I usually consider myself a master googler™ but I'm having trouble finding very much information on gravity fed protein skimmers.

It would seem to me that the pump attached to most skimmers could easily be replaced with a feed off of the overflow from the display tank. Is there something I'm missing? I'm planning on a 40b with a 200gph fuge loop. I've pretty much convinced myself I need to skim (as much as I don't like the added equipment), and I don't remember where I first ran across things, but here is some of what I found.

http://www.reefworkshop.com/DIY_gravityassistpump.htm
http://www.nano-reef.com/featured/?tank=11

My reasoning here is that I would like to keep my heat input into the system as low as possible for summer months and this would also help do away with a lot of the bubble problem in fuges.

Am I nuts?

You aren't nuts, but this kind of setup is typically reserved for recirc skimmers. Usually the flow rate from the return is a lot higher than you are looking at currently, and I don't think you want to run your entire tank return through the skimmer. Maybe I'm wrong. the heat added to a system via a skimmer pump is really minimal. Skimmer pumps are typically a lot lower in power draw than water pumps.

i would use more than 200 gph, and i would use some sort of baffle at the end so it would work simular to a aqua c skimmer. otherwise you could add a airstone to increase bubble production. if there isnt enough bubbles in a skimmer the skimate wont be extracted as it will mix back in with the water. after seeing what little skimate is produced from having the overflow water crash into the sump i dont see you collecting much without some additional form of bubble production. probly not enough to be worth your time when you can get a cheap used skimmer for $60 or less, or even a seaclone that is absolute crap would produce more and people give those away.

i think you worry to much about adding heat to the tank. a smaller pump wont add much if any heat to a tank. you could add a cheap remora or superskimmer and i dont thing you would increase the water temp any noticable amount. certonly not enough that couldnt be offset by using a small fan bloing across the top of the sump. if your worried about bubbles either design you sump correctly with baffles to control bubbles or set the flow correctly.

I've read a lot about air stones not being worth the added hassle, but I was sure I read that a skimmer's turnover rate should be between 2 and 4 times tank volume, which would be about 140 to 280 gph. Is this not correct? Main display tank will have about 600gph on a closed loop. I don't have a problem with spending for a good skimmer. I'd rather keep the maintenance and worry level lower.

The only real issue I saw after reading your post and evil's is with my plans for the 20h. While I haven't decided on a species, I want to keep something interesting, but likely more fragile in the display refugium. This means keeping the flow rate around 5-10x turnover, and keeping the equipment intrusion minimal (keeping the sump part of the fuge in the back corners and small). I think I finally resolved all of that on the way home.

If a larger skimmer is really the prescription, I could mount the intake low in the overflow chamber of the fuge (keeping the wall low enough to drain any excess into the main chamber), run up to a 400gph external skimmer (kept behind the fuge), and split the output of the skimmer between the main chamber and the intake for the return pump.

So final questions... 1) Do I really need a 400gph skimmer? 2) Do you know of anything with a setup like I'm talking about so I can see if it's possible? 3) Did I ramble into total incomprehension?

no u cant replace the pump on a in sump skimmer. the pump makes the bubbles and puts water into the skimmer. u lose the pump u lose the bubbles. no airstone will be able to make the bubbles a pump does.


i used to have a recirc skimmer hooked to my drain. the drain line will need to be split and one side goes into skimmer the other into sump. i had about 300 gph throught the sump with no problems. with a recirc skimmer the pump makes all the bubbles so no need to try and add more. if u have problems with not enough water going into the skimmer u will need to gate valve mod it. this will let u run lowwer gph through the skimmer but it will back up the water inside to create more contact time.

Let me see if I can use my awesome powerpoint skills to illustrate better.

Front and top view. I can skim 400gph with this design and only push 200gph through the refugium if I need to. Everything is external so it takes up very little room.

You make it sound in this post that you are not running a sump. Is this correct? If you aren't, you should. It will solve almost all your issues, especially when it comes to keeping equipment out of the display. I think before we go too much further with the skimmer discussion, you should explain how you tank is going to be set up in detail.

Yeah, apparently I'm sucking at this whole explanation thing. I think that's from using pictures to describe things at work all the time. Sadly, it appears that I can't even do that right as I left out the split output of the skimmer.

Two displays running on alternate light cycles. One a heavily planted refugium, and the other a more typical reef tank. My original fuge loop would look just like this but for the addition of the skimmer. The last idea I posted about looks just like this with the addition of a split output from the skimmer - one output to the fuge and the other to the top tank's intake. This is all within the confines I've got (wife demanded the stand match the rest of our furniture). You can think of it as a sump with a really big fuge.

Red under positive pressure and blue under suction. This does not include the closed loop. I just roughed this up in the last 10 minutes at work.

im confused about your tank.

lemme get this straight...the water drains into the right compartment (looking at it head on). the skimmer draws water from the same compartment and exits to the one on the left. where the return pump draws water from the left compartment and pumps it up through the red pipe?

this seems awfully complicated for something so simple. and if you plan on making your tank and sump like that, theres a HUGE chance it might overflow and you will have problems with micro bubbles in the display.

Like I said, I flubbed a bit but didn't realize it until I left work. Earlier in the thread it was recommended that I do more than 200gph of skimming, but at the same time I don't want to do more than 200gph through the fuge portion of the lower tank. The part I forgot to model in was the split in the skimmer output - one end going to the fuge, and one end going to the return chamber. I;d include valves to push ~200gph through the fuge and the remainder straight back to the top tank.

Is the output of a fuge full of bubbles? I was under the impression it was pretty good bubbleless water. If that's not the case I definitely need to change things. I'm still not convinced I need more than 200gph of skimming.

Overflowing isn't an issue. If the return pump turns off, I might get about 2 gallons of drainage (.5" drop from the main tank and another half gallon from the plumbing). I'd simply set the water level in the fuge to less than about 16 gallons total (12 inches high in this case). I didn't model any of the baffles (or all) at the correct heights, just a rough in. I'll try to finish it with dimensions tomorrow.

I really think I'm just confusing everyone at this point. What I've morphed my plan into is your basic sump design with everything normally kept in the sump plumbed in externally. The only real difference is the scale of the chambers an that they are at a 45° in the corners.

the only reason i say its going to have micro bubble problems is your bound to have micro bubbles coming out from the skimmer. and you have the output feed directly into the return chamber without going through any baffles.

another thing is when the water from the fuge flowing into the return chamber, with your pipe suspended in the middle of the chamber instead near the bottom, it will suck in some bubbles created from the water crashing down into that chamber.

also, im not really sure about having the return pump suck and pull the water at the same time. this to me just creates extra head for the pump to overcome.

can/will cause the pump to cavatate. (spelling) this is bad move the pump down shorten the pipe leading into the pump as much as posable.

I didn't think a couple of feet of head on the suction side would create a problem. It shouldn't create cavitation (requires MASSIVE pressure drop), but it could potentially cause some problems with the throughput of the pump. I'll go buy my target pump and see what kind of throughput I can get with the pump at different heights.

I'm glad you guys are bringing all of this up. It's difficult to get really solid answers. Too many forum posts to go through. Definitely won't split the output of the skimmer in that manner. Using what you guys have told me, I'll try to fix up the model tomorrow to see if it looks any better.

These pumps are not designed for suction work. They are not self priming. You would have to have the pump lower than the waterline in the sump.