• Folks, if you've recently upgraded or renewed your annual club membership but it's still not active, please reach out to the BOD or a moderator. The PayPal system has a slight bug which it doesn't allow it to activate the account on it's own.

WATER CIRCULATION

Hello Everyone,
I am setting up a 65 gallon tank and do not know what pump I should get for my sump. I will need an external pump for the sump. What is the correct GPH that should be circulated through the sump. I will also be putting two to three powerheads within the tank.
 

RichT

Officer Emeritus
Officer Emeritus
The rule of thumb, which may not necessarily be correct as it seems excessive to me, is 10x the tank volume. So a 65g tank would require a 650gph pump more or less.
 
Also, remember that what pumps these pumps are normally rated for is for 0 height. You need to find what flow rate that particular pump has at the height that it needs to pump the water to. For example, my mag 9.5 is rated at 950 gal/h at 0' height but only 750 at 5' which is the height it had to pump from my sump under the tank to the rim of the tank. Also, these ratings are based on a very specific profile of the tubing, which in all probability will be wider then what you will use for the return line (for example Mag 9.5's 950 is rated for 1 1/2" pipe, while my return line is 5/8" or something like that). So to be safe buy something that is rated more then the flow you need and you can then easily regulate it with some valve.
 
http://reefcentral.com/calc/hlc2.php is a good head loss calculator. Use this to determine which pump to get. do you have an overflow? if so what is it rated for in GPH? what kind of sump do you have? you want to make sure the sump can handle the throughput of the pump. if your not careful with these two issues you can either overflow your display tank or starve your pump.
 
Probably controversial but here's another thought on return rates.

Having a high turnover isn't always a good thing. Many people are now finding out the virtues of slower turnover (way less then 1x) rates. One really nice thing about slower flow rates is less noise. It's a lot easier to regulate the water level in the sump which is what cmsurfr22 mentioned. You also tend to have less problems with micro bubbles in the display. If you have an overflow box with siphon tubes you'll find a slower flow causes a lot less problems with bubbles getting into the siphon tubes causing problems.

Most skimmers work a lot better with a slower flow going into the sump. The skimmer has a lot more of a chance of processing closer to 100% of the "skimmed" water then if the water is just rushing past it in a high flow sump. Any of the sump water that doesn't get processed by the skimmer gets mixed back into the display with a high force and the display typically has a lot more organics in it then a slow flow setup. Ideally the water going over the overflow has all the organics and proteins in it that you want to remove from the water so you want to try and make sure you process as much of it as possible. If your return pump is higher then your skimmers pump in gph then you have a skimmer working with "clean" water instead of good "skim" water.

Also worth considering is how your refugium (if you have one) is connected into the system as well as any additions down the road you might make like the addition of a RDSB (Remote Deep Sand Bed). Fuges and RDSB work better with a slower flow.

Really the only reason I see to have a high flow return is to gain in tank circulation and this can usually be done better with powerheads or a closed loop instead.

Besides the actual return a bigger pump can be used if you plan on having phosban reactors or canister filters, etc. Instead of using multiple smaller pumps for each device you "tap" the main return to supply water to each device.

So ultimately if you know your sump processing cabability in gph then you have a target to use in picking out your return pump. The sump processing cabability could be only your skimmer or could be your skimmer and fuge if they run in parallel. Returning more then your sump can "process" will result in less then clean water introduced back into your display.

Just some food for thought,

Carlo

PS everyone thinks you need big skimmers on large systems. To a large degree you don't, infact a good quality small skimmer can keep a large system clean. The trick is to not return more water to the tank then the skimmer can process. Regardless of the size of the tank the one true fact is that the proteins/organics that the skimmer will remove is taken from the top of the water. If you make sure all this "overflow" water is processed by your skimmer with a slower return pump then you don't need a bigger skimmer. If you can't keep up with the skim/proteins on top of the water in the display, then it's time to upgrade the skimmer or add another one. There is nothing wrong with letting your skimmer process the same "sump water" over and over again to remove as much gunk as possible.

So how much returned water do you need in a "slow flow" setup? Just enough to keep the proteins off the top of the water which isn't much at all.
 
C

concept3

Guest
Hey Ron, I have a 100 or so gallon tank and only use an Iwaki 20 which is rated at about 500-540 gph. I find that this pump is sufficient for me for my return as I have two Tunze streams in the tank which is my main water circulation. I found this combo perfect as the external pump has minimal noise, doesn't add heat into the tank and consumes less power (I think 45 watts).
 
Carlo said:
Probably controversial but here's another thought on return rates.

Having a high turnover isn't always a good thing. Many people are now finding out the virtues of slower turnover (way less then 1x) rates. One really nice thing about slower flow rates is less noise. It's a lot easier to regulate the water level in the sump which is what cmsurfr22 mentioned. You also tend to have less problems with micro bubbles in the display. If you have an overflow box with siphon tubes you'll find a slower flow causes a lot less problems with bubbles getting into the siphon tubes causing problems.

...

I'm not sure if Carlo actually means "way less than 10X" or the statement is correct as is "way less than 1x".

I would agree that you don't need a 10x flow through your filtration system, especially if you use good skimming and other circulation methods. I don't think I'd want to go below a 4x or 5x turnover per hour though. I would also go with a pump rated about 650 gph, because you will loose something between the plumbing and any branching off the pump you wish to do to support other various reactors. I happen to like Iwaki pumps, but there are plenty of others on the market that are also very good.

If you use an external pump, I highly recommend using true double union valves on the intake and output of the pump. This lets you close the valves and remove the pump for cleaning, and/or repair or replacement. You can also use the valve on the output side of the pump to control total water flow.

The rest of the post by Carlo is right on the money as far as skimmers go. I would only add that there are a lot of junk skimmers on the market, especially on the low end, that are just about useless. Make sure you get a quality skimmer, and like everything else in this hobby, it's going to be somewhat expensive, but a good one will last many years.
 
Yea DaveK I did mean under 1x. Told you it might be controversial. But read on for some additional clarification. :)

I was stressing the point that it's far better to 100% process the skim then get caught up in X turnover numbers.

I just grabbed a few flow rates from a few skimmers off the net:
Current USA Fission Recirculating Skimmer 150 gal - 300gph
Current USA Fission Recirculating Skimmer 225 gal - 400gph
AquaC EV-120 Protein Skimmer with Mag-Drive 5 Pump - 225gph
AquaC EV-180 Protein Skimmer With Mag-Drive 7 Pump - 275gph
AquaC EV-240 Protein Skimmer With Mag-Drive 18 Pump - 550gph
AquaC EV-400 Protein Skimmer With Mag-Drive 18 Pump - 700gph
AquaC EV-1000 Protein Skimmer With Iwaki 70RLT - 950gph
AquaC EV-2000 Protein Skimmer - With Iwaki 100RLT - 1200gph


So let's say you have a 125 tank with a typical 20g sump. When all is said and done you have just about 125 gallons of water. If you had an EV-180 then the skimmer can process your tank 2.2 times per hour. If you had the EV-240 then you could process your tank 4.4 times per hour.

Those are pretty realistic numbers IMHO and therefore would say 3 to 4x turnover for normally well equipped skimmer systems is a good turnover rate to shoot for after taking head height and plumbing size into consideration.

Now let's take another system with an EV-240 but assume the person upgraded their tank to a 265. With sump lets say the total water volume is 300 gallons. In this case the skimmer is able to process the water 1.83 times per hour. If the person had an EV-180 then they would not quite be able to process their system one time per hour.

This last example is what I was getting at with the less then 1x processing flow rate. It's an underpowers skimmer for the tank but the best flow rate to keep the water as clean as possible would be under 1x turnover.

For most people with a decently selected skimmer they will most likely be in the 3x-4x range which happens to match many recommendations you get on return rate.

Kind of make sense and looks more realistic when you look at it this way and crunch a few numbers doesn't it?

I think the old recommendation of 10x dates back to wet/dry filters and from the time when the return was the main form of circulation in the tank. Now we have closed loops and the ever popular powerheads which negate the use of high flow returns.

For those people with smaller skimmers but with higher then "recommended" return rates you can always tap the return and and run some of the "return" back into the sump going through a small Phosban reactor running carbon. The carbon will help clean up the water and gives your skimmer a chance to process some of the water again.

Carlo

PS for the record I like to see/target about 3x turnover rate on my own systems and those I setup. This seems to be the sweet spot for cost versus clean water. Regardless I think going beyond 5x (even with a powerful skimmer) is too much. You're not really cleaning the water much more then at 3x and just using more electricity then needed on the return and skimmer pumps. Of course if you're buying powerful equipment with the intent of up sizing down the road then that's a whole nother story and probably good planning. :)
 

RichT

Officer Emeritus
Officer Emeritus
One advantage of the higher turnover rates is gas exchange / Oxygenation. I agree 10x is excessive. Personally, I wouldn't shoot for anything under 5x IMO. But that's just me.
 
I

ILCFisherman

Guest
In terms of biologically filtering the water, yes the slower the turn over the better for the tank, hence the reason you fuge circulates the water as slow as it does. HOWEVER, the more water that you circulate, the healthier and more responsive your animals will be. and this circulation should not just be linear current. So therefore you can have your tank turn over less than once an hour and it be sufficient so long as their is other significant movement IE closed loop, vortex. what i typically like to do is turn the tank itself over anywhere from three to five times with a skimmer that can do the same volume of water with a series of closed loop pumps situated in as many places as i can fit.
 
Thanks so much guys. I have a much better understanding of what I should do now. I got a pump that pumps 500 gph and a skimmer that pumps 500 gph. How does this sound. It's about 8x an hour but i can turn it down a bit if needed.
 

pgordemer

Officer Emeritus
Officer Emeritus
But lets not confuse turnover rate with exchange rate or water movement.

We all know that its important to have good water movement for good gas exchange and to make our tanks happy. Many times though people are trying to use the return pump rate from the sump to equal the overall turnover rate.

Many people want to get 10 to 30 x the tank volume as a turn over rate. Combine all the powerheards, devices and sump for that, don't try to do it all from the sump. 4 to 5 x is more than adequate for the sump.

My MAK 4 in the basement with 40 gallon breeder sump returns around 450 GPH by the time it goes through the UV, bends and 18 feet back tot the tank. But in that tank I also have 2 Koralia #4 powerheads, so now we are looking at 1600-2000 turnover.
 
Ed got exactly what I'm talking about.

What really got me thinking about slower turnover rates (for skimmers) was from working at Adventure Aquarium. On big tanks, it is just about impossible to reach one times turnover processing to clean the water with foam fractioning (protein skimming).

Just to give an example (just posted it in another thread) the Georgia Aquaiums' main tank is 6,300,000 gallons. It uses 28 pumps and 56 sand filters as well as protein skimmers to clean about 1,800 gallons of water per minute. That would be 108,000 gph.

The aquarium's 8 million gallons use 218 pumps, 141 sand filters and 70 protein skimmers to move 261,000 gallons of water per minute. "Move" is the keyword and not clean. Including all circulation they can do 261,000 x 60 = 15,660,000 or approximately 2 time turnover for circulation. But note the difference between circulation and cleaning the water. By pure math it would take over 2 days to clean the main tank water to achieve a 1 time turnover rate. It would get dirtier and dirtier by conventional wisdom using just turnover rates but because the dirtiest water floats you really only need to "skim" the surface and clean this water with foam fractioning. Other water can be cleaned with much less expensive equipment and with higher turnover rates.

With good circulation in the tank (closed loop or powerheads) you can keep the water moving (also good for O2) and allow the organics and proteins to get to the top layer so that it can be skimmed.

At Adventure Aquarium (I'm sure other public aquariums too) the water is pulled from different spots in the tank. The top layer (skim) is pulled and ran through foam fractioning. Mid and bottom tank (closed loops) would pull water and run through sand filters. This can easily be duplicated with home equipment in the form of canister filters, fluidized filters or even small sand filters depending on the size of your tank. When you think about a canister filter like a fluval it will pull water from say 8 to 12" down under the water surface. It's a "closed loop" with filtration built in.

Bottom water that is pulled is the best place to treat UV if you use it since you have the most chance of killing parasites since they live on the subtrate and rock.

I'm in the process of reworking my plumbing (never done it seems) to duplicate this somewhat. The overflow would go to my skimmers, ozone, refugium & carbon reactors while mid pulled tank water (from each tank) will go to a dedicated "closed loop sand filter" which is actually a small pool sand filter setup maybe/optionally followed by carbon if needed). This change should easily allow a 5-10 fold increase in water volume without the need for more equipment. Granted this is way overkill for most systems but the average 100-210g tank can get the same "effect" by slowing down the skimming water and adding something like a canister filter or fluidized sand filter as a "closed loop" as apposed to connecting it into the sump area.

I'm not really trying to go off topic from the "turnover rate" topic but as you can see there as many ways to run filtration that can change the ideal sump turnover rate. All in all I still think the ideal way to figure your sump turnover rate should be tied to the skimmers flow rate since this is probably the most important piece of filtration equipment in the average system and you might as well optimized it's flow rate.

Carlo
 
RonnieB1089, that sounds like a good match up when you take the head height into consideration and tube diameter used. If anything you mights still be returning a little too much as the skimmer will typically flow roughly about 1/2 the pump rating. But you're right in the right ballpark.

See how it goes. Remember you can always tap the ruturn and flow some of the water back into the sump. It's especially easy to do when you aren't running PVC and just using flexible tubing. A barbed "T" and valve with barbed fittings is pretty easy to come by at Lowes/HD or any LFS. So you always have the ability to cut down on return flow if needed.

Carlo
 
Carlo said:
Yea DaveK I did mean under 1x. Told you it might be controversial. But read on for some additional clarification. :)

I was stressing the point that it's far better to 100% process the skim then get caught up in X turnover numbers.

...

Carlo

...

I would agree with you about trying to process all or most of the water through the skimmer as it flows through the filtration system. However, I would disagree with you on an under 1x turnover. I believe it's far better to use a 5x to 10x turnover along with a skimmer that can also process water at or close to that rate.

In effect this is saying that most all skimmer manufactures are "extremely optimistic" when it comes to skimmer sizing.

An example, and I'll just do one, would be an AquaC EV 400 skimmer, with an Iwaki MD55RLT pump, which is rated at 1080 gph at 4' head. This is about the skimmer size I would use on a 100 gal, or so tank. Now the manufacture, AquaC, rates it for a tank size of 100 - 450 gal.

It would be very easy to think that OK it will "work" on a 450 gal tank, and my tank is much smaller, so I don't really need anything that large. I'll use something much smaller. However, as soon as you do that, you are now using a skimmer that is not processing the entire water volume that is flowing through the filtration system. In effect, a less than optimum system has been created.

If you now go back now and optimize the flow, by lowering it to what the skimmer can deal with, and you go much below a 4x or 5x turnover, I believe that you now have another problem. You are not processing the water fast enough for the system you have.

I should point out that this applies to typical reef filtration systems that are in use today. In other words a berlin type sump, and a large skimmer, and optionally a refugium and other filtration "goodies".

If you are talking about the flow through a refugium, or a filtration system based upon algae scrubbing, or similar biological filtration systems, then a much lower flow, sometimes under 1x may be applicable.

My point is that to have an effective filtration system, you need to do both the flow through the filtration system, and the flow through the skimmer correctly. Having either one too low will not create the best possible environment for your livestock.
 
No argument from me at all Dave on the 5x turnover rate if your equipment can handle it and you really want a lot of flow coming back from your sump. I myself wouldn't go 10x under any circumstance as it just isn't needed IMHO.

I 100% do think "most" companies are extremely optimistic in quoting skimmer sizes if you try and run them at 5x with notable high-end skimmer companies not included (i.e. Deltec, Bubble King, Volcano, etc). The same skimmers can run well on quotes system sizes at slower turnover rates. :)

Using the example AquaC EV-400 Protein Skimmer the manufacture claims it to handle 700 gph of flow through the skimmer. This is the actual flow going through the skimmer and not the pump's gph rating. Using my criteria you could go 7x on a 100 gallon tank, 3.5x on a 200 gallon tank or 1.75x on a 400 gallon tank while still cleaning the same amount of water. The skimmer itself doesn't know or care what size tank it's cleaning.

I'd think that skimmer could handle a normal bio-load on a 210 gallon system with about 3x turn over from the sump. In this case it would be able to process all the water going to it. In general from my personal experience if you know the actual flow through the skimmer and assuming it skims well if you can keep the turn over rate at 3x you'll be pretty organic free and can keep nitrates low.

I agree with your example and how the skimmer would not be processing all the water. That in itself is the "key point" I'm trying to make. It is way better to slow the flow of the return down to the point you are processing 100% of the skimmer water.

Here's another way to think about this. Lets say you have a normally stocked 55 gallon tank. Let's also assume your current skimmer kept nitrates at a tested 0 (working well).
Now lets say you move everything over to a new 265 tank. Let's also assume you DO NOT add any additional livestock (you kept the same bio-load). That's an increase of almost 5 times the water volume. Now in this situation the typical person would try and keep the "recommended" 5-10x turnover rate and would probably get a bigger return pump. Now they are moving the water 5 to 10 times faster then the skimmer can process the water and before long nitrates and other organics build up in the tank so the person now purchases a 5 times bigger skimmer.

Guess what, the bio-load didn't change, the same amount of organics would be in the water. Had the person kept the same skimmer and same size return pump they would still be able to process the water and keep nitrates at zero. The only real perceived difference is that if the person didn't change any of the pumps (skimmer or return) but doing the math they would now have a return rate 5 times lower then they previously had before. BUT it doesn't matter because the bio-load didn't change.

Does that make better sense looking at it that way?

Note: I would agree that the circulation in the "new" tank above would need to be increased somewhere around 5 fold but this could be done with powerheads or closed loop systems and not by way of a bigger return pump.

Maybe worth noting, I'm running what typical manufactures call a 250 gallon skimmer on close to 800 gallons of water with a pretty decent bio load using slow flow as I've talked about and have 0 nitrates and ORP in the mid 4 hundreds. I did recently last week add a 2nd skimmer (came with a system I purchased) but it's more for redundancy then for processing as the 1st one kept the system at 0 nitrates without help. Granted it's hard to say how much the skimmer in itself is responsible since I have a lot of rock and a 125g refugium with a DSB but at this point I have no intentions of upgrading to a bigger skimmer. If my ORP gets down into the 3 hundreds or nitrates get up to 10 then I'll know it's time to rethink the skimmer and adjust return rates accordingly.

You touched on what I think is a good point indirectly. As the size of your system increases the use of alternate tools is helpfull. These would include refugiums, remote DSB (maybe in the refugium), UV, Ozone etc. The more "tools" you have the less any one particular piece is needed or relied upon like a skimmer.
 
Top