PDA

View Full Version : Sump flow design


Lavoisier
11/10/2012, 02:54 PM
Here is my plan for my sump. My question (probably a dumb one) is if I can create a slower flow chamber with the use of the two valves pictured below. This chamber would be used for macro algae.

http://i1321.photobucket.com/albums/u542/jimselby/500g%20Build/Sump.png

Lavoisier
11/11/2012, 07:13 AM
I woke up last night and realized it was a dumb question. Does the addition of a small powerhead look better? I will add lighting above this compartment on a reverse cycle to the tank

http://i1321.photobucket.com/albums/u542/jimselby/500g%20Build/Sump-2.png

uncleof6
11/11/2012, 02:23 PM
Feed it from a tee in the return line, and over flow into the return section. Why make it more complicated than it needs to be? Pipes/valves through bulkheads/baffles is complicated, and.........I guess that sump operation is so basic and simple, that it cannot possibly work well, unless we complicate it......

And your return section is too small.......

Lavoisier
11/11/2012, 04:27 PM
Excellent. For some reason I thought the BeanAnimal should not be teed. Great solution. How large should the return section be?

SGT_York
11/11/2012, 06:09 PM
Don't T the bean animal but since you have two overflow lines use the siphon flow line to your skimmer section and the lighter durso line to the refug. (flow dependent you can easily T the durso or emergency lines) The return section needs to handle the excess water when the power is disconnected and still have enough space to stop the bubbles. (rule of thumb 1/3'rd the size but there are much better calculated methods) I would bet your refug section with the plumbing will create air bubbles close to your return pump intake. I'd raise the baffles on the refug side and have it flow to the other side compartment then into the return section to give you added space against bubbles. You could also add a carbon shelf in case you wanted to run media.

Lavoisier
11/11/2012, 07:07 PM
Thank you. I'll work up the new layout tomorrow and post it. I appreciate you thoughts.

Lavoisier
11/11/2012, 08:50 PM
I was able to get to this tonight after all. Does this look like a workable design? (I won't actually need a ball valve on the outflow, just a bulkhead)

http://i1321.photobucket.com/albums/u542/jimselby/500g%20Build/Sump3.png

uncleof6
11/11/2012, 09:27 PM
Excellent. For some reason I thought the BeanAnimal should not be teed. Great solution. How large should the return section be?

101: You tee the RETURN--the RETURN goes from the PUMP to the TANK. BeanAnimal's design is a DRAIN--not a RETURN, it goes from the TANK to the SUMP. We need to speak the same language here, otherwise, there is nothing but chaos.

SGT_York
11/11/2012, 10:20 PM
Don't T the Siphon the water will just flow straight down and not into your refug. I'd use the middle line and run it directly into the fuge. You can control the water headed into the refuguium by opening/closing the syphon on the right line.

Uncleof6 is proposing a good system to T the return line from your return pump IMO this wastes efficiency as the above works just as good.

Also your fuge drain will create bubbles. move that drain to the middle baffel and you'll be good to go.

uncleof6
11/12/2012, 01:27 AM
Don't T the Siphon the water will just flow straight down and not into your refug.

But with a valve(s) (which is used in the siphon) you can tailor the flow any way that suits you. However, this has nothing to do with why we DON"T want to tee or do anything else with the siphon: It can create starting issues--the siphon may fail to start properly.

I'd use the middle line and run it directly into the fuge. You can control the water headed into the refuguium by opening/closing the syphon on the right line.

The only problem is, when the system is adjusted properly, there is only a trickle of flow through the open channel line. Insufficient to provide flow, by anyone's standards, through a fuge. The valve in the siphon line is used to raise the water level in the overflow till water just flows in the open channel.

Uncleof6 is proposing a good system to T the return line from your return pump IMO this wastes efficiency as the above works just as good.As I have explained, it does not work, with this system. Additionally, when using a durso or other air assisted drain system, it is unwise to use a valve(s) in the line as they exacerbate the issues with these type of drain systems. They also create a plug hazard for systems using only one drain line.

The bottom line is, bean's system works as designed. A few modifications will not harm the function. However, the splitting of lines etc. is a major modification.

Also your fuge drain will create bubbles. move that drain to the middle baffel and you'll be good to go.One of the design criteria for this system was to eliminate bubbles, and noise. Your proposed modification will create both.

Lavoisier
11/12/2012, 07:48 AM
Cheeez...what a knucklehead. 101--you're being very kind, try Kindergarten! This is what I get for not reading carefully. The vice is one for which I am constantly correcting my students.

Thank you, as well SGTYork. I appreciate your patience and willingness to help.

Yes, Uncle, I totally get the reasoning which is why I was confused when I misread your excellent solution--tee the RETURN line. (Whenever I see you post my first thought is Overflow expertise. My bad).

I'll correct and re-post.

uncleof6
11/12/2012, 01:45 PM
The return section should be large enough, that the system can run for several days, without the pump running dry/sucking air--and destroying the pump. Average evaporation is 2 - 2.5 gallons per day. ATO's are a convenience, not a fail safe. They can fail. We need a passive fail safe to protect the pump--at least for a while, giving time to discover the failure of the ATO.

The system can run without a "fancy" sump with a huge fuge; the system cannot, however, run without a pump.

Lavoisier
11/12/2012, 03:19 PM
The return section should be large enough, that the system can run for several days, without the pump running dry/sucking air--and destroying the pump. Average evaporation is 2 - 2.5 gallons per day. ATO's are a convenience, not a fail safe. They can fail. We need a passive fail safe to protect the pump--at least for a while, giving time to discover the failure of the ATO.

The system can run without a "fancy" sump with a huge fuge; the system cannot, however, run without a pump.

Good point on the ATO--they are not fail safe. So, a couple of questions as I rethink my sump.

I may need to use a higher average evaporation rate with my larger tank (500g+). Would you agree? and would a 1% per day (5-6g) be overkill? Numbers are all over the place with anecdotal experience on various threads as one would expect with all the variables possible. I'll check Escobar when I get home to see if he has anything on this issue.

By "return section" do you mean the area after the last baffle and the pump?

[I had planned on a large cryptic tank beside the sump but may be safer moving the macro algae to that space entirely or doing half and half.]

Lavoisier
11/12/2012, 03:34 PM
I may have found the answer to my first question. This study (http://maxwellsci.com/print/crjbs/%283%2972-77.pdf) done with 28 8'x4' tanks of three different depths 24", 30", 36" showed an average loss of 1.2g per day. So I think 2-2.5 will be good, conservative number to use in my design.

Lavoisier
11/12/2012, 07:12 PM
So here is my redesigned sump. The return section allows for @15g which at 2.5g evaporation per day gives me 4-5 days of cushion. The center section will be filled with live rock. The Algae will be moved to a second tank/refug. beside the sump and feed with a gate valve from the return--see main build thread (http://www.reefcentral.com/forums/showthread.php?t=2229879)

http://i1321.photobucket.com/albums/u542/jimselby/500g%20Build/Sump-1.png

_shorty_
11/12/2012, 07:48 PM
Excellent. For some reason I thought the BeanAnimal should not be teed. Great solution. How large should the return section be?

Why shouldn't it be T'd.? Does it reduce overall flow rate? Or just issues getting the syphon going?

Lavoisier
11/12/2012, 07:52 PM
This build will be the first BeanAnimal I've used so I'm relying on Uncleof6's expertise along with BeanAnimal's thread. It is an issue of starting the syphon, rather than reduced flow, and simply no need to risk that critical function with a tee.