Unanswered questions about Triton Method Sump Design

[TylerS]

I'm new to the hobby and have a few questions about the "Triton Method". I have an engineering background so I'm always looking for data driven approaches. As I was doing my research I couldn't find a solid reason for water changes other than to replenish these mysterious trace elements, so naturally I was quite happy when I came across the triton method which matches what I was thinking. However I have a couple of unanswered questions.

The only good info I've come across is written by users "Triton method for dummies". So I don't know what info is based on the users recommendations, and what information is based on actual Triton recommendations and thorough testing.

Questions
1. Sump recommendation is for 10x flow rate through sump. What is the reason for this? Was there testing to prove that 10x is required instead of the more common 3-4x flow rate. Where can I read the details of the testing and the results? 10x will significantly increase overflow noise and power consumption of the return pump of my system (which will be in the basement below the tank).
Sources recommending 3-4x flow through sump:
http://www.advancedaquarist.com/2003/4/beginner
http://reefkeeping.com/issues/2008-01/newbie/index.php
http://coralmagazine.coverleaf.com/coral/20100304#pg41

2. Sump setup of having refugium prior to skimmer, as opposed to after skimmer? I've commonly read that it's best to have the refugium after the skimmer section. What are the nutrients that the skimmer removes that the algae needs? What kind of testing was done to prove this? Where can I see the data?

I'm hopping to hear from official titron reps on this.

 

[joefishUC]

I'm new to the hobby and have a few questions about the "Triton Method". I have an engineering background so I'm always looking for data driven approaches. As I was doing my research I couldn't find a solid reason for water changes other than to replenish these mysterious trace elements, so naturally I was quite happy when I came across the triton method which matches what I was thinking. However I have a couple of unanswered questions.

The only good info I've come across is written by users "Triton method for dummies". So I don't know what info is based on the users recommendations, and what information is based on actual Triton recommendations and thorough testing.

Questions
1. Sump recommendation is for 10x flow rate through sump. What is the reason for this? Was there testing to prove that 10x is required instead of the more common 3-4x flow rate. Where can I read the details of the testing and the results? 10x will significantly increase overflow noise and power consumption of the return pump of my system (which will be in the basement below the tank).
Sources recommending 3-4x flow through sump:
http://www.advancedaquarist.com/2003/4/beginner
http://reefkeeping.com/issues/2008-01/newbie/index.php
http://coralmagazine.coverleaf.com/coral/20100304#pg41

2. Sump setup of having refugium prior to skimmer, as opposed to after skimmer? I've commonly read that it's best to have the refugium after the skimmer section. What are the nutrients that the skimmer removes that the algae needs? What kind of testing was done to prove this? Where can I see the data?

I'm hopping to hear from official titron reps on this.

Hello Tyler,

The flow rates and general system designed were developed as a result of witnessing what yielded the best results for given systems. To my knowledge there was no actual scientific studies done which isolated variables and compared various scenarios. Triton is not "selling" system advice or system design so there is no need for any kind of proof. You can copy the triton design or design a system based around your own ideas or preferences. The fact that it works for so many people is proof enough in my opinion.

The flow rates of 8-10x system turnover is something that I realize seems extreme for most people, myself included to be honest. I am setting up a 600 gallon triton system for a client of mine and I do not plan on putting this amount of flow through the system. I will however use many circulation pumps within the aquarium to ensure proper mixing of all water. The filter system will have enough flow through it to adequately supply the protein skimmer and refugium with a constant flush of new nutrients. The few triton systems I have seen in person do have serious system turnover. They utilize siphons from the overflow box into the sump to minimize gurgling and air in the lines. When I spoke to Ehsan about the subject of turnover while on my Germany trip, he said that his thoughts are to use plenty of flow to absolutely ensure proper mixing of all dosed chemicals all the time. There should be as little dead spots in the aquarium and filter as possible. Providing a strong system turnover is a great way to achieve this. I have seen some setups where small powerheads were placed in the algae section in lieu of the massive turnover rates. In theory, as long as there is good mixing of all system water, I think one would be fine with much less turnover than the 10x pr hour. Again, I am an experienced American hobbyist who is learning the Triton game along with everyone. Down the road I may opt for more system turnover if I experience a need for it.

Regarding placing the algae bed before the skimmer, the reasoning for this is to use the algae bed as a sort of preflight for the main filter system. Triton tanks that are established and running well do not utilize micron socks. All the detritus from the system "feeds" the algae bed naturally. Detritus also gets coated by bacteria, gets broken down, etc. The water then passes onto the skimmer, carbon and gfo section before getting pumped up back to the display aquarium. It is a very simple and straightforward design. Another thing to consider is that skimmers are actually not that efficient at removing waste from water. It is only using air. With the high flow rates triton recommends, placing the skimmer after the algae bed will probably not change the production of skimate by much. There will be plenty of oils, fats, organics etc to be removed even after the algae bed. Again, this is where we encourage people to experiment on their own. On the UK Facebook page for triton users, there are over 500 triton hobbyists who have experimented for many years looking for slight modifications to the system that can make it perform even better. There are many posts on there that focus on sump related designs, comments, thoughts and theories. You should check it out too.

 

[thejuggernaut]

Those older flow rate numbers (3-5X) are really misguided in allot of ways. The school of thought at the time was 3 fold. First, keep your sump flow to a minimum because a return pump has to overcome head loss which makes them less efficient. Second, pushing larger amounts of water down an overflow can create noise. Finally, it is allot easier to create flow in a tank using a closed loop, which has almost no head loss, or powerheads.

But in reality, keeping a higher flow through a sump has several advantages. It allows for better surface skimming to get the proteins down to the sump. It keeps detritus from settling out. It allows you to keep your cheato rotating, so it can stay clean and allow light to hit it from all sides. Making it grow faster and in turn, process more nutrients. It allows you to filter the water faster if you use mechanical filtration (floss or socks). The argument about noise is moot if you set the system up properly, because you can flow 1000s of gallons of water completely silently if setup right. The energy consumption is minimal to add more flow. Especially if you consider that you will need powerheads in your sump to keep the detritus from settling out. Those power heads will consume power, require maintenance (cleaning) and you will probably never overcome the initial cost of them with the small amount of energy savings.

Higher sump flow has nothing to do with Triton itself. It's just good for any tank setup.

 

[TylerS]

Thanks for the replies. It's good to know what has been thoroughly tested, and where there is room for variation or improvements and optimizations on my part for my particular desires or setup.

Regarding the higher turnover rate, going from 3.5x to 10x raises my pump purchase cost by $100, and the yearly operating cost by $100 per year. It's definitely a cost that I would like to avoid if it's not necessary. This is from detailed calculations of head loss and flow rates and then data mining of various pumps costs and power draw.

I also did a lot of data mining on skimmers using the manufactures rated tank sizes and the water flow rate of the skimmer. At 3.5x I would already have water bypassing the skimmer because the skimmer water flow rate is less than the sump turnover flow rate. In which point the water would simply blow by the skimmer and back into the return pump.

 

[Sangheili]

For reference my system is thus:

100g Display Tank
20g Long Refugium
67g Frag Tank
40g Breeder Sump

The sump is three chambers. Left Middle Right
All 3 tanks drain to the left sump chamber. Frag tank and Refugium are fed from a Jebao DC9000 in the Left Chamber.
Middle chamber is skimmer.
Right chamber is display tank return, another Jebao DC9000 (but on about half speed due to small overflow piping).

Theory is the same as Triton guidelines, a decent amount of the display water heads off straight to the refugium before hitting the skimmer. Its not 100% like Triton suggest (probably more like 30%) but it follows the same theory.

I have a decent amount of head loss in the pumps.

Sump Pic:

 

[TylerS]

Thanks for your example

I looked up the specs on your pump, and at 6 feet head loss it would be pumping around 1320 g/hr. Then assuming 50% volume (since you said the DT return pump is turned down) it would be around 650 gallons/hr, so probably around 6x display tank volume goes through the overflow from the main tank.

If you have the same head loss for the frag tank and fuge then you have 1320 g/hr going through them. Since the fuge pump has 1320 g/hr and DT pump has 650 g/hr then some water is recirculated (taken from the fuge and sent right back to it). That means that depending on where your drain lines are in relation to the pump, you could actually have 100% of your display tank overflow going into the fuge (if I understand your setup correctly).
Interesting setup.

 

[TylerS]

I think it's been said many times in the hobby that there are a million ways to do things, and there's certainly lots of successful tanks with various flow rates through sump, and successful tanks with fuge after skimmer as well as fuge before skimmer.

It seems like I'm going to have to experiment. Maybe I'll be the guy to figure out which actually works best. I'm planning to have separate tanks for the fuge and sump, so I could just reroute the plumbing to vary the setup. I'll leave the return pump on unions (or maybe buy a controllable pump) so I can experiment with flow rates also. I have a friend I could borrow some pumps from and whatnot. I'll just have to make sure my overflows can handle 10x flow...

I can't find the UK facebook page for triton. Can anyone point me in the right direction?

 

[joefishUC]

I think it's been said many times in the hobby that there are a million ways to do things, and there's certainly lots of successful tanks with various flow rates through sump, and successful tanks with fuge after skimmer as well as fuge before skimmer.

It seems like I'm going to have to experiment. Maybe I'll be the guy to figure out which actually works best. I'm planning to have separate tanks for the fuge and sump, so I could just reroute the plumbing to vary the setup. I'll leave the return pump on unions (or maybe buy a controllable pump) so I can experiment with flow rates also. I have a friend I could borrow some pumps from and whatnot. I'll just have to make sure my overflows can handle 10x flow...

I can't find the UK facebook page for triton. Can anyone point me in the right direction?

Here it is. They are really great people. I recommend browsing through all the old threads as I'm sure most of the questions you have will have already been discussed in detail. This will prevent repetitive topics from popping up over and over.

[FONT=helvetica neue, helvetica, arial, lucida grande, sans-serif]Official Triton UK, Support & Advice Group[/FONT]

[FONT=helvetica neue, helvetica, arial, lucida grande, sans-serif]It is a private group so you will have to ask to join I believe. [/FONT]

 

[thejuggernaut]

Yea, it is a hotly debated topic, and one that will probably go on forever, mainly because you can create a good looking tank either way. Which way is better? I don't know. But I have had slow flowing sumps. Sumps that run around 8-10X turn over, and the one I am setting up now, I plan on have a 10-14X turnover. In my case, the extra 100 watts will only cost me $72 a year. However, that isn't really true, because the extra 50 watts of heat will offset my heaters running as long at night. It was a moot point for me.

 

[joefishUC]

It certainly is a fun topic. A few popular setups that come to mind for me besides many I have setup:

While at Sanjay's house admiring his incredibly successful 500 gal reef tank a few years ago I was amazed at how little turnover his tank had. He had photon clowns breeding inside his overflow box. That is how little flow went through his sump. Did he have wicked flow inside the tank? Absolutely.

While at Amazing Aquariums in Orange County CA a few weeks ago, I was impressed at the quality of sps growth and color in systems that literally trickle water between a shared sump and multiple display tanks. One small rio pump was doing the work of water transfer between all system parts. And again there was strong flow from vortech style pumps.

Another tank that comes to mind is Joe Yiullo's 20,000 gallon tank at Atlantis marine world in Riverhead LI. I don't have the exact specs but the last time I saw it, the setup was something like this: On average he has approximately 250 gallons in his 500 gallon sump. A single pump supplies water into his massive skimmers and various reactors and then back into the tank. The purpose of his sump is to mainly surface skim and provide a constant water level in the display tank. The flow in the tank is provided by other means. Flow is massive in the tank by the way.

Then there are tanks that are setup with the 10-14x system turnover that look amazing too. The point is that when it comes to turnover and flow, there are many ways to run a successful reef.

 

[TylerS]

Here it is. They are really great people. I recommend browsing through all the old threads as I'm sure most of the questions you have will have already been discussed in detail. This will prevent repetitive topics from popping up over and over.

Official Triton UK, Support & Advice Group

It is a private group so you will have to ask to join I believe.

For some reason there's no hyperlink. And if I google that exact phrase I still don't find it...

 

[Sangheili]

Thanks for your example

I looked up the specs on your pump, and at 6 feet head loss it would be pumping around 1320 g/hr. Then assuming 50% volume (since you said the DT return pump is turned down) it would be around 650 gallons/hr, so probably around 6x display tank volume goes through the overflow from the main tank.

If you have the same head loss for the frag tank and fuge then you have 1320 g/hr going through them. Since the fuge pump has 1320 g/hr and DT pump has 650 g/hr then some water is recirculated (taken from the fuge and sent right back to it). That means that depending on where your drain lines are in relation to the pump, you could actually have 100% of your display tank overflow going into the fuge (if I understand your setup correctly).
Interesting setup.

Yep, the way the water moves in that first chamber I think a good deal of the display water makes it into that pump. You can see the display pipe in the back against the wall of the sump, the water flows out of it and down directly towards the return pump. The other two overflows are pointed towards the 2nd chamber. The water in that first chamber actually flows in a circle due to this. Works out perfectly.

 

[joefishUC]

For some reason there's no hyperlink. And if I google that exact phrase I still don't find it...

If you type the following into your Facebook search bar it should come up.

Official Triton UK, Support & Advice Group

 

[Tim@Triton]

Try this, CLICK :smile:

 

[Newt0075]

Just came upon this thread.... Any updates? I know it has been a few years. Did you end up using the Triton method? If so, how did it go?

 

[George Lopez]

Thanks for the replies. It's good to know what has been thoroughly tested, and where there is room for variation or improvements and optimizations on my part for my particular desires or setup.

Regarding the higher turnover rate, going from 3.5x to 10x raises my pump purchase cost by $100, and the yearly operating cost by $100 per year. It's definitely a cost that I would like to avoid if it's not necessary. This is from detailed calculations of head loss and flow rates and then data mining of various pumps costs and power draw.

I also did a lot of data mining on skimmers using the manufactures rated tank sizes and the water flow rate of the skimmer. At 3.5x I would already have water bypassing the skimmer because the skimmer water flow rate is less than the sump turnover flow rate. In which point the water would simply blow by the skimmer and back into the return pump.

I have decided to make a modified triton system for my new tank. Im flowing a portion into a large fuge, then cascading into two filter socks with the rest of the water, then comes the skimmer chamber, two baffles with sponges for bubble breaks into rubble rock then to my return pump... complicated i know but I'm trying to put the best of all the sumps into it.

 

[ReeferGil]

I have decided to make a modified triton system for my new tank. Im flowing a portion into a large fuge, then cascading into two filter socks with the rest of the water, then comes the skimmer chamber, two baffles with sponges for bubble breaks into rubble rock then to my return pump... complicated i know but I'm trying to put the best of all the sumps into it.

Do you have any pics?

 

[George Lopez]

Ill take a pic of my final plans when i get home after Irma

 

[George Lopez]

Do you have any pics?

Ok so this is a more or less of my idea (30 gallon long 36 x 12 x 16 tall) may reduce the fuge a little depending on the skimmer but basically a large fuge 2 socks a skimmer/ reactor area ruble rock after a bubble break and then the pump.

 

[ReeferGil]

Looks great!