New Thoughts on "Adequate Flow"...

[Superstretch18]

Well, maybe not new, but something I got to thinking about recently, fiddling with the flow on my tank.

The question is always coming up, "how much is enough flow?" The normal answer is "### gph/gallon tank size." That's what I've always gone by, but with the ever expanding variety of different pumps, powerheads, wavemakers, etc., the "rules" are getting blurred.

This became very clear to me when I was playing with my closed loop system this weekend. Just set it up and was fiddling with return nozzles. The return pipes are 3/4" hard pvc, so I was using 3/4" 90 degree elbows for the return. This resulted in a "sandstorm" effect, that literally churned the water. I also found that changing the return to a 1 1/4" elbow resulted in a dramatic decrease in the velocity of water moving through the tank - i.e. no more sand storm. Obviously, no change in gph (or if anything, an increase), but a marked reduction in the speed that water was moving through the tank.

Well, this leads me to the thought that maybe we are looking at "adequate flow" the wrong way. Granted, on some level gph is important, but I'm thinking that velocity or psi would be a better measure of what the appropriate flow is for LPS, SPS, etc.

The best way that I can visualize this is by a pinwheel:

You can spin a pinwheel by either a breath of air or a gust of wind. The actual volume of air being moved is an order of magnitude higher with a gust of wind, but until a specific velocity is reached, the pinwheel won't spin. In the case of a pinwheel, velocity is much more important than the actual volume of air moving. I'm starting to think that in terms of a reef tank, the corals are like pinwheels.

So if my goal is to move water at a certain velocity in the tank, the next step is how the heck to measure average velocity in an aquarium setting? While I was writing this post, I stumbled upon this article comparing rotating return devices that used "test strips" placed in locations around the tank and compared them to similar strips placed on a natural reef. Unfortunately, it doesn't go into more detail as to replicating these strips, but I think the idea is intriguing:

http://www.fishchannel.com/media/saltwater-aquariums/aquarium-frontiers/sea-swirl-device.aspx.pdf

I would reach out to the author (Richard Harker), but I have no idea how to get in touch with the man.

In any case, my reason for putting this out there is to see if anyone knows of anyone that has developed a similar idea or has a better method of measuring velocity in a tank setting. I imagine that gph / diameter or area of return pipe would yield a significant number, perhaps divided by the longest length of glass in the tank (to help account for friction), but I really have no idea. I also have no way to relate this back to the numbers on a natural coral reef. Nevermind requirements for acropora vs. favia vs. goniopora, etc.

So, I put this out to my fellow members on RC for thoughts, inputs, critiques, criticisms and general discussion. Feel free to tell me "you're over thinking this," but I think there is some merit to a change in thinking. I guess that's obvious, since it's my idea...

Anyway, have at it. Thanks...

Oh and for the record, I settled on one 3/4" output and one 1 1/4" output on my CL system...

 

[KarlBob]

Lately, another parameter has been added to the mix: turbulence. Closed-loop returns and impeller-style powerheads like Maxi-Jets tend to create fairly laminar flow. The water comes out of a relatively narrow tube, all going in the same direction. Propeller-style powerheads like Koralias and VorTechs create a much more turbulent flow. They tend to have wide-open apertures, where the water is not channeled as much into a uni-directional stream.

How turbulence fits in with velocity and gallons per hour, I don't really know, yet. Some people have posted their support for maximizing turbulent flow. Others have stated that laminar flow directly out of the powerhead is fine, because live rock and corals will add quickly add turbulence to the stream.

 

[jener8tionx]

I have been thinking about this too. On my 20L I have a K1 and and a K3 on opposite sides of the tank to create a flow around the tank in a circular motion. That is 1250 GPH by their rating or 62.5X turnover. I have a bunch of different corals including an anemone and frogspawn all doing great. Also, there is a 1-2" sand bed.

If this was from a closed loop, everything would be blown apart, but this type of flow allows everything to do just fine.

So I agree that GPH and turnover are very relative numbers.

 

[username in use]

I have been thinking about this very subject for some time. I believe that everyone has been looking at flow from a very narrow viewpoint in the past, and simply placing a gph on a tank and saying put all the pumps in untill you get it, and not considering the efficiency of the flow. I think that you can get away with far less flow than conventional thoughts would allow if you increase the efficiency of said flow. On my 29t (30x12x18) I utilize a 1/2" sea swirl on the side of the tank pushing a measly 550 gph (less than 20x turnover). Everything in my tank is moving, there is a visible guire and every coral visibly changes directions all the time and has fantastic polyp extension. I think the inherent flaw in powerheads, including vortechs, and other stream style pumps, is that they dont move, even a closed loop although it can alternate nozzles cannot change the direction of those nozzles. I dont think there is enough importance placed on changing the direction of the nozzles or pumps to increase turbulant flow and randomized currents. Just my experience from staring and staring at the flow in as many setups as I can.

 

[greenbean36191]

Here's a post from almost exactly a year ago that should address what sort of flow velocities to aim for and a few simple ways we measure it in the field (we couldn't measure GPH there if we wanted to).

http://www.reefcentral.com/forums/showthread.php?s=&postid=12796534#post12796534

 

[Superstretch18]

This is exactly what I was looking for:

<a href=showthread.php?s=&postid=12796534#post12796534 target=_blank>Originally posted</a> by greenbean36191
...Also, there have been a lot of good numbers published for flow rates in nature. Some reasonable ranges are <5 cm/sec in lagoons, flats, and deep forereefs, 5-15 cm/sec on the backreef and shallow forereef, and 15-50 cm/sec on the reef crest. Unfortunately, since most reef tanks are a jumble of corals from different reef zones, it's really hard to say what the desirable or safe value would be.

There are some really low tech but proven methods for measuring these things though. One way is just to use dye, (usually fluorescein, but food coloring should work) and a meter stick (or a ruler for in tanks) and just measure how far the dye moves in a given time period. Another way is to measure the dissolution rate of a solid such as a block of gypsum, plaster of Paris, or even a wint-o-green lifesaver. There have been papers published on how to correlate the dissolution rates of all three to flow speeds. The second method is a little more involved, but it's also usually more accurate. I would check the chemistry forum first to make sure none of these things would cause problems if your dissolved them in a closed system though.

I imagine that you could even use something relatively light and with relatively neutral bouyancy (like a mysis shrimp) and measure the speed as it moves across the tank. I think it's time to break out my stopwatch and ruler!

Actually, now you guys got my gears turning. I wonder if I could apply a 10 cm x 10 cm grid to the front glass of my tank, video tape the velocity experiment and then calculate the different velocities throughout the tank. Maybe even find the "average velocity" of the tank and use that as a barameter. Shoot, that would be a good idea if I didn't have bowed glass...

username in use makes another good point about how a rotating or moving outlet can behave and create flows similar to multiple pumps on a wavemaker, but do so through less "GPH." That does complicate things further...

 

[wooden_reefer]

<a href=showthread.php?s=&postid=15216725#post15216725 target=_blank>Originally posted</a> by KarlBob
Others have stated that laminar flow directly out of the powerhead is fine, because live rock and corals will add quickly add turbulence to the stream.

I believe this to be true.

In smooth pipe the Reynolds number need for laminar flow is about 2100. Irregular shapes will create eddies. IMO

 

[sirrus6]

There is a simple flow measuring device that is reminiscent of Superstretch18's pinwheel design. It is sold by JDC electronics and it is typically used to measure linear water flow velocity in streams, etc. it is basically a propeller whose rotation (as a consequence of water flow) is calibrated to linear flow velocities.


Typical linear flow measurements (in m/s, meters/sec) in my tank are:

1) 6" in front of a Vortech MP40: 0.36 m/s

2) 6" in front of my SeaSwirl return (Gorman Rupp 520 pump, 4' head): 0.53 m/s

3) 12" in front of a Tunze 6100: 0.17 m/s

It's a little tricky taking these measurements in that the #'s will vary by 50% depending upon the angle of the flow stream to the propeller; the max values when flow and propeller are approximately perpendicular are reported above. In any other tank locations where the flow is chaotic due to converging flow streams and object-induced turbulence, the flow numbers vary all over the place.

Is there any particular measurement in my tank that you would like me to make?

Ken

 

[gabbagabbawill]

I would like to see something made that I would like to call a "flow box".

Basically, imagine a large fan about 12" in diameter that is very flat, something that looks like a vacuum impeller:

that is built into a box that will direct and provide steady water movement over a large area, but not amount to a fast current like a powerhead would. This would be a lot like the kind of flow that people try to make in tall seahorse tanks, except instead of going bottom to top, it will go side to side... It will still benefit from turbulence created by tank design and rock because it covers such a large area of flow. Add to that a "random" design, something that pulsates, either by adjusting speed up and down electronically (servo controlled DC motors would be the way to go here), or mechanically, and you may have the perfect natural ocean current simulator. NOCS Wavebox. you heard it here first!

We could design this here and find someone who does injection molding to make a prototype and test it? Anyone want to take up an ultimate online collaboration? The proceeds (if any profit were made selling the design) could all go to reef central. Am I being overly optimistic here? Apparently, there are lots of engineers and other smart people here.... lets make this happen.

 

[Superstretch18]

<a href=showthread.php?s=&postid=15255224#post15255224 target=_blank>Originally posted</a> by sirrus6
There is a simple flow measuring device that is reminiscent of Superstretch18's pinwheel design. It is sold by JDC electronics and it is typically used to measure linear water flow velocity in streams, etc. it is basically a propeller whose rotation (as a consequence of water flow) is calibrated to linear flow velocities.


Typical linear flow measurements (in m/s, meters/sec) in my tank are:

1) 6" in front of a Vortech MP40: 0.36 m/s

2) 6" in front of my SeaSwirl return (Gorman Rupp 520 pump, 4' head): 0.53 m/s

3) 12" in front of a Tunze 6100: 0.17 m/s

It's a little tricky taking these measurements in that the #'s will vary by 50% depending upon the angle of the flow stream to the propeller; the max values when flow and propeller are approximately perpendicular are reported above. In any other tank locations where the flow is chaotic due to converging flow streams and object-induced turbulence, the flow numbers vary all over the place.

Is there any particular measurement in my tank that you would like me to make?

Ken

Hold on for a moment, because I want to convert these to cm/second, so that i can compare to the same scale as greenbean gave for ocean measurements:

1) 6" in front of a Vortech MP40: 0.36 m/s = 36 cm/second

2) 6" in front of my SeaSwirl return (Gorman Rupp 520 pump, 4' head): 0.53 m/s = 53 cm/second

3) 12" in front of a Tunze 6100: 0.17 m/s = 17 cm/second

Based on the numbers above, you should be cranking enough water out to mimic the reef crest. Assuming that the corals in your tank are placed appropriately, of course. (Not doubting...)

I don't think that I need any other particular readings from your tank, but I'd be interested to hear what size tank this is. I've been trying to do my own calculations based on greenbean's numbers by lowering the water level in my sump to put some microbubbles into the tank and then measure flow with a tape measure and a stopwatch and it seems that I'm in the proper range with my current set-up. FWIW, I've encountered the same issues with this method trying to make out flow rates in the turbulent areas of the tank. Actually, since I've been paying attention to it, I'm shocked how turbulent and variable the flow in my tank is, considering that it is all coming from stationary outputs.

I do think that it is interesting to compare the gph rated flow with your readings:

Vortech MP40 = 3,000 gph vs 36 cm/second
Gorman rupp 520 = 1,450 gph at 4' head vs 53 cm/second
Tunze 6100 = 3,170 gph vs 17 cm/second

In particular, how the Tunze and Vortech are rated similarly with regards to gph and have a similar set-up, but drastically different velocities.

Still trying to decipher what all of this means, but the added information is helpful...

And holy cow that JDC flowmeter is $500! I was just going to buy one until I saw the price tag! Very cool unit though...

 

[Superstretch18]

<a href=showthread.php?s=&postid=15256078#post15256078 target=_blank>Originally posted</a> by gabbagabbawill
I would like to see something made that I would like to call a "flow box"...

This actually gets to the crux of the dilemma...

The device you mentioned would probably move vast amounts of gph, but I don't think it would produce enough velocity to maintain sps that are used to reef crest flow rates. Depending on the coral, this may or may not be beneficial...

 

[sirrus6]

Vortech vs. Tunze

Remember, I reported linear flow velocities. The volumetric flow velocities (i.e., gallons per hour) are the product of the linear flow velocity and the cross sectional area of the flow stream. As you get further from the pump, the flow stream will widen, and thus must slow down (smaller linear flow velocity) since a constant amount of water (= volumetric flow) is being pushed out by the pump.

So, 1" in front of the Vortech = 110 cm/s; 3" in front = 73 cm/sec, and 6" in front = 36 cm/s.

Ken

 

[Superstretch18]

<a href=showthread.php?s=&postid=15256366#post15256366 target=_blank>Originally posted</a> by sirrus6
Vortech vs. Tunze

Remember, I reported linear flow velocities. The volumetric flow velocities (i.e., gallons per hour) are the product of the linear flow velocity and the cross sectional area of the flow stream. As you get further from the pump, the flow stream will widen, and thus must slow down (smaller linear flow velocity) since a constant amount of water (= volumetric flow) is being pushed out by the pump.

So, 1" in front of the Vortech = 110 cm/s; 3" in front = 73 cm/sec, and 6" in front = 36 cm/s.

Ken

Duh, thank you. Sorry, I had a moment. I think I really am over thinking this now...

 

[gabbagabbawill]

What I'm proposing is something that could alleviate most flow issues sub "reef-crest" flow, which could be maximised using power heads that are already available to the market.

I'm suggesting something that would be a diffuse flow as soon as it leaves the device, instead of 1-2 feet in front of it.

 

[wooden_reefer]

For the open loop, for a system that has a CL, for the purpose of maintaining the bulk conditions of the water, gases, filitration etc, the flow rate need not be very high, IMO.

May be as low as 4x is good, IMO .

 

[H20ENG]

gabbagabbawill
Could you not make this device by cutting most of the impeller housing away on a large powerhead? It seems I remember some folks simply removing the housing and letting the impeller run.

Or am I not following you. Thanks for the link Greenbean, I knew I remembered a thread with good flow #s

 

[mhltcob]

I use a similar method of flow measurement to what was mentioned. I use a neutrally bouyant pellet food, a piece of eggcrate cut in various lengths to measure distance or a ruler/yardstick, and a camera to capture video of the pellets moving. I than get a minimum, maximum, and average velocity for each area. It is very simple to do, costs nothing (assuming you have a digital camera), takes me less than an hour to do even big tanks (1000 plus gallons), and gives the aquarist a much better idea of how your water moves around a fixed object (coral).

It's good that there are finally starting to be people that reallize how useless turnover is in measuring water flow. Considering how important various flow velocities are in regulating respiration, feeding efficiency/total capture, bleaching resistance, etc. in our corals it is about time. By coming up with water flow measurements we can compare values accurately from tank to tank, and from tank to the reef environment. As mentioned flow velocity will change depending on location within the aquarium. Two aquariums (with identical hardware) with 40 x turnover but different aquascaping might experience completely different flow patterns/velocity. A system with a large piston system or wavebox might have lower turnover but much higher flow than a system with alternate devices. A system with a high volume/low velocity circulation device might provide superior flow conditions for coral health over a higher velocity pump with the same GPH rating.

 

[dots]

Your correct, turnover is like "watts per gallon" and someday will be replaced, but just like the other its a generalization.

Remember, do not fall into that trap trying to equate our small closed systems with that of the ocean, which has constantly moving water from different areas carrying a lot more dissolved oxygen to the coral, as well as toxic substances away.

Most will agree that we have come away from the higher velocities shooting across the tank to a more distributed movement.

I would like to see Vortech for thier next program in which the pump on one side of the tank ramps up to a consequetivly higher rpm it cycles up and down. Conversly, the other(s) do the opposite. This way the general tendency of the water flow shifts left to right much like a tide.

By the way: Flow=Area*Velocity which is a volume flow rate, and a key point in Fluid Dynamics. I see a lot of you making the connection between Area and Velocity, and how they relate and change with repect to one another. :thumbsup:

 

[greenbean36191]

Remember, do not fall into that trap trying to equate our small closed systems with that of the ocean, which has constantly moving water from different areas carrying a lot more dissolved oxygen to the coral, as well as toxic substances away.

Actually, for physiological purposes nature and our tanks are quite comparable- assuming "toxic substances" means metabolic products (CO2, O2, radicals, DIN), which are what water flow is important in removing. The oxygen environment in a typical reef tank and on the reef is nearly identical.

 

[108reefer]

so, a closed loop with an OM 4-way doing opposite corners would be the right setup? to move the tank water in opposite directions every cycle?