Sihponing rate and return rate

[desm0nds]

Hi

There has been many posting on ideas, suggestions and product in the forum. Of which the DIY sump tank attracted me most. Personally I have a 3ft marine tank setup. And after reading all those, I am motivated to build a sump tank for myself.

Problem is, I am no engineering expect in the field of fluid flowrate though I am an in the engineering field of electronics. Therefore I would need many advice from you people out there. Be it an engineering expert of the reefkeeper with many years of experiences. You will be very helpful.

To begin with, I have attached a picture of my thoughts. I have not built the sump tank yet so lets just assume its a typical kind of sump setup. i.e. 3 compartments with filter element, deep sandbed and a return pump.

Siphoning will takes place from the main tank to the sump tank. There should be 2 intakes. i.e. a surface skimmer intake and a bottom intake. Siphoning will only stop when water level falls below the T-joint between the 2 intakes. This has been tested and proven.

All pipes will be 13mm diameter. 90 degree joint and T-joint will be used.

A return pump will be used to channel the water back to the main tank. The pump has a rating of 2200 litre/hour at max ht of 1.95meter.

Will this system be in balanced? Or will the siphoning rate be greater than the return pump flowrate? In either case, the main tank or the sump tank will be overflowed.

If this is not in balance, how can I achieve a balance system? i.e what is the formula to be used to calculate for the correct pump rate?

Some posting said that "whatever goes in must comes out". Is this statement true in the sense that the siphoning rate will be dependent on the return pump rate?

There are some posting that uses double U in teh piping to achieve different flowrate. i.e. the diff between the output and the water level of the main tank. The greater the difference, the faster the flowrate. But then again, what will be the formula used to calculate this height difference?

Hopefully there can be answer to these. I do believe that many of us out there are having the same questions as well.

Thanks in advance.

[thc]

Check this out RC

Make use of the calculator, it should help

[desm0nds]

Thanks thc. The website is useful but it doesn't really answer to my question. It does give me the true flowrate of the return pump based on the number of angle joints, diameter of the pipe used, the total vertical and horizontal length used and the power of the pump used. However it doesnt calculate the true flowrate of the siphoning path. All it says for a typical GPH drainage, wat will be the recommened pipe size. It doesnt takes into consideration the number of joints and the distant.

Do you have any other ways to find out?

[porc]

There are some posting that uses double U in teh piping to achieve different flowrate. i.e. the diff between the output and the water level of the main tank. The greater the difference, the faster the flowrate. But then again, what will be the formula used to calculate this height difference?

haha! I had a long argument with a fellow bro here on this. you must have read that. my explainations and examples were as clear as I could make it... I still maintain that what goes in will come out and is independant of the height difference. well, you can choose to believe him or me...

your current design is based on a full siphon on the return pipes. IMO not a good design because it is very difficult to tune and the "what goes in = what comes out" rule doesn't apply. For the rule to apply you just need to add a few more bends in the piping. See my diagram below. Similar to yours except no surface skimming and the additional bends. the blue part shows the water and in this case explains why what goes in will come out. hope this helps..

[desm0nds]

Thks porc. A diagram speaks a thousand words. In the daigram, you have indicated L1, L2 and L3. I pressume that L3 is the water intake. L2 will be the minimum water level when lower than this, siphoning will break. And L1 above L2 will be dependent on the Return pump rate. i.e. Amount of water back to the main tank. The faster the water get pumped back into the main tank, the faster will be the flow into L3. Can I say that?

What about the surface skimming part? How can I still coporate a surface skimming action into the design?

[Eric]

Hi

There has been many posting on ideas, suggestions and product in the forum. Of which the DIY sump tank attracted me most. Personally I have a 3ft marine tank setup. And after reading all those, I am motivated to build a sump tank for myself.

Problem is, I am no engineering expect in the field of fluid flowrate though I am an in the engineering field of electronics. Therefore I would need many advice from you people out there. Be it an engineering expert of the reefkeeper with many years of experiences. You will be very helpful.

To begin with, I have attached a picture of my thoughts. I have not built the sump tank yet so lets just assume its a typical kind of sump setup. i.e. 3 compartments with filter element, deep sandbed and a return pump.

Siphoning will takes place from the main tank to the sump tank. There should be 2 intakes. i.e. a surface skimmer intake and a bottom intake. Siphoning will only stop when water level falls below the T-joint between the 2 intakes. This has been tested and proven.

All pipes will be 13mm diameter. 90 degree joint and T-joint will be used.

A return pump will be used to channel the water back to the main tank. The pump has a rating of 2200 litre/hour at max ht of 1.95meter.

Will this system be in balanced? Or will the siphoning rate be greater than the return pump flowrate? In either case, the main tank or the sump tank will be overflowed.

If this is not in balance, how can I achieve a balance system? i.e what is the formula to be used to calculate for the correct pump rate?

Some posting said that "whatever goes in must comes out". Is this statement true in the sense that the siphoning rate will be dependent on the return pump rate?

There are some posting that uses double U in teh piping to achieve different flowrate. i.e. the diff between the output and the water level of the main tank. The greater the difference, the faster the flowrate. But then again, what will be the formula used to calculate this height difference?

Hopefully there can be answer to these. I do believe that many of us out there are having the same questions as well.

Thanks in advance.

Just install one ball valve for the return pump discharge and one ball valve for the siphoning.

Return valve open full then play ard with the siphon valve to balance the flow.

If you want to upgrade to a more powerful return pump for higher circulation rate, just install a T joint with another ball valve to control excess flow back to sump compartment.

p/s: I suggest you use a 5000l/hr and above return pump instead of the 2000l/hr cos after minus off the head loss, you dont have much flow left in the tank. The higher the cycle rate, the cleaner your water is due to higher amount of filteration.

I've seen a drastic improvement in my water quality after changing from a 2500l/hr Eheim to a 5500l/hr Hailea return pump. In progress of upgrading the return pump to 2 x 6000l/hr right now & Teeing off any excess flow to my chiller and protein skimmer.

[desm0nds]

Eric, I did think of incorperating a ball valve. In fact this afternoon, I snipped down to King George Ave to get myself 2 ball valves. This would be my final resort if all else failed and I didnt know if this method will work?

Have you tried it already?

You mentioned about using a 5500l/hr powerhead for the return line. My setup is currently only 13inch pipe. Can the pipe takes that kind of high speed flowrate?

One more question. Since you are having high speed return path into your main tank, do you still have got or still need a powerhead inside your main tank for water movement?

[desm0nds]

Apologies. I did mentioned I am using 13inch pipe. This is a mistake. A pipe such big in size would be crazy. It should be 13mm diameter.

[Eric]

Eric, I did think of incorperating a ball valve. In fact this afternoon, I snipped down to King George Ave to get myself 2 ball valves. This would be my final resort if all else failed and I didnt know if this method will work?

Have you tried it already?

You mentioned about using a 5500l/hr powerhead for the return line. My setup is currently only 13inch pipe. Can the pipe takes that kind of high speed flowrate?

One more question. Since you are having high speed return path into your main tank, do you still have got or still need a powerhead inside your main tank for water movement?

I'm using the ball valves method to control the sump level.......24mm siphon to sump with 20mm main return & 13mm chiller return to tank.

Powerhead is for water circulation inside the main tank, Return pump flowrate is the tank volume cycle rate. Both is as important.

Imagine you only have PH inside the tank and its creating good water movement, but then the water quality will be the same, its just circulating inside the tank. If you have a high cycle rate, the water will undergo more frequent filteration and skimming thus cleaner water will be always inside your tank circulating.

13mm should be alrite, if you scare the flowrate too high, divert the return into two lines blowing from each corner. That way you even save on powerheads. Hehe.

[desm0nds]

Thanks Eric for sharing. You mentioned that if the flowrate is too high, I can divert the return path into two lines. Do I then have to have 2 line of siphoning line to the sump tank? Like the theory of "What goes in must comes out", if I have 2 input and 1 output from main tank, either the main tank will overflow or the sump tank will get dried up. Isn't this true?

[RockyBoy]

13mm pipes are too small for pipes from main tank to sump. Would be noisy. Better get at least 25mm.

[desm0nds]

I am afraid you are right, Rockyboy. But I have got only 13mm elbow, coupler, pipes etc. So probably I would like to stick to 13mm. Perhaps I could have 2 siphoning paths and then use Eric's suggestion of a 5500l/hr return pump over 2 return path.

Or for simplicity, I should just resort to getting the 25mm parts and pipe. But I have got a question. Do I have to use a 13mm to 25mm reducer at the return pump? Cos I believe the pump outlet will be only 13mm?

Any suggestions of which method will be better?

[aStRoBo|]

13mm pipes are too small for pipes from main tank to sump. Would be noisy. Better get at least 25mm.

wad about for return pipe if he were to use 25mm pipe from tank to sump?

[RockyBoy]

If I'm using 5,500l/h, I'll use 32mm pipe liao or 2 x 25mm pipe. I prefer to have unobstruct flow of water to the sump.

My pipes from overflow initially is 20mm (short distance), then increased to 25mm. My water flow is 2500l/h.

For your return pump, u can use 13mm for 2,200l/h or 20mm for 5,500l/s

[desm0nds]

a bit confused now. Pardon me.

Rockyboy, so you are saying that for 55ool/hr, the recommended diameter for the siphoning pipe is 32mm while for return path is 20mm.

Why is there such a difference? In such case, will the system still be in balance? i.e. a bigger pipe will channel more water than a smaller pipe.

wat if I were to use 25mm pipe for both siphoning and return path?

[Eric]

a bit confused now. Pardon me.

Rockyboy, so you are saying that for 55ool/hr, the recommended diameter for the siphoning pipe is 32mm while for return path is 20mm.

Why is there such a difference? In such case, will the system still be in balance? i.e. a bigger pipe will channel more water than a smaller pipe.

wat if I were to use 25mm pipe for both siphoning and return path?

In short desmond, the Siphon (25mm) to sump piping should be bigger than the Return (13-20mm) sump to tank. For the return pump connectors, you can always play around with the reducers.

You dont have to worry abt material balancing cos you have the T-joint to channel excess flow and ball valves to regulate.

Oh ya, remember to put a check valve at the return pump discharge to prevent any back flow due to electrical failure or you can just put the return piping nearer to the water surface so the backflow will be limit to your sump capacity. I would suggest using a check valve though.

[desm0nds]

Ok so probably I could summarised as follows:

There will be one siphoning path of 25mm down to the sump. At the end there will be a ball valve.

On the 5500l/hr return pump, 13mm pipe will be used whereby a T-joint is used to divert any access flow back into the sump which is controlled by another ball valve.

Instead of using a check valve which I heard it will be "sticky" after long usage, I would use the hole method to break unwanted siphoning at the return path end.

Well, it does sounds workable to me.

Btw Eric, how much does a 5500l/hr pump cost? Where can I get for a good price? Any good shop that you can recommend?

[porc]

Sorry for the late reply... was busy last 2 days... from your discussions it seems you are currently considering something like your original design. You can try that out but here are my 2 cents more...

1) too many parameters are being controlled. your pipe diameter, no. of elbows, pipe length, pump capacity, water level, height of input and output (ie. pressure head), valve opening etc all affects the frictional force and therefore the flowrate, which you are hoping to control. you will end up with a system with a narrow stability band. My proposal is actually an overdesign in terms of flowrate using larger pipes etc but more failsafe. The only parameter here to consider is the allowance for L1-L2 (see details below).

2) therefore tuning may not be easy. but I must also qualify here that I have no real hands-on experience on these systems (as I find the design not up to my own Hazops standards to be implemented and tested), but I do understand such systems theoretically and they are prone to instability outside of the equilibrium state especially if they are introduced abruptly(eg. you stick your hand in and increase the water level by a fraction of an inch, or some debris clogging your pipes etc). However some reefers here have done it with no problems so it's not impossible.

3) One of my more highly probable hazop scenarios would be power failure (ie. loss of flow in return pipe) followed by power recovery (resumption of flow). after loss of flow your tank will not empty (good!) because you have drilled your hole. but the hole will also cause a break in the siphon (bad!). when power returns your overflow is not primed and you know what happens next... if you have a system that solves this problem then good.

As for your earlier questions I'll answer them since you've asked...

Thks porc. A diagram speaks a thousand words. In the daigram, you have indicated L1, L2 and L3. I pressume that L3 is the water intake.

Correct!

L2 will be the minimum water level when lower than this, siphoning will break.

In a way you are correct. but based on normal usage water level will not fall below L2. Assuming your return pump fails, your water level will fall from L1 to L2 and stop there. The water (ie. siphon) in the 1st bend will still be there and ready to continue when your pump kicks back in.

And L1 above L2 will be dependent on the Return pump rate. i.e. Amount of water back to the main tank. The faster the water get pumped back into the main tank, the faster will be the flow into L3. Can I say that?

Absolutely correct. most people have trouble understanding this part. It is precisely this water level difference here (and not the outlet) that is pressuring the water thru the pipe. The higher the pump pump rate, or the smaller the pipe size, the more allowance you need for L1 above L2.

What about the surface skimming part? How can I still coporate a surface skimming action into the design?

Yes. you can incorporate a surface skimmer in exactly the same way you designed your original piping in your first post. I just don't have it here because the diagram was drawn much earlier and I just reused it.

There are other ways if incorporating a surface skimmer, but there are more also considerations if you want a mid-tank input. We can discuss more if you are interested.

cheers,

porc

[desm0nds]

Hey porc, u r back!

Yup, we have had some discussion going on here on other way of design.

No deny that so far, all method should works and the only difference will be the constraints.

If I were to incorperate a surface skimmer as in my original design into your design, what then would be the required distant between L1 and L2. Assuming that 25mm piping will be used for the siphoning path and 13mm piping for the return path powered by a 5500l/hr.

Last nite, I did a test with your design on 13mm pipe for siphoning and return path over a 2200l/hr return pump. I just could get the distant between L1 and L2 right. Somehow or rather, either the water level in main tank raised much above the surface skimmer inlet before the water level stablise or the water level falls below the surface skimmer inlet. The water level will just now stay at the desired level above the inlet. What could have gone wrong?

At the same time, I also tried out the ball valve method. Well, it does works but it takes quite an amount of time to get the flow to stablise.

Maybe I should be incorperate both ideas into 1. With Porc's double U design and Eric's ball valve design. In this way, I could still restart the siphoning after pump recovery from failure and at the same time, achieve a system balance by double controlling the flowrate with the ball valve.

Having the best of the both wolrd.

[porc]

Hey porc, u r back!

Yup, we have had some discussion going on here on other way of design.

No deny that so far, all method should works and the only difference will be the constraints.

If I were to incorperate a surface skimmer as in my original design into your design, what then would be the required distant between L1 and L2. Assuming that 25mm piping will be used for the siphoning path and 13mm piping for the return path powered by a 5500l/hr.

Last nite, I did a test with your design on 13mm pipe for siphoning and return path over a 2200l/hr return pump. I just could get the distant between L1 and L2 right. Somehow or rather, either the water level in main tank raised much above the surface skimmer inlet before the water level stablise or the water level falls below the surface skimmer inlet. The water level will just now stay at the desired level above the inlet. What could have gone wrong?

I assume you mean that you adjusted L2 right? L1 would then naturally equilibrate itself.

Firstly, I think 13mm siphon piping is not big enough. imagine on one hand your powerful pump is pumping in 2200l/hr through a similar 13mm piping, and on the other hand we want a natural equilibrated water level difference (L1-L2) to push water at the same rate. I would expect the the level difference required to be quite big. 25mm piping should be good to handle the large flowrate with a minimal L1-L2 difference.

Secondly, I suspect your design is slightly complicated by your wanting for a bottom inlet in addition to the surface skimming. If L1 is too high L2 will correspondingly rise above the skimmer, utilizing both inputs but eliminates the skimming action. whereas when L1 is too low the water can still use only the bottom inlet alone and thus L2 is below the skimmer inlet..

I really cannot think of a good solution to this... but if you evaluate your current tank situation and think that there is already enough circulation in there anyway then maybe a bottom inlet is not neccessary and you can use the following design explained in the next post which I have successfully tried.

Maybe I should be incorperate both ideas into 1. With Porc's double U design and Eric's ball valve design. In this way, I could still restart the siphoning after pump recovery from failure and at the same time, achieve a system balance by double controlling the flowrate with the ball valve.

Having the best of the both wolrd.

hmmm... where are you going to place the valve? If you place after second bend don't forget that there is an open top there. whatever you restrict will just overflow from the open top. If you place the valve in the first bend itself, you are effectively just adding more friction thus requiring a larger L1-L2 level difference... that's what I can analyse for now... well, if you do try it out let us know the results...

[porc]

This is the design that I'm using now. The only difference is in the addition of a cylindrical pipe of larger size than the overflow pipes. In this case L1, L2 and L3 is equivalent to those in my earlier design.

during operation, water level in the tank will be at L5 which is slightly (abt 1cm) above L4 (which is the length of the cylindrical pipe) and overflowing into the cylinder. The input to the overflow is placed into this cylinder and is actually drawing water from this cylinder. Similar to before, L1-L2 is the water level difference driving the flow through the pipes. Adjust L2 such that L1 is as close to L4 as possible. so that there will be minimal bubbles from water splashing into the cylinder. make L3 as low as possible so that bubbles will not get into the pipes. bubbles accumulating in the pipes would break the siphon sooner or later.

when the return pump stops, water level in the tank will drop only to L4, whereas water level in the cylinder will drop to L1. siphon is still maintained. when the return pump regains function everything will continue working as normal...

BTW, a disclaimer here... i don't take credit for these designs, which are based on simple principles. many other people have thought of these designs too and have successfully implemented it. whenever i say "my" design i simply mean the design that I'm using...

[desm0nds]

hmmm... where are you going to place the valve? If you place after second bend don't forget that there is an open top there. whatever you restrict will just overflow from the open top.

Had wanted to ask you, the opened end top. What's it function? I guess its for any unwanted air bubbles that is created during the siphoning process to escape? But by having an opened end at that position, won't it create a noisy "draining" or whistling sound?

but if you evaluate your current tank situation and think that there is already enough circulation in there anyway then maybe a bottom inlet is not neccessary

The bottom inlet is meant to suck up any debris which is being lifted by the water current. Currently I am only using a 1600l/hr powerhead to achieve the water circulation. So I dun think that the current will be strong enough to lift those bottom debris to the top so as to be sucked in by the skimmer inlet.

What do you reckon?

Btw, what is your recommended return rate for a 3ft tank? I would like to gather more feedback on this. Understand that 5500l/hr is quite common but due to budget constraint, well... sigh.

[delirium]

Would like to share my tank setup. Main tank is 3x1x1.5' and sump is DIY standard 2' tank.

From top of the tank to sump is 3'. Using 12mm piping for siphoning and i think 8 or 10mm for return. Cant remember. No powerhead in tank as the return is split into 2.

Back of tank has 2 holes drilled. 1 hole is cover up at the top half and it siphons the water while the other hole takes in air while it skims the surface. Return pump is stated 2500 l/hr. 1 return pipe goes down half way into the tank while the other is set just under the water level. So when the pump goes off, siphon breaks automatically.

[porc]

Had wanted to ask you, the opened end top. What's it function? I guess its for any unwanted air bubbles that is created during the siphoning process to escape? But by having an opened end at that position, won't it create a noisy "draining" or whistling    sound?

the open end top is to allow air in sspecifically to break the siphon in the second bend, otherwise it's a full siphon again and no different from your original design... it is actually part of a durso design that can actually help reduce noise. you add an end-cap there, drill a hole and add a small air-valve to control the air intake. supposed to reduce the slurping sound but i find that it makes no difference for my case so i just leave it open... check out http://www.DursoStandpipes.com

The bottom inlet is meant to suck up any debris which is being lifted by the water current. Currently I am only using a 1600l/hr powerhead to achieve the water circulation. So I dun think that the current will be strong enough to lift those bottom debris to the top so as to be sucked in by the skimmer inlet.

What do you reckon?

do you have a fine sandbed? if your circulation is so strong to stir up debris wouldn't it stir up the sandbed too. don't know how well it would work. I have no experience on this. i normally just clean my sandbed while doing water changes...

Btw, what is your recommended return rate for a 3ft tank? I would like to gather more feedback on this. Understand that 5500l/hr is quite common but due to budget constraint, well... sigh.

hey, i believe that before this you tank is already cycled and has been working without a sump right, which actually means that filtration has already been going on successfully in your main tank itself already. IMO a sump is just additional filtration or more space for a refugium, skimmer and other equipment etc so flowrate is not too crucial. I myself am using just a recycled big boy overhead filter pump. rated only 1200l/h but definitely much less with the 1/2 m height it is pumping up...

sometimes people tend to overkill on technology. I believe sometimes by simple logical reasoning you can realise that some "recommendations" are unneccessary...

[porc]

some picts to share. forgive the quality and mess...

overall view... the bends are contorted in this way simply to conserve space...