skimmer bar overflow questions, please help

[cheezischrist]

hello all, i am pulling my hair out trying to find answers to these questions myself, and figured i would just ask already. please any imput would be extremely appreciated.i am basically copying and altering this guys skimmer bar setup. i really like this idea, but am having a hard time figuring out a few things.
http://www.youtube.com/watch?v=WHn7_ZqZljs&feature=plcp
its a standard aga 75g tank, so bottom drilling is out of the question, and ive really grown attached to this type of setup, but i will be altering it. i still plan on using 1" pvc for the overflow bar, and 1" bulkheads, and the two end holes will be 6" down from the top of tank, and 4.5" in from each side, but i also wish to include a 3rd hole in the middle, and use a "t" instead of the top elbow and connect the 2 overflow bars together. hopefully increasing rigidity, and failsafing should one of the drains clog. and will also be doing the 1/8" slot cut across the top of the tubes, allowing 1" from each fitting uncut. this should leave me roughly 32" total of surface skimming. and i will also be doing the 2 returns, with split locline, similar to his single one, centered between each drainpipe. each overflow will have its own drain pipe also, 2 going to the first chamber with my skimmer in my sump, and one going to the 3rd refugium chamber. te return will be in the middle section of the sump. for my sump i will also be using a 30 long aquarium i had laying around. there will eventually be a second 30gallon macro/ xenia and anthelia refugium directly connected to the main tank, but thats a future addition.now on to my questions.
1) approximately how many gph's will this overflow be draining? been rackin my brain withh calculators and formulas trying to get an answer and just dont really understand them. and should i move up to 1.5" pvc for the overflow instead? or keep it 1" on the inside and bump the outside up to 1.5"
2)can i drill 5 1 3/4" holes evenly across the back of this tank, or should i lower the returns to increase stability? and if so, how much. since ill be using split loclin, im not too concerned with their placement.
3)the slots on top of the skimmer bar, should i bump them up to 1/4? and if so, will that increase the drain gph?


thank you in advance for any input or answers here. my sump and return plumbing are going to be much different then his in te video, im just more confused and concerned with the overflow questions i have now. still a long way from this actual build.

[cheezischrist]

noone?

[nolken]

you should be able to handle up to 1500 gph given you actually have at least 23" of linear overflow. 1" pvc should be fine with 2 1" overflow drains. if you do add a third it will just be a fail safe if the others get clogged. What i would do is put it together inside the tank, don't cement it (who cares if it leaks inside the tank), and try 1". If that doesn't work you lost an hour of time and maybe 10 bucks in fittings. It'll be easy since you didn't cement it.

I'm sure the 5 holes can be done and work just fine, given that the tank has a decent glass thickness. the tightly installed bulkheads probably provide plenty of support to take care of the support lost from drilling the holes. What is the thickness of the glass?

definitely go 1/4 in. 1/8 will cause way to much drag. keep in mind you will need 32 in of pipe overflow length, giving you 64 in from the total of both sides. divide that in half because of the slits and subtract about 20% for tension drag and you got about 25 in of linear overflow, just a little over the 23 in the calculator said you need for 1500 gph.

[cheezischrist]

wow, thank you, not only for an answer, but for an in depth one. i was just waiting for the bashes i usually get on my ideas. . and i didnt plan on cementing anyting in tank, for toroug cleaning processes. anything outside the bulkheads would definately be glued. and just to reassure myself, it would be better to have all 3, so i had a failsafe should one or 2 get clogged correct? the slots and outputs into sump would be part of the weekly maintnence. im not sure of the width of the glass, but its pretty thick and heavy. probably at least 1/4". i think im going to stagger the holes just for my own piece of mind, i did have a tank blow out before, and 90 gallons going through 3 floors was not pretty
thank you though for answers.

[Ted_C]

thats funny - I'm also copying his design - on a 48" 120 gallon tank I had already planned the same - two 90's and one tee in the middle.

The guy in the you tube video actually gave two values for his pipe slit width: 1/8" and 1/10" (in a later video). His return pump is only producing about 300 gph I think (maybe less)

I'm the commenter asking about how he cut the pipe in the first video

I cant get to the pics of my drilled tank from here at work - But I will try to remember to post a reply when home this evening.

[RocketEngineer]

From personal experience I recommend you DON'T use this setup. I used this design on my 75g setup and have had to run an MD5 at half its flow rate or run the risk of overflowing my tank. The problem with this design is that the flow falling from the slot into the pipe is only about a 1/2" drop. The result for me was the center of the pipe was full to the slot while the ends would surface skim. If I were to do it over, I would go with a Coast-to-Coast internal overflow with an external box for the drains. Its a little more involved but it takes up less space in the tank while letting me use a BeanAnimal drain setup.

RocketEngineer

[Ted_C]

One other idea I had that's a split from your outline and the youtube guy - instead of all three drains going directly to the sump separately, I was going to take all three 1" drains down to a 1.5 (or larger)" horizontal pipe (with 1/4" fall for every foot of run) - the majority of the output for this would go to my skimmer compartment in the sump and a T with a gate valve would control the output to my refugium (my skimmer compartment will be on the right of my stand, the refugium will be to the left of the stand - with an extra large return compartment in the middle.

My return will be a 1262 eheim and I'm expecting to get around 800+ gph to the Tee where I will have it split into two returns (and split inside the tank again with locline fittings to 4 outlets).

I'm a week or two out from plumbing - so I'll definitely be posting some pics on this design.

The good thing about this design - if it doesn't work out - like in RocketEngineer's example - it can definitely be adapted to a coast to coast with a beananimal overflow. Your coast to coast will need to be a little deeper than usual (since your drains are 6" the coast to coast will need to be something like 7" deep so you can turn the 90 degree elbow down) and the drains wont be right next to eachother like in BeanAnimal's example but spread out over the length of the tank.

[leeit2me]

I agree with RocketEngineer. I had this setup like yours with my acrylic 60gal. I would have to dial down my return pump by half. It was scarey when I had it full blast and slowly watching the water level rise and rise and the slotted overflow couldn't keep up. It does a great job at surface skimmer, but just not enough to handle a simple Rio 1700 power head return pump. So, I went with a coast to coast and am a happy camper now.

[nolken]

You know its funny cause what RocketEngineer said is exactly what I was thinking would happen. the water wont flow through the pipe fast enough. since it's a flat surface it won't flow like it would on a slant.

one idea to solve this problem would be to take a heat gun to the center and ends of the pvc overflow and form it so it has a slight slant and cut level slits across the water line (the slits in the middle would be deeper than the slits on the end). I think that would work but sounds like a royal pain to do.

+1 on the coast to coast. original idea was cool, but the coast to coast is a tried and true fail safe idea. It may work better with 3 drains though. i'm anxious for you to try it but remember once you drill the holes there's no going back.

[uncleof6]

Originally Posted by nolken

You know its funny cause what RocketEngineer said is exactly what I was thinking would happen. the water wont flow through the pipe fast enough. since it's a flat surface it won't flow like it would on a slant.

True, however this is not why this contraption does not work well at all.

one idea to solve this problem would be to take a heat gun to the center and ends of the pvc overflow and form it so it has a slight slant and cut level slits across the water line (the slits in the middle would be deeper than the slits on the end). I think that would work but sounds like a royal pain to do.

+1 on the coast to coast. original idea was cool, but the coast to coast is a tried and true fail safe idea. It may work better with 3 drains though. i'm anxious for you to try it but remember once you drill the holes there's no going back.

To do any surface skimming, the slot has to be at the water surface. This means there will be air in the drain lines, for one--as the "weir" is actually the drain intake. With air in the drain line, you can't put as much water through it. Hence, the need for a low flow rate. Secondly, the slot is not the same as a "weir," technically it is, but it is too narrow on the downstream side to handle any significant flow without backing up. Which leads to the water level rising, diminishing surface skimming, floods, as the system tries to reach siphon. The physics are complicated, again requiring a lower flow rate.

The design is a poorly conceived notion, to eliminate the need for an internal overflow--"box style"--with separate elbows. A space saver, if you will, that does not work as advertised. But just like the "pvc overflow," which is a rather poorly conceived notion as well--in terms of real world performance, this is a waste of time.

Cut a 4" pvc pipe in half, and stick it on the back of the tank, and plumb out of it, and you would be getting closer.....but what an unsightly mess that would be. NO i don't recommend doing that either--2 - 3" internal overflow box, and put the plumbing in a box outside the tank, with several good size holes for pass through.

[nolken]

With it being slanted the ends would be under water so no air could get into the drain lines, and the bottom of all the slits would be at water level. and yes it would work better as a square tube so the ends wouldn't be narrow. I didn't care go into details because i thought it would be exactly as you mentioned a complete waste of time for the amount of performance and space you would actually save. My only point was that something "could" be done to make a design like that work. It's not worth thinking about in too much detail

[uncleof6]

Originally Posted by nolken

With it being slanted the ends would be under water so no air could get into the drain lines, and the bottom of all the slits would be at water level. and yes it would work better as a square tube so the ends wouldn't be narrow. I didn't care go into details because i thought it would be exactly as you mentioned a complete waste of time for the amount of performance and space you would actually save. My only point was that something "could" be done to make a design like that work. It's not worth thinking about in too much detail

With the "overflow" high enough in the tank, to perform any surface skimming, the system will suck in air, whether the ends are submerged (slanted) or not. If the hole thing is submerged--and no air enters, then we deal with a siphon--that will not perform any surface skimming. This cannot be fixed. It is inherent in any drain inlet, expected to perform surface skimming. The long pipe, or a single upturned pvc elbow, behave the same way.

[bubba3085]

I was toying with this idea as well and this is what I came up with to deal with the issue of the surface skimmer only being capable of handling a low amount of flow. Using a siphon in conjunction with the surface skimmer and hooking up a ball valve to control the amount of flow left for the surface skimmer to handle. Granted I have no clue what I am doing and am trying to figure out where to go just as the OP. Any feedback is appreciated.

[uncleof6]

Unless the slit is submerged by better than an inch, this design will suck in air, and a siphon will not form, and you are back to square one. if the slit is submerged, it will not do surface skimming. Not any point to experimenting with this design.

Why on earth would you want to match your flow rate to the flow rate of the skimmer? (what the skimmer can process.) There is absolutely no connection between flow rate through the sump, and skimmer performance. Perhaps the simplest concept involved here to understand, is that regardless of flow rate through the sump, more water is going to pass the skimmer by, than will be processed by the skimmer.

[bubba3085]

Thanks; if you look at that picture the bulkhead on the right would be the siphon. The red line I drew there is a piece of PVC sheet cut into a circle plugging the tee so it is in essence a 90 with support for the skimmer bar. The skimmer bar would do just that; skim the surface at whatever flow rate was quiet. As you stated maybe 50-100 gph. The siphon bulkhead would handle rest of the flow from the return pump maybe 250-300 gph. The flow through the sump would be around 400 gph.

[uncleof6]

Originally Posted by bubba3085

Thanks; if you look at that picture the bulkhead on the right would be the siphon. The red line I drew there is a piece of PVC sheet cut into a circle plugging the tee so it is in essence a 90 with support for the skimmer bar. The skimmer bar would do just that; skim the surface at whatever flow rate was quiet. As you stated maybe 50-100 gph. The siphon bulkhead would handle rest of the flow from the return pump maybe 250-300 gph. The flow through the sump would be around 400 gph.

I think I follow you here. However, it is not safe to run a siphon without a dry emergency in case the siphon plugs, or it will flood your house. The skimmer bar is not an emergency back up. Again, do it right the first time, and save yourself the aggravation. Install an overflow box as already discussed

[bubba3085]

I'll have to entertain that as I'd rather not answer to the wife as to the flood in our living room. Thanks for your perspective uncle.

[nolken]

maybe i'm sounding dumb but how exactly is the air entering the drain pipe?

[nolken]

I mean i see why the top one may not work, but why not the bottom? the drain will drain as fast as it needs to whether there is air in the drain or not. there is no need to create a siphon as we are not trying to raise water higher than the water level. as long as the drain pipes are big enough to handle the water flow air shouldn't matter. as the pipes require more flow they will fill with water pushing the air out. Our house drains always have air in them. air will only effect a siphon which isn't needed since he is drilling through the tank.

[nolken]

I mean i see why the top one may not work, but why not the bottom? the drain will drain as fast as it needs to whether there is air in the drain or not. there is no need to create a siphon as we are not trying to raise water higher than the water level. as long as the drain pipes are big enough to handle the water flow air shouldn't matter. as the pipes require more flow they will fill with water pushing the air out. Our house drains always have air in them. air will only effect a siphon which isn't needed since he is drilling through the tank.

[uncleof6]

Originally Posted by nolken

I mean i see why the top one may not work, but why not the bottom? the drain will drain as fast as it needs to whether there is air in the drain or not. there is no need to create a siphon as we are not trying to raise water higher than the water level. as long as the drain pipes are big enough to handle the water flow air shouldn't matter. as the pipes require more flow they will fill with water pushing the air out. Our house drains always have air in them. air will only effect a siphon which isn't needed since he is drilling through the tank.

It isn't rocket science. If you wish to surface skim the water, the slit or whatever mod you make will need to be at the surface, and for this setup to function, as reiterated several times (by more than one person,) the flow rate must be such that air does enter the system. Even at full siphon, unless it is ~ couple inches deep in the water, it will suck air--which is not very accurate, because with air in the line, a siphon will not form. Air entering the device at any point, will: yep you got it--introduce air into the system. It cannot be avoided. Try as you will, however you are not going to beat the physics.

The "durso standpipe" is a prime example of the physics at work--other than those involved with flow over a "weir" and this contraption includes those physics as well. It is a low flow device--and was well relegated to the junk pile quite some time ago.

The logic error you are making is that air only affects a siphon. This is not so. It is the air that makes the durso such a pita. The only difference between this and a durso, is where the air enters. The slit does not handle the flow rate to keep the air out. whether it is angled or not--unless it is submerged to the point that it siphons, rather than acts as a weir or "waterfall". At this setting, there is no surface skimming.

The bottom line is: build something that works--and works well, out of the box.

[nolken]

so are you saying that the air is entering the pipe but not escaping because of the constant inrush of water through the slits? my thought is that the design is the same as the coast to coast design, the only difference is the size/depth of the box and the location of the drain. the main difference that i'm guessing would be causing that air is the thickness of the pipe. maybe you're thinking the pipe is "closed off" with no open top? i understand that if the slits are too shallow the water won't flow in fast enough and the water surface will go above the slits which will in turn create a system with no air but would not perform any surface skimming in a closed pipe. just like a pipe with simple holes drilled in it.

I think now i understand 100% what you are saying because i'm coming up with a way where what you're saying is in fact a problem. i think i may not be explaining what is in my head 100% accurately, and i don't feel like doing a detailed autocad drawing. the square "pipe" would be more like an overflow than a pipe, just a really small overflow the size of a pipe

[Ted_C]

Air enters the system in two places - one at the skimmer bar and three times at the three capped vents behind the tank (two for the you tube guy). As long as the skimmer bar slit doesn't restrict the flow too much - a siphon should never be reached. A siphon is not the end goal for this design - rather it's to have the water sliding down the outsides of the pipe with plenty of air.

Now on to some calculations for the skimmer bar slit. The you tube guy figured out that he needed 5.5 inches of slit to drain 280 gph (the maximum flow rate for his return). Assuming he meant a skimmer bar slit for 1/8" - that's .6875 sq inches per 280 gph.

If it's a linear equation (which I suspect it is) - then 1500 gph would require a slit with 3.68 sq inches.

That would be a 1/8" slit with a total length of 30 inches. By my calculations in a 48" long tank with three holes cut into the back of the tank - subtract 6" for the two end holes, subtract 4 inches for the 2 90's and one tee, and subtract another 4 inches for the 1" of non-cut pipe on each end (for rigidity) - that leaves 34" of available skimmer bar slit divided by two pipes (with 4.25 sq inches of skimmer bar slit). That should support and fully surface skim 2000 GPH.

[nolken]

Originally Posted by Ted_C

Air enters the system in two places - one at the skimmer bar and three times at the three capped vents behind the tank (two for the you tube guy). As long as the skimmer bar slit doesn't restrict the flow too much - a siphon should never be reached. A siphon is not the end goal for this design - rather it's to have the water sliding down the outsides of the pipe with plenty of air.

That is my thoughts exactly.

[RocketEngineer]

nolken and Ted_C,

The problem with this design is the water does not have enough vertical drop from the edge of the weir to the outlet pipes. In simple terms, the volume downstream of the overflow edge is insufficient to allow enough vertical drop to supply the outlets with high volumes of water. What you get is a pileup in the horizontal pipe that puts an upper limit on the flow rate down the outlets.

Scientific answer:
Flow rate is determined by the height drop between the inlet and outlet. In our case, this height drop is applied along a horizontal pipe and must therefore be provided by the water level upstream rising in relation to the outlet height. The problem with the slotted horizontal pipe is that you reach a point along the pipe where the outlets are so far from the inlet edge that the water level necessary to push water towards the outlet equals the height of the inlet. At this point, the water level in the horizontal pipe IS THE SAME AS THE TANK. No surface skimming can be achieved because water has totally filled that portion of the overflow pipe.

Put another way, in order to achieve surface skimming along the entire length of the horizontal pipe, the horizontal pipes must be sufficiently short as to keep the water level in the horizontal pipe below the lip of the slot in the pipe. In order to achieve high flow rates, multiple outlets would be necessary with additional outlets needed for each increase in flow.

In contrast, a simple glass box with three 1"drain pipes (a BeanAnimal setup) can handle a huge range of flow rates, is simple to adjustable to match the pump, fail safe and achieves surface skimming equal to or greater than a narrow slot in a pipe.

uncleof6 and I aren't trying to pull your leg. We really do understand the physics involved and are trying to help. In my case, I built a system like this and learned first hand IT DOESN'T WORK. One of the reasons my 75g has algae problems is a lack of flow to the sump which limited how much water the skimmer and refugium could process. My 125g will fix that problem.

If you don't believe us, then go ahead and build it the way you want. Check the quote in my signature and you will see I do understand about trying something rather than taking someone's advice.

Good Luck,
RocketEngineer