MrOm's comment makes perfect sense to me and I can't see how he infers anything about the efficiency of the actual protein skimming device.

I don't know why you keep trying to change the focus of the thread, which is throughput for a sump.

If you stop at a gas station and fill your tank to the top with 50 litres of gas, then proceed to overflow another 950 litres of gas (20 x the amount needed) on the ground, where does that rate on the efficiency meter? Would you do this if you read it in a book? Would you reconsider the practice if someone offered you irrefutable proof that it isn't necessary?

I agree that the most efficient skimmer feeding system is a direct feed from the display drain, but most protein skimmers are built by manufacturers to cut production costs, so they use a one pump does all system.

If you are directly feeding a recirculating skimmer from a drain, you should always have an emergency bypass T or Y in the event that the skimmer clogs, can't keep up with the sump return pump, or a surge of water in the tank while you are servicing it. In the event that any water takes the T or Y, the system is no longer FIFO or 100% efficient as MrOm would say :)

With a little valve tweaking you can get close enough to a perfect match with the sump & skimmer pumps to call it "efficient". The good thing about directing the skimmer effluent over a glass partition and trying to match the pump outputs is if you don't get a perfect match, the system is se;f regulating, in that you don't have a skimmer zone that runs dry when the skimmer pump outcompetes, or an overflowing skimmer if the return pump outcompetes.

As bandit pointed out, our goal here is to move water from the display tank to the sump and back for the purpose of filtration. If the filtration only requires a slow throughput, then that's all we need to give it.

If your recirculating skimmer needs 300 GPH throughput, it doesn't matter how big your tank is, you need to install a sump return pump that delivers 300 GPh to the display and subsequent sump/skimmer.

Just like i said match your flow with your equipment.

As I stated in an earlier post, the Sicce PSK 2500 pump used on many skimmers has a maximum output of 660 GPH. I would suspect less than that with the high range of air to water ratio that pump offers. Another very popular skimmer pump is the Laguna Max-Flo which puts out 450 GPH in most cases. These are pumps that you will see on high end skimmers that are rated for tanks up to 300 gallons.

A magic number like 10 or 20 x the volume of the tank turnover has absolutely no bearing on the issue at hand when the protein skimmer used is the same for a 75 or 300 gallon tank. The display/sump throughput should be 500 GPH which is equal to 6.6666 x for the former and 1.6666 for the latter.

I would suggest you use a bucket and stopwatch to meter the exact output of the skimmer once it is broken in and running at maximum efficiency with respect to air intake/production. Only then will you have hard numbers of skimmer pump water flow rates.

Some of the large Beckett skimmers require a very strong feed pump, but these are typically plumbed right into the display tank with an influent at one end and the effluent at the other. They may be sitting in a sump or catch basin, but they are usually plumbed independent of them.

Ok this whole thing is bing made a lot more complacated that it has to be. let me try clarafy and put an end to the back and forth which is getting no where and just confusing people reading this.

Ok so I will put forth 2 senarios for you to read and tell me which one works better all equipment is the same so tank is 80 gal sump is 30 gal, skimmer is fed by a 500gph pump

so senario 1, lets match the skimmer so the return flow is 500 gal per hour

Senario 2, lets use a bigger return pump so the flow is 1500 gph.

and heck let throw one more in
senario 3 lets match the tank, flow rate is 30 gal per hour.

so in senarios 1+2 the skimmer is given all the water it can handle but yet in senario 3 you are starving the skimmer.

so lets pick a number and say that this skimmer is 20% efficient (picked 20 as it is easier to work with) which means in 1 hour this skimmer will reduce the "crap" in 500gal of water by 20%

so our total water volume in these systems is the same at 100gal

system 1 the skimmer has all the water it can handle so it reduces the waist to it theoretical maximum reduction. as per the article Mr Wilson poster that is about a 80% removal.

in system number 2, the exact same amount of water is processed, so you get the same results as system 1.

but in system 3 where we matched the tank size for a 1X turnover, the skimmer only sees 80gal in that hour, but since it skimms the hell out of that 80 gal it is reduced to the max pull down so 80% clean 20% dirty then mixed with the 100% dirty water in the display so 80 gal at 20% mixed with 20 gal at 100% is going to give you about a 35% dirty mixture as aposed to the 20% from the first 2. so you can see if we matched the sump flow for a 1X turn over it would be even worse as we would have 20 gal of 80% clean water mixing with 80 gal of 100% dirty, which would end up with a number of 84% dirty.

So what do we get from this, not enough flow is very bad, to much flow works good but isn't efficient in power used for moving the water.

heaters will work just as good in high flow as in low flow but they transfer more heat to a perticular sample of water in a lower flow, but in a closed system where we have constant cirulation the end result will be the same. and to low a flow can cause heat gradiants which you don't want. as for UV this is a hard one as you have to match the flow written on the UV unit.. mine wanter 400 to 550 gph so I gave it a 550gph pump and knew there would be between 50 and 100gph of losses, chiller same thing match the flow rating.

so the way I had my tank set up I created massive amounts of flow in the display to keep everythign suspended. my surface was very turbulant so I never developed bio film that needed to be skimmed off the surface. but I still used a smooth overflow to reduce noise and I had the water flowing over it at a depth of about 1/4" so I would have enough water flow to carry suspended junk to the sump.

from the sump I had one pump tt remove water for my skimmer, this sucked from my settling pond and returned at the return pump chamber, my heaters were located in the settleing pond where the flow was a little lower. I had two pumps pull water out of the return pump chamber and send it to the chiller and the UV then return at the very bigining again to help keep the flow in the bigining section very fast and turbulant so any suspended matter wouldn't fall out befor the settling pond.

the result of this system was very very little cleaning of crap from the bottom of the desplay, but lots from the settling pond which is what I wanted. I would shut down the return pump, skimmer pump and UV pump, suck all the water from the settling area (about 20) then put my new salt water in there and water change and clean up was done. I don't like refuge, algae or anything else in the sump, as they do need a low flow, so I had my refuge remote and I pumped water to it and it trickeled back into the display directly.

Steve

I really don't see how that simplified matters.

In example one we have high efficiency.

In example two we have a waste of resources (extra drain noise, additional plumbing, more expensive pump, saltcreep, microbubbles, added heat and noise from larger pump etc. etc.). There is a lot more to it than "to much flow works good but isn't efficient in power used for moving the water." Protein skimmers are designed to maintain bubble stability long enough to deliver the "bad stuff" to the collection cup. Once you lower salinity, add ozone, or decrease the concentration of "bad stuff" you compromise the stability with a net result of premature bubble merging and popping (wet foam). This is why protein skimmer must be sized correctly for the tank and subsequent bioload it is intended for. Installing an oversized skimmer on a tank will often result in poor foam development.

In example three we are giving the skimmer a deficit of "dirty" water, and falling short of the manufacturer's feed requirements. Whatever the shortfall is, will be your rate of inefficiency. Unless you are prepared to challenge the skimmer manufacturer's recommended feed amount for a slower feed, you have a net loss of efficiency. Extra passes through the skimmer with the same effluent water that has just been skimmed is not the same as increasing contact time and does not "skim the hell out of the water".

Your math is incorrect in scenario three as well. The skimmer is fed new "dirty" water at a rate of 80 gallons per hour. Providing you agree that a FIFO system is the most efficient, some of that water will be processed once, while other molecules will be processed many times. There is actually a formula for this in Escobal's book that you disagree with.

How can you assume that you are getting "max pulldown" by reskimming the same water over and over at that unknown rate. Based on a feed of 80 GPH and skimmer pump output of 500 GPH, the water will pass through the protein skimmer 6.25 times per hour, providing your design incorporates a way of telling how many times each molecule has travelled through the skimmer before returning to the display :) A FIFO design can assure that all 80 GPH passes through once, but after that it's entirely random.

Heaters rarely come on in reef tanks, but I still install them nonetheless. Debating their maximum efficiency at the cost of the other equipment you will actually use on a daily basis is moot. There is a difference between manufacturer's directions and scientific formulas for zap rates. If you want efficiency, then use Escobal's formula. Here is a simplified version you can follow. http://www.aquariumadvice.com/articl...ers/Page1.html
A typical 25 watt UV sterilizer works most efficiently at 80 GPH throughput.

Proteins are polar, with one end being hydrophilic (attracted to water/repelled by air) and the other hydrophobic (repelled by water/attracted to air). Protein constantly switch polarity back and forth, so when we collect surface water it is always rich in hydrophobic proteins, or surfactants (surface active agents).

Chaetomorpha and Gracilaria do very well under high flow rates, but it isn't a requirement.

I don't understand why you wanted high flow in your sump? I like the idea of a settling chamber, but it should be slow moving and at the base of the first zone (skimmer/settling/mechanical filtration). Also why not use one pump for the chiller and UV unit? By returning these effluents (UV & chiller) to the beginning of your sump, you are irradiating the water two or more times, and getting a heat gain by not sending the chilled water directly to the tank. It looks like your system is a "water juggler" with a lot of resources expended to complete a simple task. By you description, your sump served two purposes 1) House the protein skimmer. 2) House a settling chamber. The rest of the equipment you mentioned is either a header tank or inline. Why not put the skimmer in the refugium header tank?

Whenever I see this statement written it always means the same thing. you simply do not believe it and your way is the best way and you cannot prove it, you have been offered a number of alternatives with proof yet you want us to believe you rather than have a civil discussion, you are tired of trying to get us to believe that we are wrong.

I really do not care if we confuse people, I am sure if they ask the correct questions they will get the correct answers, and it isn't because we always did it that way.

Even if we confuse people then at least they start to think about it one poster even admitted that he tried something in blind faith and found it to be so good he tells everybody, until he tried it he simply did not believe it.

I answered an opening post about flow trough the sump, I answered with facts yet for some reason because it isn't not what you are doing then it is wrong.

I did call skimmer manufactures, many of them and was given the same answer, over sizing the skimmer or giving the skimmer more water than it requires is a absolute waste of money.

To be fair, the overflow system made sense and I couldn't find any fault in it. Plus, once it was set up I had hard proof it worked, even though I didn't set it up exactly as I should. If someone then told me it was noisy or didn't work I'd probably laugh before I got angry.

But I am not totally convinced about everything in this thread. It is mostly theoretical. Logically it makes sense but there is no way I could prove it for myself. Logically a high sump flow makes sense, too. I read StirCrazy's posts and he has good points. This is why there is debate. In the beananimal thread, you noticed that the nay-sayers eventually fell away. I don't think they will ever fall away here. Countless people have healthy, successful reef and marine tanks using strong flow through the sump.

FWIW, the Beananimal OF system is AWESOME! There, I said it again.

Well my new clownfish are sick so I'm going to read up on curing them now.

Are you technically not doing the exact same thing? Stir crazy has spent hours on his posts and his examples do have some relevancy but you have shot them down each and every time because you believe you are right...

After reading all of this I can quite clearly see that the gentlemen from Toronto have there facts strait. In there FIFO sump model I'm positive you could achieve higher skimmer efficiency, but lets face it folks the simple truth about it is that very few people have set up there tanks to be FIFO style...
If your tank is not set up to be FIFO lowering your turn over to 1X is useless as you're not filtering all of the water. As Mike has pointed out this is all theoretical, there are so many different variables in the equation to the perfect skimmer its redundant to argue back and forth.

Its funny that this article was posted "http://www.advancedaquarist.com/2010/1/aafeature" as it actually downplays the importance of a skimmer in the reef setting:

"None of the skimmers tested removed more than 35% of the extant TOC, leading to the conclusion that bubbles are really not a very effective medium for organic nutrient removal."