currently in use in planted tank gas management. It is however an attempt to add to the current dialog and explore possible improvements and perhaps a new direction.

I've always felt the the meticulous metering of CO2 was a possibly weak aspect of the hobby. The very length of the Dual Stage Regulator topic is testament to the attention this particular area attracts. The topic is littered with very creative methods of achieving levels of control that would not otherwise be available without the outlay of large sums of money. The fact that the normal cost would be out of the reach of the average person says to me that the difficulty of metering gas into our tanks with the level of control needed is prohibitively complex and perhaps a new approach may be indicated. In searching for a better method I explored rotometers as perhaps a better way. There was some promise, but the meters which would give control and feedback at the very low flow levels were again too costly.

Had I been looking for a low water flow solution there were a number of comparatively economical rotometer choices. So I asked myself … What could be done to move away from trying to control very low gas flows and go instead to controlling low or even moderate water flows? These are some of the items I came upon:

-----
1.The water metered into the tank would need to have a CO2 concentration that was predictable and constant.

2.It should be non-dependent on things such as end of tank dumps and needle valve variations to allow for the use of less expensive and less precise equipment.

3.It should also assist in the balance of night time O2 levels.
-----

The next post will be an offering (for discussion) of a Tank Gas Management Reactor that I'd like to discuss as a possible direction.

Pat
 
Last edited by a moderator:

pat w

Member
Nov 4, 2009
462
0
16
Daphne, AL (east Mobile Bay)
Say hello to Frank, my new design for a Tank Gas Management Reactor.

attachment.php


The general idea is to circulate water while feeding enough CO2 to take it as close to full CO2 saturation as possible. I attempt to achieve that with the generation of three flow paths.

attachment.php


First the Primary Saturation Flow, which is begun by the Circ Pump and is passed back down to the bottom level to feed the pump again; very simple. Directly out of the pump the water is channeled into a nozzle into the narrower venturi tube. A gas feed line is inserted much like a mazzei to introduce CO2 or O2 into the flow at point “B”. A filter sponge or similar substance would be placed into the flow path to block the bubbles of gas from passing into the lower level insuring a solid water stream into the venturi.

attachment.php


Second is the Induced Saturation Flow. This is the flow of water drawn from behind the venturi and should be restricted to the upper level with the clear PVC. This should be the only path for undissolved gas bubbles and the bubbles should circulate continuously. A disc of fine screen material might be placed in the path to chop the bubbles into a fine mist to promote easy circulation.

Last is Pass Through Flow which is induced by whatever pump you use to pump water out of your tank, through the reactor, and back to the tank. The flow will, in actuality, be a mixed pattern as it is mixed with the water circulating within the reactor itself. Before anyone raises the question; yes the flow will eventually make it out of the reactor. I've seen many similar systems at work where flow is passed through a system while a separate pump recirculates and mixes the contents of the system. If you use a filter to supply this flow you will need to supply a additional bypass path with a another ball valve to restrict bypass flow and force the desired MAXIMUM flow through the reactor. This MAXIMUM will later be metered down to regulate the gas delivery to the tank.

attachment.php


CO2 is supplied by the usual CO2 tank/regulator/solenoid however the demands on these components will be greatly reduced. It is delivered to the reactor in relatively brief bursts that will produce a sizable CO2 blanket at the top of the large clear vertical PVC tube with the “A” on top. The “A” is a float switch which takes over the job of switching the CO2 solenoid on the regulator. As CO2 is used the volume of gas in the collection tube will reduce and the water level will rise. The float switch will then turn on the solenoid to replenish the blanket. No more end of tank worries from less expensive regulators and the need for a needle valve is greatly reduced or eliminated entirely. At night the 3-way valve is switched off and instead of CO2, air is injected into the venturi to supply the needed O2. (not shown is a needed check valve) The size of the CO2 collection chamber will allow for the measurment of the actual CO2 delivery rate in milliliters per minute by mesuring the rate of rise in the water level.

attachment.php


The final flow back to the tank is regulated by any flow restricting valve and then to any of several low cost flow meters.

More to come on theory and operation.

Pat
 
Last edited by a moderator:

pepetj

Lifetime Members
Lifetime Member
Oct 8, 2008
188
2
18
59
Santo Domingo, Dominican Rep
Can't wait to read in extensive the rationale of your interesting proposal. How long have you had a prototype working? What have you measured? If I get your idea right it means that your proposed system may noticeably extend the life time of our pressurized CO2 cylinders meaning longer service time per cylinder before refill is needed?

Pepetj
Santo Domingo
 

pat w

Member
Nov 4, 2009
462
0
16
Daphne, AL (east Mobile Bay)
pepetj,

No prototype as yet. Only theory and design, so far. I'll be getting the parts together over time.

Can't say I'd expect any improvement in economy over a more conventional reactor design. The main focus was to move away from regulating low levels of gas flow and trying to use fluid flow instead.

More later,
Pat
 

pat w

Member
Nov 4, 2009
462
0
16
Daphne, AL (east Mobile Bay)
Basic Theory of Operation

I’ve based my design on some premises that may or may not be valid. Some are dependant on others so if one of the earlier ones fails there will be a sort of domino effect.

-------
1. The venturi will draw gas which is not under pressure into the side port. This one is pretty safe as it is used widely in other areas.

2. As the gas to water ratio of the Induced Saturation Flow increases this gas draw will decrease until finally it stops completely while there is still a mix of water and gas in the flow path. This is the big one. If this fails the whole deal is off as eventually the induced flow would be in a state of vapor lock and there would be no difference between this reactor and any other.

3. The foam at “C” will stop the flow of gas to the inlet of the pump but not reduce water inlet pressure enough to cause cavitations. This may be addressed with some sort of baffle system that would rely on the floating of the bubbles and the sinking of the water, but not so much if the bubbles are chopped up too finely.

4. The Induced Saturation Flow will continue to dissolve gas until ‘saturation’ is achieved (Assuming that Pass Through Flow is closed). This should reduce the gas to water ratio and allow more gas to be drawn in until item 2 comes into play.

5. When Pass Through Flow is open an equilibrium will be obtained between the saturation level and the gas to water ratio; dependant on the Pass Through Flow Rate.
------

As the system runs with Pass Through closed (or the supply pump turned off) there should be an increase of gas in the Induced Flow area, but as the gas levels here rise the compressibility of the gas in the mix should reduce the suction allowing less and less gas in until it eventually stops. This assumes that there will be none of the gas entering solution. As gas is dissolved, the gas in the mix will decrease to a point that will allow make up gas to be drawn in.

Now introduce Pass Through Flow. The new water in the system will reduce the gas to water ratio and will be at a lower saturation level allowing for more of the gas to enter solution and subsequently more gas being drawn into the venturi to make up for the loss. The natural tendency of the system to move toward equilibrium should cause the rate of the gas drawn into the venturi to be exactly the same as the rate the gas is dissolved and delivered via the outlet to the tank. This could be tested by placing a bubble counter on the outlet of the CO2 collection tube and observing the action.

A manual switch could be placed in parallel to the float switch in the CO2 collection tube and used to pre-charge the tube with an excess of CO2. This would allow the user to time the rise of the water level in the tube. This should give a solid metric on the amount of CO2 being delivered to the tank as opposed to simply a rate.

At night the 3-way valve will cycle and room air will begin to mix in the induced flow. In a short time the CO2 should be well below potentially hazardous levels. Cycling back just before lights on would bring CO2 levels back up.

Comments and/or Critiques Please ...
Pat
 

pat w

Member
Nov 4, 2009
462
0
16
Daphne, AL (east Mobile Bay)
Float switches; nice one for $11.24
http://www.mcmaster.com/#liquid-level-switches/=9fit3g

3-way valve
http://cgi.ebay.com/1-4-Electric-Solenoid-Valve-3-WAY-120-Volt-AC-B30N-/290482296480?pt=LH_DefaultDomain_0&hash=item43a21816a0

And if you have a smaller tank and a need to go with a lower Pass Through flow rate than 50 GPH Dwyer Visi-float has you covered, including the metering valve, for $46 and change in brass; $56 in stainless which replaces the Tom Aquatics flowmeter and the ball valve.
http://www.gogenlab.com/products/product/F4320-visi-float-flowmeter-4-scale

Pat
 

pat w

Member
Nov 4, 2009
462
0
16
Daphne, AL (east Mobile Bay)
andrei_0;57278 said:
I think the water is more likely to flow like this:View attachment 1865

I guess I needed to do more than just describe the venturi operation. I plan to build up a nozzle that will direct flow into the opening of the narrower section causing an induced flow in the direction previously indicated.

Perhaps this will help.

attachment.php


The nozzle will be narrow enough to fit inside the "T" and allow a good amount of flow around it to be drawn into the narrow section of pipe. This should cause the flow at the other end to split; some going down to make up for the suction of the pump; the other turning up to makeup for the flow induced by the venturi flow.

Thanks for the comment though as now that I see the nozzle in place it looks as if the gas feed line might be better placed in the reducer at the end of the nozzle.


Pat
 

pat w

Member
Nov 4, 2009
462
0
16
Daphne, AL (east Mobile Bay)
Here's another design. Same idea, push the CO2 concentration in the reactor as high as possible and then regulate the level in the tank by controlling the water flow rate back to the tank. The CO2 supply can be in a separate reservoir as long as there is an expansion path for the water back to the tank and the reservoir is lifted so the water preasure doesn't force the CO2 out and into the reactor. It's important that the venturi draw regulate the flow of CO2 into the reactor. I'm still working on this part.

The image below is just the saturation reactor without the CO2 reservoir.

attachment.php
 

pat w

Member
Nov 4, 2009
462
0
16
Daphne, AL (east Mobile Bay)
Left C;57535 said:
Would a flow rate indicator like this Fisher 10A2235A model be useful to you? It can measure from 0.05 to 0.60 gpm of water or 0.10 to 1.30 scfm of air. There are over 10 of them and they are not very expensive. http://0041f2c.netsolhost.com/Ratosight10A2235A.pdf
http://cgi.ebay.com/FISCHER-PORTER-...566?pt=LH_DefaultDomain_0&hash=item56325e3666

Its the right style. I'd have to find a metering valve to go with it. A ball valve would probably work fine. The biggest question is I don't have any idea what level of flow I'll need so the range of the flowmeter is pure guess work right now. Funds are low, so I'll have to wait on anything.

Thanks for the heads up.
Pat
 

pat w

Member
Nov 4, 2009
462
0
16
Daphne, AL (east Mobile Bay)
Something like this is what I was thinking I'd need.

http://cgi.ebay.com/DWYER-RMC-SERIES-FLOWMETER-RMC-141-SSV-1-1-0-GPM-/110603369991?pt=LH_DefaultDomain_0&hash=item19c0797e07

When I had an external reactor I was running it with a Rio 180 (120GPH) I figured I was getting about 80-100 gph. My thinking is that 6-60 with the water fully saturated should be about right. But as I said, it's all guesswork right now.

My first expenditure will be to build the venturi. Without it the flowmeter will be worthless.

Pat
 

pat w

Member
Nov 4, 2009
462
0
16
Daphne, AL (east Mobile Bay)
Venturi build a success - Build pics

I completed and did a brief test of the venturi.

The nozzle is simply a piece of 1/2" PVC glued into a threaded adapter I drilled out to clear the OD of the pipe. I placed a 1/2" to 1/2" slip to thread adapter to one end and installed a 3/8" barb for the nozzle jet. I mounted a garden hose adapter to the other end for the test. This will be removed later to install the pressure feed from the pump.

attachment.php


The body is a pair of "T"s; one 1-1/2" std. and one 1-1/2" to 3/4" reduction. The left and bottom 1-1/2" openings are fitted with 1-1/2" to 1-1/4" female thread reducers. This will make mods and cleaning easier. The exit at the right is fitted with a 1-1/2" to 3/4" reducer slip to female thread for the exit pipe.

The 3/4" opening at the top is for the feed from the tank(filter) and was plugged for the test. It will be fitted with a barb later.

The nozzle assembly is inserted from the end with the 3/4" "T" and screwed down to place the opening of the nozzle jet just inside the 3/4" opening on the other end.

attachment.php


attachment.php


attachment.php


It needs to be just at the opening to leave room for the water that has to flow around it and into the exit pipe.

Then I fit a 3/4" pipe on the end and a 1" barb and hose on the bottom for the test.

attachment.php


I filled a 5 gal. bucket and attached the venturi to the garden hose on my well pump. When I turned on the water the jet of water produced good suction on the 1" hose that left a noticeable red ring on my arm. When I inserted the end of the hose into the bucket the venturi lifted the water ~ 2.5 feet with no problem and emptied the bucket in rather short order. I didn't have my stopwatch at the time but I estimate somewhere in the vicinity of 30 seconds give or take.

I'll post an accurate flow measurement soon.

Awaiting more funds to move on.

Pat
 

pat w

Member
Nov 4, 2009
462
0
16
Daphne, AL (east Mobile Bay)
Venturi Flow Data

Flow data:

5 gallons (orange utillity bucket from Home Depot full to rim)

Head - 0 (I placed the venturi on the ground next to the bucket)

Time - 20.3 seconds to the point the one inch hose began to suck air at the bottom of the bucket

Flow = (5gal/20.3 seconds)*(3600 seconds/hr) = 887 GPH

To this add the primary flow from the garden hose that induced the venturi flow and I should have ~ 1600 GPH flow in the saturation loop that will be about 1 - 2 liters volume tops. That should give me a turnover rate in the sat loop of 1.5 loops/second. {estimate}

Pat
 
Last edited by a moderator:

pat w

Member
Nov 4, 2009
462
0
16
Daphne, AL (east Mobile Bay)
I had a spare minute so I added the gas injection tube. It should use the venturi vacuum to draw in either CO2 or room air.

attachment.php


With the resounding success of the venturi I'm a little worried the draw will be stronger than I expected and I may need to restrict the flow into the tube somehow.

Pat
 

pat w

Member
Nov 4, 2009
462
0
16
Daphne, AL (east Mobile Bay)
All that work and today I find these:

473-825.jpg
10-4305.jpg


PVC-Venturi-T's.

I've decided to redesign using them and adding a second venturi. It'll be upstream of the other and will draw in gas bubbles and feed the downstream venturi a mix, call it a foam, which could reduce the the draw of CO2 in hopes of reaching an equilibrium. The compression of the foam in the venturi should produce a higher solution rate, too. The upstream venturi would be sent a solid fluid stream and therefor would have stronger draw. That would be the best entry point for the feed from the tank or filter.

Just curious, there seems to be more than a little interest but very little feedback. Just wondering.

Pat
 

Tug

Lifetime Charter Member
Lifetime Member
Jan 5, 2009
1,150
9
38
Washington, DC
So's not to interrupt your genus?

Hi pat,
Wouldn't it be better to introduce oxygen (air) to point B, 24/7 if possible?
pat w;57212 said:
3.It should also assist in the balance of night time O2 levels.
-----
 
Last edited by a moderator:

pat w

Member
Nov 4, 2009
462
0
16
Daphne, AL (east Mobile Bay)
Tug;58107 said:
Hi pat,
Wouldn't it be better to introduce oxygen (air) to point B, 24/7 if possible?

Hi back,

You're probably right. The best draw will always be near the venturi orifice and introducing O2 with the CO2 shouldn't have a serious negative affect on the solution rates of either. I just don't know how the nitrogen is going to affect things. I have the sinking feeling there will be huge undesolved nitrogen content in the thing after the first night that would throw everything off.

Here's the redesign with the new venturi-T's and a side port 5-way cross from the same place.

attachment.php


The pump drives the first venturi with a solid fluid column and draws a mix of fluid and air and drives the second venturi with the resulting foam mix which draws the CO2 in at the top. A second intro point could be added easily to include the air feed 24/7. The tank/filter inlet at the 5-way would use the first venturi’s draw to keep that flow consistent. The back pressure from the first venturi would help drive the output to the flow meter.

Oh … and if there was any genius involved, I’d be able to come up with a way to pay for the thing.

Edit: Exerp from a PM to GerryD on his suggestion

As far as following the thread, I know I've come up a little short describing it all. I'll try again here. You might want to get a snack and something to drink ... this is long.

Design Objectives:
1. Create a system that allows the dosing of CO2 as a flow of CO2 saturated water; a kind of CO2 stock solution in hopes of reaping some major benefits.
1A. Fluid flow is easier to control than gas flow and controlling that flow can easily be metered with rotometers. Gas rotometers for the gas flow rates we have are prohibitively expensive.
1B. Since the design is dependant on heavily saturating water with CO2 most of the failures will result in a reduction of CO2 being delivered to the tank thereby reducing the potential of gassing fish.
1C. A liquid can be more easily introduced into the existing filtration or tank circulation system for even distribution.
2. Make the system resistant if not immune to the evils of some of the less expensive and less precise components such as end of tank regulator dumps, cheap, flakey, hard to adjust needle valves to name two.
3. Provide a means to accurately measure in real world units the amount of CO2 being used to assist in assessing the impact of changes and to provide some feedback on utilization to assist in refill scheduling.
4. The system should have large tank capacity. Hopefully 500g+ in the base design and upscalabe with larger venturis and pumps.
5. Do all this to the extent possible by relying more on principles of physics and less on ever more costly apparatus.
6. Make it as compact and user friendly as possible.

Principles of Operation

Venturis: The venturi is commonly used in the Spa industry to induce air flow into spa jets by forcing a jet of water into a slightly larger enclosed flow-path. The jet forces the contents of the flow-path to move downstream forming a small vacuum in the cavity behind. Fluid or gas or a mixture of both can be pumped into and through the venturi cavity into the down stream plumbing. For spas the induced flow is exclusively air and the motivating flow is exclusively water. The two are not mixed till they enter the downstream plumbing.

My hope is to use a venturi, and at first drawing water only, drive a second venturi down stream. Then, by way a small access point, CO2 can be introduced into the flow at the vacuum point of the second venturi. The flow is then looped back to supply the induced flow feeds to both venturis. This may be the point that throws folks. It is at this point that the return water for the pump is separated and allowed to drop into a downward moving current at a nearly 10:1 lesser velocity to allow any bubbles not caught in the venturi action to float back to the intersecting feed point. By keeping the CO2 bubbles in a constant rapid circulating flow and occasionally subjecting them to the added pressure of the second venturi orifice the CO2 saturation levels should be elevated well above anything yet experienced using existing reactor designs.

Now that I have what I hope will in effect be seltzer water, introduce an external supply of water fresh from the filter or from the tank itself into the low pressure point at the venturi feed intersection. Use the back pressure of the first venturi to feed a valve equipped rotometer(edit:The rotometer is conected between the Outlet to Tank above and the tank http://www.gogenlab.com/products/product/F4320-visi-float-flowmeter-4-scale). The rotometer will be the metering device that is used to control the flow of the CO2 saturated water back to the tank thereby controlling the CO2 levels there. A clearly visible indication of flow rate will replace the subjective use of counting bubbles.

CO2 Delivery:
Excessive gas in the system will effectively kill the intended operation of the second venturi. In order to prevent this the system must be able to be self-regulating in its CO2 supply. I intend to employ the vacuum of the second venturi and the small pressure that exists as a result of the elevated tank water. As I stated above the system operation is first described as a fluid only system. With no gas in the system the transfer curve for the venturi indicates somewhere in the vicinity of -4 psi vacuum when powered by a solid fluid stream. What no one has been able to tell me is what happens to the transfer curve as a mixture of fluid and gas is passed through the orifice or what variations in the relative concentrations of each will do. I'm expecting the vacuum generated to be reduced by as much as 2 psi. down to -2 psi.

Now to the other side of things. Follow me for a moment. Place on the floor a small low container with a bottom side mounted return tube to reach up to and into the tank as an expansion line. Place two gas tubes on the top, and mount a float switch inside at the top with the wires extending out of the container(it's how it's made). Attach one tube through some kind of flow control (a simple air line valve should do) to the CO2 solenoid on the regulator. The other tube is attached to a "T" and one side of that to a flow valve as a bleed device; the other side of the "T" is attached to the CO2 input of the reactor THROUGH an adjustable check valve with the high pressure side to the container. Open the bleed valve and start a siphon to begin filling the container. When the water reaches the bleed valve close it. The container is now completely full of water with an expansion line back to the tank. The CO2 solenoid is wired through the float switch in the container. When the solenoid is powered the CO2 will feed into the container and displace water back into the tank. The displaced water will eventually drop low enough to allow the float switch to open stopping the flow of CO2 with about 2" or so at the top of the container. Since the container is connected to the tank via the expansion line the CO2 will be under a slight pressure. In my tank that will be between around 1.55 psi and 1.8 psi depending on the level of water in the tank (this number allows for a 4" drop in tank level that I never allow to occur but I thought it prudent to make the provision). Whew; ... all that to get a positive 1.75ish pressure on the high side of the check valve.

Now some simple arithmetic. 1.8 to 1.55 psi on one side of the check valve; -4 to ... lets say -3 psi at the venturi of the reactor. that gives a range of 5.8 down to 4.55 psi across the check valve. Set the cracking pressure to about 5 or 5.25 psi and as the CO2 is pushed into solution in the reactor the check valve should cycle and replenish the supply. As the CO2 is delivered to the reactor the water in the container will rise eventually allowing the float switch to close to replace the CO2 in the container.

Well that's all for now. See if you can figure out how I satisfy item number 3 in the objectives?
*******************
Thanks again to Gerry for the interest and getting me to lay the whole thing out a little better.
 
Last edited by a moderator:

pat w

Member
Nov 4, 2009
462
0
16
Daphne, AL (east Mobile Bay)
I've found this adjustable check valve for $20.50 + shipping. (1/4 MPT,Brass)

3-50 psi cracking pressure.
Here's the idea: set the cracking pressure to ... let's say 25 psi ... then set the regulator outlet pressure to around 15 psi. When the venturi draws a vacuum of more than -10 psi then the check valve will open and feed the reactor. After the CO2 content of the "foam" is high enough to reduce the vacuum back down to the reseat level, the valve closes until the CO2 dissolves and the vacuum goes back up.

Thoughts?

Pat