Sump basics

pat w

Member
Nov 4, 2009
462
0
16
Daphne, AL (east Mobile Bay)
@Tim - I have to thank you for the discussion. It has given me ideas to improve my sump insert design. I've been basing all my thinking on the hodgepodge rubbermaid kludge setup I threw together after my canister mishap. Now, thanks to some of your questions, I've re-thought the design and modified to include a separate return pump chamber of only 2 gallons. This will limit the amount of water exchanged between the sump and the tank during a power loss or an accidental loss of Overflow siphon. This will allow for a higher water level in the tank and not be concerned with the volume pumped back to the tank in the event of siphon loss. Power loss issues are still managed by the adjusting the overflow weir level till the OF stops and the return siphon breaks kick in.

Here's a 3D model of the design in and out of a 10 gallon leader tank.


Sump insert w-tank by pat w1, on Flickr


Sump insert by pat w1, on Flickr

You could easily drop the same boxes in a 20 long, increase/decrease the size of the pump chamber to suit individual needs, etc.. I have the acrylic and a cut plan. Next step is to get some MDF and cut out some router patterns.

Pat
 
Last edited by a moderator:

shoggoth43

Lifetime Charter Member
Lifetime Member
Jan 15, 2009
1,092
11
38
The small pump chamber has a pro/con and it's the same thing. It's a small chamber the pump sits in that's baffled off from the rest of the sump. Pro: if the overflow fails you can only pump what's in the chamber up to the main tank. If you've done this right, you won't be able to raise the water level in the main tank above the rim and flood. It's likely you'll damage the pump although a float operated switch might be used to prevent this. Cons: you will get flow fluctuations from the pump as the water level rises and falls and the head pressure above the pump changes. Higher water level will give more flow. It's a smaller chamber so evaporation will change this level quickly. In what might be a 20 gallon sump, you may only have a few gallons available. This may not get you through a long weekend if you go away.

I've always set the overflow height to some distance above the bottom of the upper tank trim. Maybe 1/2 to 1" just so I don't see the waterline or have any ripples / waves big enough to show. To set this I drain the tank down somewhat and then setup the overflow. I then fill the tank up until the overflow kicks in and fills the sump. Once the level is high enough to prime the pump I start it and continue to fill until I get the desired level in the sump and stop filling.

Siphon break holes are the better choice vs. check valves. You can drill a couple in ascending order. You can also just run the return barely under the surface and as the water siphons back in the siphon will break when the end clears the surface. Check valves FAIL. They need maintenance. So you'll need two valves, a check valve and another valve to close off the water flow while you work on the check valve ( for those who run the return up through the bottom of the tank ). Did I mention that check valves fail? No, that's not the voice of experience or anything... :rolleyes:

Coast to coast overflows are very nice, but if you aren't running a lot of flow, you may find that they do a poor job of skimming debris off the top as the amount of flow over the top will be too shallow. Better surface skimming will be had from an overflow with no teeth, but you will have a "thinner" water layer going over the top. Simple eggcrate light diffuser material can be used on the inside of the skimmer as teeth to keep larger floating plants out.

If you have drilled the back wall of the tank for a bulkhead your overflow box need not be overly tall or deep or even box shaped. A simple pane of glass slanted outwards to 2" gap at the top should suffice. Putting the bulkhead for an overflow box near the top mitigates the risk of a box/bulkhead failure. The tank will drain only to the lower edge of the hole in the back. I'll let you do the math on how much water can come out if the box fails and you drain to the bottom of a bulkhead 4" down vs. a standpipe in the bottom of the tank. If the tank cracks because you put the bulkhead too close to a corner then you've got other issues.

Rounding the inside edge of the overflow box will help water pour in more easily and helps reduce splashing and noise. Where possible you will want to run at least two lines from the overflow to the sump. One will be blocked slightly by a gate valve so that there is always some amount of water above it. This is the main line to the sump and as there is never any air allowed to slurp into the line, it will be silent and you will not inject bubbles into the sump and this will help retain CO2. The second line is a standpipe a little taller than the first. You will adjust the valve on the first line so that water just BARELY spills over into this. Just a tiny trickle. This is your safety line. Any clogging of the main line will have all the water flowing into this one. If that happens you will get a noisy overflow so you know to investigate. This is the Herbie design. The Bean design will add a third standpipe.

You CAN easily do the Herbie design with a hang on back siphon type overflow if it was designed for two hoses. If you do this, it is absolutely critical that you make sure the water level in the overflow box in the back is much lower than the level in the tank. Siphon speed and flow is very dependent on the height difference. If you start a siphon in the tank and hold the other end over a bucket, you will drain the tank much faster if the end of the tube is down in the bucket vs. holding it just barely lower than the tank level. The same applies here. Make sure you've got a few inches of drop and you won't have to worry about your overflow slowing down.

Siphon overflows are workable. You want to make sure the flow through the tube is as fast as possible to make sure bubbles don't have a chance to form or build up. This may require getting a smaller U tube, or switching to two smaller ones or whatever. Don't be afraid to experiment. Alternately, skip the potential problem with the siphon breaking in the U tube and have the tank drilled for a bulkhead. Gravity will trump the siphon U tube every time and will provide much higher flow rates. That said, I run a siphon overflow with the Herbie design on my big tank without any real concerns or noise. It will eventually be drilled but I'm in no hurry.

-
S
 

tjbuege

Lifetime Charter Member
Lifetime Member
Jun 25, 2009
119
0
16
Minnesota, USA
Soggoth43,

Thank you for the lengthy reply. Lots to think about. And lots of questions keep coming to mind... :)

How much water is really needed in a sump? What is the desired level? Does the volume of water have any bearing on the effectiveness of filtration? I would think the biological (wet/dry) portion, along with some sort of foam filter between the wet/dry and the main sump reservoir are the important things. What I'm thinking is, the sump water level is probably not going to be running anywhere near the top of the sump tank. If it's too deep, then it will flow back into the wet/dry area and it will no longer be wet/dry, but just plain wet. :)

Also, how much does the water pressure (i.e. level of water in the sump) above the return pump really come into play? If the return pump is sitting in 10" of water, or if it's sitting in 5" of water (due to evaporation, for example), is that really going to have a big impact on the flow of that return pump? I'm thinking the distance from the pump output to the actual aquarium will have the most impact on flow, as that is the greater distance, and it's on the output side of the pump. And it's a fixed distance. Of course, I could be completely overlooking something here, so please correct me if I'm wrong.

Evaporation is a variable factor in this, I know. My tank will be 36x18 (assuming outside dimensions), so one inch of water is about 2.6-2.7 gallons, depending on glass thickness. Not sure how long it takes for that much to evaporate in my house, but I understand that it will drop the level in the sump faster than it would in the aquarium if I were running a closed canister system. An inch drop in the aquarium equates to about a 2 3/4" drop in the sump. Right now, my 20g long (30x12) loses about 1" of water every 2-3 days. That leads me to believe with the larger tank I'm planning, I'll probably lose 2-3" of water in the sump every couple days. Is it normal to have to constantly top off the water in the sump to keep it running? Is this a daily routine? It's another consideration in my decision to go with a sump. I know I could probably plumb in some sort of auto top-off, but I'm not ready to get that complex yet.

I'm also not willing to drill holes in an aquarium at this point, so I would opt for an external overflow box. Coast to Coast overflows... that refers to a built-in style of overflow that spans the entire length of the back wall of the tank, right? And requires drilled bulkheads? It sounds interesting, and I understand the benifits of that design, but again, I don't want to drill right now. :)

Good tip on rounding the inside edge of the overflow box lip. Makes complete sense. I bet that has other applications in the aquarium, too.

Interesting about the two output lines. I think I'm starting to understand the basis of these silent overflow designs. I'll have to reread them now that you gave a good explanation on how they work. If I purchase an overflow box (as I probably will this first time around), the models with two drain lines are quite pricey. Is the two line design critical in silent operation, or is it possible to run silent with a single return line? Also, do you run the two separate lines all the way to the wet/dry portion of the sump, or do you combine them at some point above that? I suppose rereading some of these articles will answer that question.

Regarding your comment about overflow siphon and water level in the overflow box, I'm struggling a bit to understand this. If the water level in the back box is lower than that in the front box, and if you shut off the pump, wouldn't the water levels want to equalize, dropping the water level in the front box, breaking the siphon? Or are both ends of the U tube the same length, and the part inside the tank is just in deeper water? Having a hard time picturing this one. Regardless, if I purchase an overflow box, it's design should be such that this isn't an issue, right?

Speaking of commercially-made overflow boxes, I'm still wondering which brands are recommended? CPR, Eshoppes, Tom Aquatics, Lifereef? From what I can see: CPR doesn't use a siphon tube, so trapped bubbles could be a problem; Tom Aquatics looks like it has extra "stuff" involved (not sure what I'm seeing in the pictures); Eshopps PF-1000 has two outputs, uses a siphon tube, and won't break the bank (about $72); Lifereef get's high reviews, and uses a siphon tube, but their double drain model is expensive... $200. Right now, Eshoppes looks to be the best value, but if it's junk, I need to know that.

Thanks for the discussion and the ideas!
 

shoggoth43

Lifetime Charter Member
Lifetime Member
Jan 15, 2009
1,092
11
38
I'll try to address the other stuff in a bit but:

For a prefilter you can either put one in the overflow, or put one in the sump. Many reefers don't bother and just dump into their skimmer chamber and/or prefer to just use their refugium as a settling chamber. For planted tanks this is also an option as you could easily throw some floating stems or other plants in there and use a reverse lighting scheme to allow for some "filtration" at all times. None of these plants need be submerged. Regular houseplants can suffice as long as they'll grow. Japanese Peace Lilies are often used for this. Any food/debris that accumulates would just settle to the bottom and you could use this as a grow out chamber for your cuttings, shrimp, etc. Those that escape will likely end up sucked into the pump and blasted into the main tank to become food for fish, although I doubt a full grown cherry shrimp would make it past the impeller and to the tank in one piece.

Prefilters by way of sponges or socks are another option. Tom has stated that he thought socks outgassed too much CO2. I've been trying to design a method where the sock remains submerged to reduce any splashing which might outgas CO2 as most sock holders hold the sock out of the water somewhat and water pours down into it. I would see this as a problem but less of an issue with a reef where you generally don't want CO2 to hang around. A side benefit to the sock is that any fish that make their way into the sump/prefilter would be easy enough to rescue without risking them being injured or pinned which seems to happen in the overflow prefilters despite my efforts.

Usually the hose from the overflow runs right into the biotower and the wet/dry media. Sometimes there is a prefilter chamber ahead of this. Neither of these methods is required. Many reefers will Tee off their return line and run a portion of their outflow to their refugium. There is no reason why you couldn't dump the prefiltered chamber directly into the pump chamber and then run a portion from the pump to the wet/dry biotower. A side benefit is that if you shut down flow to the tank, you keep your filter running so as to avoid die off should your maintenance run for a longer period of time. Some people feel that it is more efficient to run the pump without restriction and this method allows you to run any "extra" flow you don't want in the tank back into the filter. Some pumps do run more efficiently, cooler, and quieter doing this. Submersible pumps are excellent for this as you only need to throw it into the sump and keep it under water and you will have minimal restriction. A strainer over the pump inlet of the typical black basket can restrict flow by a surprising amount. In some cases 50% or more. If you use one, go up a size. Again, you could run a prefilter, biotower, and refugium ( planter tray or whatever ) all off the same pump/sump.

If you have multiple tanks it is often easier to have one sump/filter and have an overflow and return run to each tank. You can do this either with an overflow and return in each tank that run directly to the sump, or you could have each tank flow into the next with the ones on the ends running to the sump. It is possible to do this with a cannister filter and U tubes but water evaporation can cause issues whereas the sump will keep the water levels in the tanks constant. A sump is also a great location for sensor probes, reactors, heaters and other equipment. While you can buy inline versions of many of these, it's often simpler and possibly cheaper to buy the typical in tank versions and place them in the sump. Either method will tend to keep the main display tank less cluttered.

For pumps, do NOT restrict inlet sizes. There is a great thread where just increasing the size of the inlet dropped pump temperature and noise significantly. I think it was Gerryd's tank. Any restriction should be done AFTER the pump on the pressure side. Use at least the same size inlet as the pump takes. If you can, running a size larger to the pump and then dropping to the right size just at the pump should maximize flow, especially if there are any turns in the pipe. Use spaflex or tubing instead of rigid pipe where possible. The sweeping turns will reduce head pressure. It's easier to make a gentle turn than it is to slam into a 90 degree elbow. You can use cone distributor as a manifold instead of multiple T fittings.

Valves come in generally two varieties that we care about. Ball valves and gate valves. Ball valves turn 90 degrees. They are intended for one purpose. ON/OFF. They are not intended for flow adjustment although this can be done. The way they work does not allow for great control or repeatability. Higher quality/expensive ones are smoother and can be better for this. Gate valves require multiple turns to operate and provide much finer control. These are the valves which are used on skimmers and other places where fine control is required. Consider union valves where you may need to remove something for cleaning such as a pump or CO2 reactor or similar. With the exception of a CO2 line from a solenoid, just skip the check valves.

Tank returns can be over the rim or through bulkheads in the back or bottom of the tank. If you choose a lower bulkhead placement you are likely to find a leak at some point. Think about this and consider the consequences of a fast leak, or cracked bulkhead, and how you'll fix it. If you're not concerned, go for it. If this seems like a bad idea or something that might be really difficult to fix quickly, go over the top and use either a siphon break or barely subsurface return as your siphon break. LokLine is a wonderful option and something you should consider.

-
S
 

shoggoth43

Lifetime Charter Member
Lifetime Member
Jan 15, 2009
1,092
11
38
Water level in the sump is not too critical from what I've found. I'm lazy though and don't like filling it too often. I don't have a sponge in my wet/dry although I do try to keep the water level just below the media for maximum effectiveness. I have seen a design where the media was fully submerged giving the same effect as a cannister. I don't care for this only as it can potentially reduce O2 levels if flow stops for whatever reason. Conjecture ahead -> Having the media "dry" would seem to give it the most O2 without competing for what's already in the tank. Kind of like how an emersed plant has ready CO2 access without using as much of the CO2 dissolved in the water.

I have noticed that flow will drop significantly as I let the water level drop. I'm using an external pump and the water level from just below the media to where the pump could suck in air is only 3 inches at most so it's noticeable on my setup. If I ran a higher water level it may be even more noticeable but might not be an issue with a different pump.

It's pretty common to dump water into the tank. I might throw a pitcher of water in every day or so, or just hear the pump sucking air and throw a few pitchers in and let the water levels settle down to where they need to be. It isn't a big deal although I have run an auto top-off in the past. I just didn't find it to be worth the effort to maintain and pouring water in the tank gives me a chance to flush out any stagnant spots I might find.

Coast to coast is just as you said. It doesn't have to run the width of the tank and you don't need the overflow on the back. It can be on the side if that makes things easier for something like a peninsula setup.

You can do the silent overflows with one return. However, it won't be stable and you'll fiddle with it often and when something goes out of whack you'll either suck air in the line ( this is good ) or it won't keep up ( this is bad ). You really need two lines to keep it stable and safe. The main line runs to the wet/dry. If you bought a two return design wet/dry you can run both line to the wet/dry tower. If not, just run the safety line to the main chamber instead. If you leave it above the water line you'll hear a trickle of water noise which will lead you to A) investigate what's going on which is a good thing and B) probably need to use the rest room. Keep all lines separate for best results.

The U tube will remain deep enough in the water in both boxes to keep the siphon. I'm using a commercial eShoppes filter and overflow. It's their smallest dual return model. On one side I have a small fitting I shortened to place the sponge filter on. I stack the other sponge filter on top of that. On the other side I shortened to tube to be just slightly higher than the internal plastic wall, maybe an inch higher. This line is the safety line. At the end of the main return line, just before the wet/dry I have a gate valve to adjust the flow. It then hits a 90 degree fitting and goes into the top of the wet dry. The safety line goes into the other hole in the top of the wet/dry. Putting the gate valve at the other end seems to let the water fill up the tube easier and purge out all of the air. Doing it the other way doesn't seem to work so well and since it would be out in the open would encourage "adjustments" from guests. If there's a flow problem with the pump the overflow will suck air in. If the prefilter clogs you'll hear either the trickle noise or if it REALLY clogs you'll get that horrible gurgling toilet flushing noise of a straight standpipe. Any noise is your queue to investigate.

-
S


tjbuege;73854 said:
Soggoth43,

How much water is really needed in a sump? What is the desired level? Does the volume of water have any bearing on the effectiveness of filtration? I would think the biological (wet/dry) portion, along with some sort of foam filter between the wet/dry and the main sump reservoir are the important things. What I'm thinking is, the sump water level is probably not going to be running anywhere near the top of the sump tank. If it's too deep, then it will flow back into the wet/dry area and it will no longer be wet/dry, but just plain wet. :)

Also, how much does the water pressure (i.e. level of water in the sump) above the return pump really come into play? If the return pump is sitting in 10" of water, or if it's sitting in 5" of water (due to evaporation, for example), is that really going to have a big impact on the flow of that return pump? I'm thinking the distance from the pump output to the actual aquarium will have the most impact on flow, as that is the greater distance, and it's on the output side of the pump. And it's a fixed distance. Of course, I could be completely overlooking something here, so please correct me if I'm wrong.

Evaporation is a variable factor in this, I know. My tank will be 36x18 (assuming outside dimensions), so one inch of water is about 2.6-2.7 gallons, depending on glass thickness. Not sure how long it takes for that much to evaporate in my house, but I understand that it will drop the level in the sump faster than it would in the aquarium if I were running a closed canister system. An inch drop in the aquarium equates to about a 2 3/4" drop in the sump. Right now, my 20g long (30x12) loses about 1" of water every 2-3 days. That leads me to believe with the larger tank I'm planning, I'll probably lose 2-3" of water in the sump every couple days. Is it normal to have to constantly top off the water in the sump to keep it running? Is this a daily routine? It's another consideration in my decision to go with a sump. I know I could probably plumb in some sort of auto top-off, but I'm not ready to get that complex yet.

I'm also not willing to drill holes in an aquarium at this point, so I would opt for an external overflow box. Coast to Coast overflows... that refers to a built-in style of overflow that spans the entire length of the back wall of the tank, right? And requires drilled bulkheads? It sounds interesting, and I understand the benifits of that design, but again, I don't want to drill right now. :)

Good tip on rounding the inside edge of the overflow box lip. Makes complete sense. I bet that has other applications in the aquarium, too.

Interesting about the two output lines. I think I'm starting to understand the basis of these silent overflow designs. I'll have to reread them now that you gave a good explanation on how they work. If I purchase an overflow box (as I probably will this first time around), the models with two drain lines are quite pricey. Is the two line design critical in silent operation, or is it possible to run silent with a single return line? Also, do you run the two separate lines all the way to the wet/dry portion of the sump, or do you combine them at some point above that? I suppose rereading some of these articles will answer that question.

Regarding your comment about overflow siphon and water level in the overflow box, I'm struggling a bit to understand this. If the water level in the back box is lower than that in the front box, and if you shut off the pump, wouldn't the water levels want to equalize, dropping the water level in the front box, breaking the siphon? Or are both ends of the U tube the same length, and the part inside the tank is just in deeper water? Having a hard time picturing this one. Regardless, if I purchase an overflow box, it's design should be such that this isn't an issue, right?

Speaking of commercially-made overflow boxes, I'm still wondering which brands are recommended? CPR, Eshoppes, Tom Aquatics, Lifereef? From what I can see: CPR doesn't use a siphon tube, so trapped bubbles could be a problem; Tom Aquatics looks like it has extra "stuff" involved (not sure what I'm seeing in the pictures); Eshopps PF-1000 has two outputs, uses a siphon tube, and won't break the bank (about $72); Lifereef get's high reviews, and uses a siphon tube, but their double drain model is expensive... $200. Right now, Eshoppes looks to be the best value, but if it's junk, I need to know that.

Thanks for the discussion and the ideas!
 

tjbuege

Lifetime Charter Member
Lifetime Member
Jun 25, 2009
119
0
16
Minnesota, USA
Hi Shoggoth43,

Ok, lots and lots of stuff for me to digest, before I respond with more questions. :D

But ... do you suppose you could post some pictures to your setup? The Eshopps Overflow box you mentions is the one I was looking at. I'd specifically like to see pictures of your modifications and setup. I think that would help me understand it.

Now, I need to go read and reread your responses...
 

scottward

Guru Class Expert
Oct 26, 2007
958
10
18
Brisbane, Australia
Check valves FAIL. They need maintenance.

Sorry to interrupt the detail of sumps (interesting reading), but then again this is MY thread so :p

Just a quick one regarding check valves, I'm going to fit a swing check valve between my circulation pump and my AM1000 to prevent gas backflowing into the pump during a powercut. Shoggoth, regarding the failure of check valves, are you talking about gunk building up on the seat preventing it from sealing closed properly? This check valve has a maintenance port (or whatever it is called), plus I intend to regularly test it by deliberately shutting off the power when I have at least an inch of CO2 build up inside the AM1000. I don't want to use a long enough run of hose between the pump and reactor such that the high point can contain the gas, nor can I leave the bleed valve open (i.e. dual venturi; too much churning, too noisy etc). I feel that I have better odds here using a check valve. Just thought I'd quickly ask....Tim is busy 'digesting' anyway.....
 
Last edited by a moderator:

tjbuege

Lifetime Charter Member
Lifetime Member
Jun 25, 2009
119
0
16
Minnesota, USA
scottward;73867 said:
Sorry to interrupt the detail of sumps (interesting reading), but then again this is MY thread so :p

Hey, interrupt away! :) Although, I seem to recall being the one to suggest a separate thread on sumps... LOL :)

scottward;73867 said:
Tim is busy 'digesting' anyway.....

Yeah, I feel like I've been eating at an all-you-can-eat buffet, stuffed to the gills :) and still trying to eat more!
 

shoggoth43

Lifetime Charter Member
Lifetime Member
Jan 15, 2009
1,092
11
38
If you maintain the valve you should be fine. Gunk builds up on seals. It can gum up the hinge on the valve, or the sliding action on the other style. I've also seen them "bounce". The seal compresses and causes a slight rebound and then the pressure of the water slams it back only to bounce. It's a very characteristic "Thunk-a-thunk-a" noise. Each bounce lets a little bit of water through. Over ten/fifteen minutes, it's no big deal. Over a weekend where you weren't around would probably not be an insignificant amount of water. Maybe it would eventually settle down and seal properly but it didn't seem to be stopping. Maybe it was just the one valve we were using but if it happened once...

I didn't see this in the hinged style but it really comes down to making sure you maintain it. Like everything else, at some point "you" as in general users will get to the "meh" stage and skip a month. No problem. Then it becomes every other month, and did you do it last month? I think so, except maybe it was the month before, or before that, etc.

Whether or not your particular application will have this issue or even result in a catastrophic problem would depend on what happens if the valve fails. If you just have a pump that takes a while to prime it may not be a big deal. If this is the only thing holding the water in a tank on a bottom mounted bulkhead you need to maintain and test without fail. I'd much rather go over the side and use a siphon break and take the slight aesthetic hit than risk the entire tank spilling out over a long weekend.

Also, I did get some photos but I'm having some trouble getting them to upload.

-
S


scottward;73867 said:
Sorry to interrupt the detail of sumps (interesting reading), but then again this is MY thread so :p

Just a quick one regarding check valves, I'm going to fit a swing check valve between my circulation pump and my AM1000 to prevent gas backflowing into the pump during a powercut. Shoggoth, regarding the failure of check valves, are you talking about gunk building up on the seat preventing it from sealing closed properly? This check valve has a maintenance port (or whatever it is called), plus I intend to regularly test it by deliberately shutting off the power when I have at least an inch of CO2 build up inside the AM1000. I don't want to use a long enough run of hose between the pump and reactor such that the high point can contain the gas, nor can I leave the bleed valve open (i.e. dual venturi; too much churning, too noisy etc). I feel that I have better odds here using a check valve. Just thought I'd quickly ask....Tim is busy 'digesting' anyway.....
 
Last edited by a moderator:

tjbuege

Lifetime Charter Member
Lifetime Member
Jun 25, 2009
119
0
16
Minnesota, USA
shoggoth43;73876 said:
Also, I did get some photos but I'm having some trouble getting them to upload.

I haven't been able to upload photos, either. What I ended up doing was posting them to my own website and then linking directly to them. But if you figure out how to post let me know.
 

chopsticks

Prolific Poster
Jan 20, 2011
70
0
6
Costa Rica
scottward;73867 said:
Sorry to interrupt the detail of sumps (interesting reading), but then again this is MY thread so :p

Just a quick one regarding check valves, I'm going to fit a swing check valve between my circulation pump and my AM1000 to prevent gas backflowing into the pump during a powercut. Shoggoth, regarding the failure of check valves, are you talking about gunk building up on the seat preventing it from sealing closed properly? This check valve has a maintenance port (or whatever it is called), plus I intend to regularly test it by deliberately shutting off the power when I have at least an inch of CO2 build up inside the AM1000. I don't want to use a long enough run of hose between the pump and reactor such that the high point can contain the gas, nor can I leave the bleed valve open (i.e. dual venturi; too much churning, too noisy etc). I feel that I have better odds here using a check valve. Just thought I'd quickly ask....Tim is busy 'digesting' anyway.....

You can run the reactor with a powerhead in the sump and feed the output of the reactor into the inlet of the return pump, this way you avoid the check valve and the gas messing with the return pump. I use this setup in my wet/dry.
 

nipat

Guru Class Expert
May 23, 2009
665
0
16
scottward;73867 said:
Sorry to interrupt the detail of sumps (interesting reading), but then again this is MY thread so :p

Just a quick one regarding check valves, I'm going to fit a swing check valve between my circulation pump and my AM1000 to prevent gas backflowing into the pump during a powercut. Shoggoth, regarding the failure of check valves, are you talking about gunk building up on the seat preventing it from sealing closed properly? This check valve has a maintenance port (or whatever it is called), plus I intend to regularly test it by deliberately shutting off the power when I have at least an inch of CO2 build up inside the AM1000. I don't want to use a long enough run of hose between the pump and reactor such that the high point can contain the gas, nor can I leave the bleed valve open (i.e. dual venturi; too much churning, too noisy etc). I feel that I have better odds here using a check valve. Just thought I'd quickly ask....Tim is busy 'digesting' anyway.....

Objection. Don't use check valve if you can have simple siphon break hole.
Siphon break hole is difficult to "seal" unless it's done by intention.
Leaves sticking there just cannot do it.

I used to run my tank with a sump. I was afraid of flooding the room
with tank water too. So I tested and tested, siphon break hole just worked.
If you are paranoid, make it two holes. I can't understand it why some people
prefer check valve.

PS: And you could test siphon break to see if it's compatible with your
CO2 reactor scheme before going out to buy a check valve. I would think
the siphon will be broken before air reaches your reactor.

PS2: And IF air reaches the reactor during power outage, so what?
I mean power outage doesn't happen often (or does it in your area?)
And it's just a seldom issue to deal with. For my point of view, it's a much
smaller issue than check valve failure. (Imagine a leaf sucked down
the hose and stuck at the check valve:p )
 
Last edited by a moderator:

shoggoth43

Lifetime Charter Member
Lifetime Member
Jan 15, 2009
1,092
11
38
After reading it a bit more, it seems more that his check valve idea is to prevent the gas in the reactor chamber from back flowing into the pump and causing it to have a mostly dry impeller chamber. This might cause difficulties with a restart. As I use a needle wheel design for CO2 and direct that into the main pump intake, any lengthy outage on the main pump can cause a large amount of CO2 to build up in there. When the power comes on it sometimes has a tough time clearing out the CO2 and repriming. I think this is more what he's trying to do.

If this is the case, then I'd put in the siphon break at the top of the tank as usual. A failure of the check valve would not be a catastrophic event and would just cause pump restart problems but would not risk any sort of flood. I may have misread his plumbing arrangement though.

-
S

nipat;73891 said:
Objection. Don't use check valve if you can have simple siphon break hole.
Siphon break hole is difficult to "seal" unless it's done by intention.
Leaves sticking there just cannot do it.

I used to run my tank with a sump. I was afraid of flooding the room
with tank water too. So I tested and tested, siphon break hole just worked.
If you are paranoid, make it two holes. I can't understand it why some people
prefer check valve.

PS: And you could test siphon break to see if it's compatible with your
CO2 reactor scheme before going out to buy a check valve. I would think
the siphon will be broken before air reaches your reactor.

PS2: And IF air reaches the reactor during power outage, so what?
I mean power outage doesn't happen often (or does it in your area?)
And it's just a seldom issue to deal with. For my point of view, it's a much
smaller issue than check valve failure. (Imagine a leaf sucked down
the hose and stuck at the check valve:p )
 

nipat

Guru Class Expert
May 23, 2009
665
0
16
shoggoth43;73898 said:
After reading it a bit more, it seems more that his check valve idea is to prevent the gas in the reactor chamber from back flowing into the pump and causing it to have a mostly dry impeller chamber.

But this is a sump (not canister filter). I mean the pump is submersed in the sump, so the pump
shouldn't run dry.
 

RukoTheWonderDog

Junior Poster
Feb 9, 2011
27
0
1
Western NY
nipat;73899 said:
But this is a sump (not canister filter). I mean the pump is submersed in the sump, so the pump
shouldn't run dry.

I have an awful time with drylock in the pumps on my tank. The reason I removed the powerhead & tubing used to atomize the gas before my pumps was due to the increased buildup of gas in the pumps with all the extra tubing. Even though the pumps are submersed, they will still build up a pocket of gas which can prevent the pumps from priming and occasionally causes my QO impellers to run the opposite direction.

Concerning the water level in the tank; as already pointed out this is not super important. Don't overfill the sump (leave room for the tank water backflow when the power is off) and keep enough water in the sump to keep the pumps from sucking air. One real benefit of the wet/dry design is that all of your media will always be wet as long as water is flowing through the sump. The water level in the sump will rise and fall, but the mechanism used to spray water over your media will keep it wet regardless of the sump water level. A non wet/dry sump design could expose the upper levels of media to the air if the sump level drops below the media level. Drying out bacteria colonies is a good way to kill them, and dead bacteria can do no work.

Evaporation in a drilled tank will obviously be higher, although I am of the opinion that the biggest factor in evaporation is your humidity in the house. My 180 is on the first floor, while the others are in the damper basement. In the summers I see high evaporation in the 180 due to decreased humidity from AC. The untreated basement is damp due to condensation and wall seep and the tanks there lose very little water from evaporation. In the winters the humidity upstairs is higher than the basement (due to furnace humidity controls) and I find that my water levels in the basement tanks (all undrilled tanks w/ canisters) drop faster than my drilled 180. I bought a humidity gauge to help keep an eye on this. When humidity levels get below 30% or so, I know I'm going to be topping off some tanks within 48 hours.

I am also of the opinion that PVC check valves are not needed, and add another point of failure. If you design your system well you won't need it. I've done extensive power out tests on my tank and have the peace of mind that the siphon breaks will prevent overflowing. I wouldn't personally put my trust in a $8.74 piece of PVC.....but that's your call.

Keep in mind that many of the reef guys have been using sumps for decades without problems of overflowing. None of the saltwater guys I know use check valves for the reasons mentioned in the previous posts.
 

shoggoth43

Lifetime Charter Member
Lifetime Member
Jan 15, 2009
1,092
11
38
I wasn't able to upload the pics, but if you want to PM me I might be able to email them to you.

-
S

tjbuege;73858 said:
Hi Shoggoth43,

Ok, lots and lots of stuff for me to digest, before I respond with more questions. :D

But ... do you suppose you could post some pictures to your setup? The Eshopps Overflow box you mentions is the one I was looking at. I'd specifically like to see pictures of your modifications and setup. I think that would help me understand it.

Now, I need to go read and reread your responses...
 

tjbuege

Lifetime Charter Member
Lifetime Member
Jun 25, 2009
119
0
16
Minnesota, USA
Are there advantages to running an exteral return pump in the sump, as opposed to just dropping a pump into the dump water directly? I would think it would be easier to just drop a pump in instead of drilling a hole in the side of the sump and running plumbing to an external pump. More room for leaks if you don't seal things right.
 

RukoTheWonderDog

Junior Poster
Feb 9, 2011
27
0
1
Western NY
tjbuege;74093 said:
Are there advantages to running an exteral return pump in the sump, as opposed to just dropping a pump into the dump water directly? I would think it would be easier to just drop a pump in instead of drilling a hole in the side of the sump and running plumbing to an external pump. More room for leaks if you don't seal things right.

External pumps become common in larger tanks; they are usually not utilized for smaller tanks. Most external pumps are high flowing and would be overkill for most tanks under 200 gal. External pumps are usually more expensive, are usually louder, and usually consume more power. As you also pointed out, they do also offer more complexity and places for leaks, but a well constructed system should be leak free.

The setups I have seen in person that use external pumps were saltwater reef tanks. They did so because sump space was needed for protein skimmers, refugiums, etc. I don't know a single freshwater guy with a tank smaller than 300 gallons that uses an external pump.
 

nipat

Guru Class Expert
May 23, 2009
665
0
16
Hehe, ;) It's Tom's 120 gallon.
http://www.barrreport.com/showthread.php/8333-120-Gallon-ADA-like/page4

5862f9f8.jpg
 

shoggoth43

Lifetime Charter Member
Lifetime Member
Jan 15, 2009
1,092
11
38
Mine does. ;)

After I redo the sump with a larger tank I'll probably just get a submersible model and throw it in the sump for space savings and noise though.

Another possible reason the reefers use them is just because they want/need/use high flow in the tank. Faster turnover through the sump helps with that and you get more cycles through the skimmer. Those without skimmers using just 'fuges can't/won't use the high flow so a smaller pump is typical. External or submersible works for that application. Different applications and different schools of thought. You definitely don't need 50+ turnover in a seahorse tank, but you might do something like that in a stony coral tank. I suspect my tank would go from planted to "sticks with swirling sand" if I tried that. The more water I pump back to the tank the lighter my wallet gets when the electric bill comes in so I try to keep turnovers to the sump to a minimum and prop pumps to make up the difference in the tank.

-
S

RukoTheWonderDog;74098 said:
External pumps become common in larger tanks; they are usually not utilized for smaller tanks. Most external pumps are high flowing and would be overkill for most tanks under 200 gal. External pumps are usually more expensive, are usually louder, and usually consume more power. As you also pointed out, they do also offer more complexity and places for leaks, but a well constructed system should be leak free.

The setups I have seen in person that use external pumps were saltwater reef tanks. They did so because sump space was needed for protein skimmers, refugiums, etc. I don't know a single freshwater guy with a tank smaller than 300 gallons that uses an external pump.