Sump Design

[othercents]

I was reading some articles about flow to and from the sump along with part time skimming that Mr. Wilson who works on a 1350g aquarium in Canada wrote. These articles can be found here. Because of these articles I started looking seriously at sump designs and came up with the following design below based on those articles. I don't have a specific aquarium build I am working on now, but this would be part of my future 120g build I am planning for next year.

My specific sump would run in a 40G Breeder with a Aqua-Medic T1000 being directly fed by the aquarium overflow. Based on the magic number Mr. Wilson mentioned for skimmer flow the return pump would be pushing 160gph back to the display, however this could be faster or slower depending on your tank size.

Let me know what your take is on the design and the idea of running the skimmer only 12 hours during the day hours and shutting off flow to the refugium when the refugium's light is off.

David

BTW. I know I misspelled recirculating in the Sump Design.

 

[that0neguy1126]

There are lots of people who run the simmer part time. A lot of people shut it off after feeding.

The thing that seems low to me is the return pump foe.160 gph back is a little more than 1 tank volume per hour. Most tanks run 10x return flow.

Sent from my HTC Incredible

 

[othercents]

There are lots of people who run the simmer part time. A lot of people shut it off after feeding.

The thing that seems low to me is the return pump foe.160 gph back is a little more than 1 tank volume per hour. Most tanks run 10x return flow.

Sent from my HTC Incredible

I know... I have a question into Mr. Wilson about that specifically. The magic number used to calculate flow for skimmer was 1.3333 x tank volume and he also recommended that speed for the refugium. It specifically required a good surface skimmer in the aquarium pulling all the water directly from the surface of the display into the skimmer. I see the problem with going faster is that you have a tendency to push the water through the skimmer so fast that the proteins get pushed out at the same time. If you are only using the Berlin method which is just a sump with skimmer then return, there is no reason to increase the flow through the sump faster than what the skimmer can handle and we are only debating how fast water should flow through the skimmer, so that the skimmer can do it's job. Once you add a refugium, now you are trying to decide flow based on the refugium and sometimes getting lower skimmer performance because of the increased flow. Different types of refugiums also require different flow (IE. Mangrove vs Cheato).

If you felt as though you needed to increase the flow through the refugium then I probably wouldn't use a direct feed skimmer (even though they are typically more efficient). I would probably use a skimmer that pulled water from the surface of the first chamber in the refugium and then dumped the water into a separate chamber. I also wouldn't use an overflow from the first chamber since that would allow your proteins to escape. I would use an underflow, so that the skimmer can have extra time to pull the proteins from the surface. However doing it this way would need me to redo the autoshutoff.

David

 

[Boogie]

My first tank had a sump just like that. Had to tune the flow to the fuge side by drilling holes. Worked well except it was small and had no place to locate skimmer.

 

[othercents]

10x flow in fuge. Is that based on system size, display size, or fuge size? Most places are saying fuge size.

My first tank had a sump just like that. Had to tune the flow to the fuge side by drilling holes. Worked well except it was small and had no place to locate skimmer.

The fuge section is 32x10x10. To get the 10x flow I need to slow down the flow coming from the skimmer section. I'm thinking that I can mount the shutoff lower which will bring more flow into the fuge especially if I make it work as a siphon. just use a ball valve to tune it back.

David

 

[chrislorentz]

10 times is based on display size. some people go upwards of 20 times turn over for sps. However that is total not just return. I run a smaller return and supplemental with korilla's and the such.

 

[that0neguy1126]

usually 10x for just the return and 25x to 50x total. My tank for instance has 2400 gph from the 2 return pumps. and then I have 2 MP40's which I think can each push 8000 gph.

I think the way your tank set up is different though. I haven't had a chance yet to click the links and read them.

 

[chrislorentz]

WOW I personally dont want flow through my refugium that fast.

 

[that0neguy1126]

refugium != sump

 

[chrislorentz]

? huh

 

[that0neguy1126]

Just because tu have 10x flow going to the dt from the sump doesn't mean u have to have that threw the fuge

Sent from my HTC Incredible

 

[othercents]

usually 10x for just the return and 25x to 50x total. My tank for instance has 2400 gph from the 2 return pumps. and then I have 2 MP40's which I think can each push 8000 gph.

I think the way your tank set up is different though. I haven't had a chance yet to click the links and read them.

What is the benefit of more flow through the sump if your skimmer and refugium are both struggling to keep up? What is the magic number for a Berlin system (sump with skimmer only)? In that system there is no reason to have a higher flow through the skimmer. You want to make sure the proteins are skimmed instead of blown back into the display and all water goes through the skimmer before returning to the display. Granted if you are using the same skimmer for a 75g display as someone with a 120g display then the flow to the skimmer should be consistent. The display size doesn't make the skimmer more efficient, so it is probably best that we can configure the flow to all skimmers to the max display size instead of the current display size.

What should the flow rate be for a 12g refugium? Why would the same size refugium on a 50g display have a different flow than the same refugium on a 120g display? Isn't the flow rate set based on how quickly the refugium can do it's job? If a 120g gets benefit from 1200gph through a 12g fudge then shouldn't a 50g get the same benefit from the same flow rate? Maybe we should more be talking about right sizing the fuge to the correct tank size (IE the fuge should be 20% the capacity of the display with 10x flow from the display).

Btw. This is a future design and discussions about the recommendations Mr. Wilson made. I don't have this setup, but don't see any problems with the configuration unless the recommended flow rates are somehow wrong.

 

[Wicked Color]

Lol, is that a top view of the blue print? That's a crazy design, I want to see it IRL.
I have a mag 18 return on my 75, its way over 10x return volume, no skimmer, I did have one but took it off for a while.
If your running the skimmer part time, the return % wont matter as much as there will be more dissolved proteins.

 

[othercents]

Lol, is that a top view of the blue print? That's a crazy design, I want to see it IRL.
I have a mag 18 return on my 75, its way over 10x return volume, no skimmer, I did have one but took it off for a while.
If your running the skimmer part time, the return % wont matter as much as there will be more dissolved proteins.

Yes, it is the top view. My original design was the same except for the link between the first and third chamber. I did it this way to make the sump more of a display sump that I can put a mantis into. I like that link between the first and third chambers since it allows you to tune the flow going through the refugium or bypass the refugium when the light is off and the Co2 increases. This becomes more of an effort to only feed the tank with the good parts of the refugium and eliminate the bad.

My point on the skimmer is no matter how much proteins are in the water, how much flow becomes too much? 20,000 gph going into a direct feed recirculating skimmer sized for a 120g display might just blow all the protein through the skimmer without being effective. When would it be ineffective? To be specific lets use a Aqua Medic T1000 with the stock pump recommended by Aqua Medic for recirculation. The skimmer should be able to run at the same speed no matter what display size you use, however the weir might be less effective in some tanks that would make the skimmer perform worse, or the bio load much lower.

David

 

[dv3]

i think there are 2 different discussions going on here ...lol...FYI neither one will really affect how healthy your tank is they are just rules of thumb and there is no magic #....thats all

 

[SAZAMA]

kiss

 

[othercents]

i think there are 2 different discussions going on here ...lol...FYI neither one will really affect how healthy your tank is they are just rules of thumb and there is no magic #....thats all

There were two articles. Flow to and from the sump which included the recommendation not to flow more water than your skimmer can handle and the second article about part time skimming which also had an section about shutting down the flow to the refugium when the refugium light is out. The refugium flow was part of a different article.

Since a comment about 10x the display came up, I am only trying to find out if the skimmer and refugium doesn't get any benefit from the faster flow then why are we pushing 10x the display through the sump. I did see a comment somewhere else that the refugium should be 10x the refugium size which in my case works with the 1.3333 x display number mentioned in the article.

I have kind of moved the discussion (probably with myself more an with anyone else) into a discussion about sump and skimmer flow is more based on the skimmer and the refugium size and not the display size unless your weir is not properly sized to surface skim the proteins fast enough.

However you are right that getting the perfect or magic flow through the sump won't matter as much as good husbandry of the reef. This discussion is only good in theory and practical application is up to each individual.

 

[dv3]

now im really confused...lol
good to see you doing your homework though ....every little bit helps i guess