Diy Pvc Overflow Conundrum

[Maehlice]

I'm trying to redesign the overflow system on my aquarium and am failing miserably.  I need some help.
 
 
DIY PVC overflows don't work (well, at least not at any decent flow rate), so I built an HOB overflow box, which has so far been working great:  I pump water in, it overflows out, and the cycle repeats.
 
The 'problem' with the overflow system, though, is that it draws water from the top.
 
I want to draw water from the bottom.  My hope is that this will reduce (if not eliminate) the need for detritus vacuuming, which has become an all-too-frequent occurrence with my 16" pleco.
 
Outside of mechanically pumping the water out, a siphon is the only option I know.  I'm using a 3/4" PVC pipe as the siphon right now and hope for a flow of 500+ GPH.
 
The issue with a siphon is that if the return pump fails or the siphon draws out water faster than the pump can return it, the aquarium empties itself onto the floor.
 
So my plan was brilliant (or so I thought).  I would overflow the water into an intermediate container deeper and taller than the aquarium.  In this case, I used a 2" PVC column fitted with a tee at the top to serve as an overflow.
 
As the column of water overflows, the siphon draws out water from the main aquarium.
 
THE PROBLEM:  The intermediate container won't fill up fast enough.  It empties as quickly as water enters it, but it just isn't siphoning in fast enough.  This is the problem I need to solve.
 
...
 
I've since broken this thing down, so I lack a photo.  The diagram below shows what I last had.  The water siphons out of the main tank (red/orange) into a 2" PVC column (green) and then out of its top into the refugium (blue).
 
That system only flowed at about 120 GPH.  If I remove the 2" PVC column of water and let the water siphon out into a container, it flows somewhere around 500-600 GPH, which seems about right according to this data.
 
 
 
THE QUESTION:  What about that column of water (green) is preventing the siphon (red/orange) from reaching its full flow rate?
 
My only thought is that the downward flow of the siphon is being restricted by the upward flow of the column (orange vs green).
 
If this is true, then my question becomes:
 

1. "Just how large must the column of water be to eliminate enough resistance?"
    
2. Or, what if I create a "U" at the bottom of the siphon inside the column so that the siphon's flow is already in the same direction as that of the overflowing column?
    
3. Or, am I just approaching this the wrong way or seeking something impossible?

 
 
 
THANKS FOR ANY ADVICE!
 
 
 

 

[DrRob]

Fascinating. I'm understanding, I think, what you're trying to do.
 
The area that strikes me as the slowing down point is the section of 2 inch pipe with the syphon tube down it. You'll have surface friction there on both the inside of the outer pipe and the outside of the inner pipe, just after the water has to turn 180 degrees. My suspicion is that you're got highly turbulent flow in there resulting in a lot of flow loss.
 
Creating a U at the bottom of the syphon will just add another 180 degree tight bend although it could, in theory, help if it makes the flow more laminar.
 
Is there any reason that you can't simply create an overflow loop, possibly plumbing the syphon tube into the bottom of the green section via a reducing bush and your 180 degree bit? Thereby removing the friction component, actually increasing the available volume and keeping the idea you're looking at.

 

[Maehlice]

The area that strikes me as the slowing down point is the section of 2 inch pipe with the syphon tube down it. You'll have surface friction there on both the inside of the outer pipe and the outside of the inner pipe, just after the water has to turn 180 degrees. My suspicion is that you're got highly turbulent flow in there resulting in a lot of flow loss.

 
I had completely disregarded friction.  The opposing flows coupled with friction makes a lot more sense to me now.
 

Is there any reason that you can't simply create an overflow loop, possibly plumbing the syphon tube into the bottom of the green section via a reducing bush and your 180 degree bit? Thereby removing the friction component, actually increasing the available volume and keeping the idea you're looking at.

 
Do you mean something like this (see attachment below -- figure 1 at left)?
 
(fig.1)

        (fig.2)

 
I had tried that already, also.  It resulted in a negligible increase -- if any.  I use a one-gallon jug and mental count to track the flow, so the increase could have been my imagination or a slow count.  It certainly wasn't anything to call home over and definitely not what I need.
 
The other thing I tried is figure 2 at right (except imagine a 2nd vertical green line at far right).  By putting the output inline with the flow like this, I was able to increase the flow more than with the bushing idea but still not to the level I need it.
 
Looking back again at the friction and turbulence, I think I see why it still didn't work.  I think the left column is still causing a sort of stagnant drag preventing the water from picking up any great speed.
 
I think I might know how to solve the problem or at least increase the flow some more.  However, I'm running out of PVC and connections.  I keep a lot on hand, but not like this.  

 I have ABS, though.  Is ABS "aquarium safe"?
 
Anyhow, my next idea looks like this (see below).
 

 
I've not yet done this, and I'm kinda reluctant to now without some feedback.  I'm also not sure I can handle the disappointment when this doesn't work.
 
And if it does work, I'm not sure I can even handle the disappointment of how "unsleek" this has become.  My first idea was quite sleek and discrete.  This is getting ridiculous.
 
I don't have anything larger than 2" PVC and ABS, so my experimentation limits are about to be met.  If my budget permitted, I'd go buy 6" PVC and re-test the first idea.  Unfortunately, I have no other use for 6" PVC, so if the idea failed, I'd be completely out of that money.
 
 
Again, thank you for all your input and advice.

 

[Maehlice]

To my utter surprise and dismay, using a considerably larger 6" PVC column with the original design resulted in MUCH LOWER FLOW of about 80 GPH !?!?  

 
I even tried using three 45" elbows to angle the water flow up towards the surface.  No change.
 
I'm convinced now a column of water is not going to work here.  

   The water can't effectively flow.  There's too much drag created by the "dead zone" at the bottom and too much turbulence in-between.
 
Since I've already spent enough time and money on this, I think I'm going to see it through to the end and try the "loop" design.  I think if the water is permitted to flow around in a loop, it might actually work.  Given my track record thus far, I know it won't, but I might as well give it a go if only to learn the outcome.
 
Because a larger column decreased the flow, I think drag is the major culprit.  If the spirit moves me, I think I may try using a smaller 1 1/2" PVC column just to see the effect.
 
 
At the very least, I've succeeded in building a system with fail-safes to siphon 120 GPH from the bottom of my tank.  While this was not the success for which I hoped, it is at least some modicum of success.

 

[Maehlice]

For those of you still interested enough to have read this far, this is what I hope to be one of two final updates.
 
Yes, two ... which indicates it still isn't working.
 
 
To recap, I broke down and bought 6" PVC!  

  No matter what i tried, It didn't work using any combination of elbows and overflows and heights (but no loop).  The flow was no greater than before.  Again, I think it may have been worse.
 
For giggles, I also ran a 3/4" siphon into a 1 1/2" column, which also had no effect.  The flow didn't even reduce.  

 
To confirm I wasn't going stir crazy, I tested (again) the siphon outside the column into a 5 gallon bucket.  As usual, the flow started off slow, but then BOOM!  The flow was so fast, I barely had time to put down my timer before the bucket was overflowing.  So, I've at least confirmed (again) the siphon itself is not at fault.
 
I then went on to test (again) the flow rate just of filling up a 6" column of water with the 3/4" siphon.  Instead of starting with the column full, I started with only 1 gallon (it holds 5 gallons also).  
 
It. Took.  Forever.  
 
Unlike in the 5 gallon bucket, the siphon never really got going -- thus proving to me the cause must be turbulence resisting the flow of the siphon.
 
 
So I think the key here is a very wide column of water (12" or so in my case).  Based on the bucket test, a 15 gallon drum (14" x 26") should do the trick.  However, all I have around here are 35 and 55 gallon drums.
 
When I get back from holiday next week, my final test will be into a 35 gallon drum.  Based on my bucket test, I have no doubt it'll finally work.
 
I can hardly wait.

 

[cornclose]

I will start my brief response by stating that I haven't read *all* of the above in detail, but would just like to point out that I suspect the syphon rate is going to be governed by the 'head of water' from the level of the syphon entry point (the main tank) to the syphon exit point (the sump/bucket etc).
 
It's probably why when creating all the other contraptions that you get a much reduced flow. The differential level between the syphon entry point (the main tank) to the syphon exit point (the sump/bucket etc) will govern the flow rate (along with the diameter of the syphon tube of course). The bigger the differential in level, the more 'head' of water there is and the faster the flow rate. So I suspect that whatever you try to do you will be disappointed unless you can maintain that head of water.
 
Its the same fundamental scenario with a return pump - the higher a pump has to send the water to return it to the main tank, the slower the flow rate will be.
 
I got around the unreliable syphon/overflow systems on my setup by using a syphon from the main tank down to my low level sump. However, I added two 24VDC solenoid valves in series within the syphon tube. These valves are powered to open, so if I lose power, they both close. I put two in series simply to avoid a potential situation where if one doesn't seat properly and jams slightly open, it would overflow my sump. Given the MTTF figures for the valves and multiplying those up for two devices in series, I get an astronomically low likelihood of ever having my sump overflow upon a power failure.
 
I also have level switches in my sump which control not only the syphon solenoid valves (allowing water into the sump) but also my return pump feeding back to the main tank. The level switches are also wired in a fail safe configuration. Once I set the return flow rate to match the syphone rate, we're all done and we have a fail safe system with litle or no pump/valve cyling.

 

[cornclose]

Ok having read the above more thoroughly I will ask one question; When you are testing the syphon into a 5 gallon bucket, what is the distance between the top surface of the tank water to that of the surface of the water in the bucket that fills ?
 
The reason I ask is that on all of the above images, the difference in level between the surface of the tank water (blue) and the surface of the 'contraption' (lets call it that..) water (green) is next to nothing!
 
You will never get good flow whilst this differential level is so small, no matter what you do with pipe/column diameters etc - the same fundamental problem exists with them all. At least that's how I'm seeing it for now, so my initial post above stands!
 
Let us know what happens next.

 

[Maehlice]

The reason I ask is that on all of the above images, the difference in level between the surface of the tank water (blue) and the surface of the 'contraption' (lets call it that..) water (green) is next to nothing!
 
You will never get good flow whilst this differential level is so small, no matter what you do with pipe/column diameters etc - the same fundamental problem exists with them all. At least that's how I'm seeing it for now, so my initial post above stands!
 
Let us know what happens next.

 
I suspect the same, but when testing into the contraption (it really is just that

) vs the 5 gallon bucket the initial reaction of the siphon is wildly different.  Both the bucket and the contraption's lowest point are the floor and started from nearly empty.  The only difference at that point is the diameter of each container.
 
I'm not expecting my next test to work, but I am at least hopeful.
 
Testing should commence tonight or tomorrow, so I will surely have news for you this weekend.  Stay tuned!

 

[cornclose]

It's not the lower level of the siphon that matters though, it's the upper level of the water. Granted, the contraption and the bucket may be at the same level when the siphon starts, but for the same unit volume of water that leaves the siphon, the level in the contraption will increase at a much greater rate than it would in the 5 gall. bucket, thus the head in the contraption will reduce much more rapidly and impede the flow accordingly.

 

[cornclose]

Any news ? The suspense is killing me!

 

[cornclose]

What happened ? It's been a while...

 

[Maehlice]

I'm sorry for the delay.  I had to disappear for a while.  At least for now, I'm back again.  I'll try to give warning should I disappear again.
 
It's exactly as you said -- "it's the upper level of the water".
 
Regardless of the size of the contraption, it starts off with a high flow and then diminishes to a trickle once the water level within it matches that of the aquarium.
 
 
 
What I did instead this weekend was make a DIY canister filter out of two 5 gallon buckets.  One was cut in half, swiss-cheese-drilled on the bottom and filled with rocks and pebbles and filter floss.  Finally, I tossed a couple refillable carbon filters on top and capped it.
 
Water is pulled through the system with one of my Mag Drive 500 GPH pumps out the top via bulkhead.  The water enters through the lid (via bulkhead) and is piped through to the bottom (via another bulkhead in the swiss cheese bucket).  The water then rises through the filter media out the top through the aforementioned bulkhead/pump back into the aquarium.
 
This way, I can pump and filter water off the dirty aquarium bottom as I originally wanted -- just not as low tech as I was hoping.