Aquatic Plant Forum banner
1 - 20 of 37 Posts

·
Registered
Joined
·
4,333 Posts
Discussion Starter · #1 · (Edited)
OK this is a bit of a book, but I found fluid mechanics to be one of my favorite courses while I was in engineering school. Contrary to much of the scientific world, the behavior of fluids is not at all intuitive. Once you learn it, you possess "secret knowledge" which is kind of fun. :) I'll try to clarify a couple of ideas here that have common aquarium applications - first, the flow of fluid through a pipe, and second, the behavior of a pump.

Concept #1 - Flow of water through a pipe

If you know anything about electricity you know that flow (amps) equals the pressure (volts) divided by the resistance (ohms). Water in a pipe acts the same way. If you want more water you either increase the pressure or decrease the resistance.

The flow of water through a pipe is described by this equation. Don't freak out!!! The math is actually pretty simple.

Q = (pDPr4)/(8hL)

Q is the flow (pick your favorite unit - liters per minute, gallons per hour, etc.) Flow is what we pay for, and what we hope our pumps deliver.
p = 3.14159.......
P = Pressure (psi, meters of H2O, mm of mercury, or whatever)
r = radius of the pipe
h = the viscosity of the fluid - in our case water
L = the length of the pipe

We can't do anything about the viscosity of water or pi, so ignore those. From the equation you can see that flow is proportional to length of the pipe. A 100' garden hose will put out half as much water as a 50' garden hose. Check it out sometime. Fifteen extra feet of tubing will decrease your flow by quite a bit.

The pressure, for our purposes, is produced by the pump of interest. More on this later. Suffice it to say that if you double the pressure, you double the flow. Remember that canister filters are really nothing more than a pump attached to a media container.

Now, the grand-daddy of them all. The one that really matters. Flow is proportional to the radius of the pipe, raised to the fourth power. In english, this means that if you double the radius of your pipe, then you'll increase the flow rate by 16 times!!. Even small differences in diameter will make a large difference in flow rates. You'll get 2.1 times more flow from 3/4" tubing than you will from 5/8" tubing.

The reason that the radius is so important is that it affects the velocity of fluid in the pipe. Areas of high velocity are turbulent, which means the water molecules are crashing around randomly, wasting energy instead of flowing nicely along toward the destination like we want them to. Think of turbulent flow like friction. Even the very large turbines in the power plant at the bottom of Hoover dam are more efficient when velocities are low. This is why the intake and outlet pipes are so big.

Try to wrap your head around this idea too: The flow rate is the same everywhere in the system. If you use 1" intake tubing and 1/2" return tubing, the water simply moves 4 times faster (and has more turbulence) in the 1/2" return line.

Every fitting you put in the line also acts to add resistance. An single elbow will produce as much resistance as several inches of straight pipe. A partially closed ball-valve will act like many feet of extra pipe - dropping flow considerably (duh, that's what a valve is for). Even a completely open valve will introduce a small amount of turbulence in the line, robbing you of energy, and decreasing flow. Add 20 feet of extra tubing, an in-line heater, a UV sterilizer, and a CO2 reactor to a little Eheim 2213 and you'll see that flow rates go to almost nothing.

Why don't the filter manufacturers use huge tubing for everything? Economics. Big tubing is expensive, and isn't easy to work with. They compromise, and we're stuck with whatever they decide to produce. For the most part, the compromise is reasonable.

Concept number 2 - how a pump works:

The flow going into a pump is exactly equal to the flow coming out of a pump (conservation of mass). So what, exactly, does the pump do? It raises the pressure of water in the output line. The pumps that we use are really horrible at developing suction on the intake line. They require a constant supply of fluid, and if it comes in already under a bit of pressure, so much the better. This is why Eheim recommends that their filters be located below the tank. The water column in the intake line acts to "preload" the pump. If you place the filter above or level with the tank, the pressure inside the canister is negative. This will lead to impeller cavitation (noisy and shortens impeller life), and gasses will come out of solution (we'd like the O2 and CO2 to remain dissolved).

Every pump has a pump curve associated with it. The shape of the particular curve is dictated by the design of the pump (rpm, impeller shape, intake & outlet geometry, etc). Eheim is kind enough to publish theirs. For an example, I'll reference the 1260 hobby pump (orange line), which I'm using in my new setup.



You can see from the graph that the two variables are H and Q. We already know that Q is flow. H stands for head. Head is basically a measure of resistance but it's a little hard to explain. Let me try.

Eheim will tell you that the rating for the 1260 pump is 40 liters per minute, or 635 gallons per hour. This is true for a zero-head condition. Zero-head means no resistance - no attached tubing, no valves, no strainers, etc. Nothing. Squirting out into thin air. By the way, the flow ratings for canister filters are given for zero-head conditions too - no media, no tubing, no bends, and no strainers. Actual observed flow rates are much lower due to the resistance of all the stuff we hook up to the pump.

We whine and complain that our 1260 pump only produces 20 liters per minute of flow. From the graph we can see that we have about 2.0 meters of head. This is how much flow you'd see if the pump were delivering water to a location 2 meters higher than itself (like pumping up out of a sump). If you try to use this pump to deliver water up three flights of stairs - forget it. Once you get up to 4 meters, the flow rate goes to zero. We call this the shutoff head. This maximum pressure that this pump can deliver is equal to 4 meters of water or about 5.7 psi.

Here is the cool part. The entire resistance of your system (tubing, filter media, elbows, strainers, spraybars, gunk in the lines, etc.) can be represented by a single value for H. If your 1260 pump is only putting out 20 liters per minute, the resistance (friction) in your system is equivalent to pumping water straight up a 2 meter pipe. You can decrease H (and therefore increase Q) by using fewer fittings, shorter sections of pipe, cleaning the filter media, or most importantly, by using larger tubing.

One last thing to remember. Don't forget that for a canister filter, or an in-line pump, the water in the intake tube will help to preload the pump. Think of this example. If the level of the water in your tank is 2 meters above the pump, the pressure on the intake side is already equal to 2 meters of H2O (about 2.8 psi). The pump will add up to another 4 meters of H2O (5.7 psi), for a total of 6 meters (8.5 psi). No matter where it is located (within reason), this particular pump will be able lift water up to 4 meters above the level of your tank.
 

·
Registered
Joined
·
367 Posts
:faint:


My head just asploded. You owe me a new one.
 

·
Registered
Joined
·
342 Posts
that's really cool.
I've always found the more you know about how something works the better you are at using it.

It makes sense now why the hose on my new eheiem is bigger on the intake side than the exhaust side.

I do have one question though what about the shape of the hose, does that make a difference.
I mean the fluval ribbed hose and the eheim smooth round hose.
 

·
Registered
Joined
·
2,253 Posts
Moo said:
I do have one question though what about the shape of the hose, does that make a difference.
I mean the fluval ribbed hose and the eheim smooth round hose.
I believe the shape of the hose will make a difference. For the fluval hoses, I think the ribbing is to keep it from kinking. At the same time, that's really gotta put some turbulence in the fluid flow as well as providing a place for stuff to get stuck and clog the hose.
 

·
Registered
Joined
·
2,499 Posts
A decent book on the subject is Aquatic Systems Engineering: Devices and How They Function by Pedro Ramon Escobal

An internally ribbed hose will flow less then a smooth one. Internal ribs increase turbulence and have a greater surface area. Fluid molecules touching the surface of a hose flow slower then those in the center due to surface tension. Thats very simplified but it gives you the idea.
 

·
Registered
Joined
·
4,333 Posts
Discussion Starter · #8 ·
Ribbed tubing does increase the turbulence, but not by as much as you might think. Like gnaster said, in a long section of straight pipe the water molecules don't all move at the same velocity. The molecules at the center of the pipe move very quickly while the ones right up against the wall don't move much at all. Think of a big river. The flow is much faster right in the center.

Things that interrupt the flow of the high velocity stream are the ones that really add lots of resistance. Think of a partially closed ball valve or an abrupt turn. In the fluval tubing the water velocity in the region near the ribs is quite low, producing a modest amount of turbulence, but not so much that it defeats the pump. The net result is some wasted energy, but much less than you'd get from a kinked tube.
 

·
Registered
Joined
·
4,333 Posts
Discussion Starter · #9 ·
Moo said:
It makes sense now why the hose on my new eheiem is bigger on the intake side than the exhaust side.
Well, yes, and no........

The reason the intake hose is large is to ensure that negative pressure isn't developed somewhere on the inlet side. In and of itself, this is a good thing. What bothers me is that they get cheap on the outlet side. You don't really need outlet tubing that is as large as the inlet side, but you would get better flow trough the system.

Incidentally, Eheim's overall concept of the canister filter is that flow velocity should be modest to allow for good contact time with the filter media. This design allows bacteria a chance to do their thing. The flow rates for most of their canisters are pretty modest compared to some of the others. One benefit of lower flows is that very little nose is produced, which has always been one of their biggest selling points. For a planted tank, mechanical filtration is probably more important to us than biological filtration, since the plants take care of much of this. We're also concerned with an even distribution of nutrients (CO2) around the tank, which often requires supplemental flow.

I'd really prefer bigger tubing on their systems since it would result in a large improvement in flow rates, but then maybe the biological properties of the filter would be degraded and the thing might make a bit more noise.
 

·
Registered
Joined
·
1,002 Posts
Very nice! Thanks for taking the time to summarize that! :)

It explains why the pump in my plant filter is cranky sometimes with starting back up after I'd had it unplugged. Not enough water pressure on the inlet side. I really should plug the holes where the bulkheads are, set it lower and convert it into a sump.

I also have a couple of 100' denitrification coils (using drip irrigation tubing) set up that are being run by a powerhead. I have heard clattering from the impellor on that one and it's a fairly new powerhead. Perhaps cuz the outflow is so restricted? There's also a couple of foot of head on it.
 

·
Premium Member
Joined
·
1,786 Posts
I LOVE IT!!

By the way, any fun analogies using arteries and veins with healthy human blood and the flow dynamics after eating McDonold's or Dunkin Donuts versus some apples or something healthy?
Or before and after a good work out and drinking Gatorade?
:happy: :bounce: :pop2: :cool: :bump2: :tongue1: :rofl:

Sorry, but this guy is really smart and his writing style is really easy for me to understand.......

And I like to tap into his brain with his occupational experience.....

And I have gotten hooked on CSI....

And my humor is at times whacked.......

So how would Men's Health Magazine spin this.......?

Any takers?
 

·
Registered
Joined
·
52 Posts
Excellent thread, this explains why using pipe cleaners has such a dramatic effect.

guaiac_boy said:
I'd really prefer bigger tubing on their systems since it would result in a large improvement in flow rates, but then maybe the biological properties of the filter would be degraded and the thing might make a bit more noise.
Would not increasing the radius on the outflow pipes result in a lower 'head' for the pump? Or rather, would larger pipe result in a more noticeable drop in flow when the head is increased, and that the line on the graph would be a lot 'flatter'?

I also thought one of the reasons that Eheim use smaller hoses on the return side was to maintain a stronger current (as opposed to flow) out of the return pipe / spray bar and into the tank. A larger pipe may flow more water in, but the strength of the current would be less. Or am I missing something? (sorry I did not study any of this at school).
 

·
Registered
Joined
·
4,333 Posts
Discussion Starter · #13 ·
The pump curve is determined by the physical characteristics of the pump, ie impeller speed, voltage, blade size, casing geometry, etc. Changing the attached tubing will take you to a different point along the curve, but it will not alter the basic curve in any way.

"Current" is the same as flow. You will get a higher velocity stream with a narrow tube (or by using a small outlet nozzle) but using a tube of smaller diameter will not in any way create more current or flow.
 

·
Registered
Joined
·
455 Posts
One big reason the outlet hose is smaller is that water is heavy. Reducing the hose diameter just a little bit relieves the pump of a lot of work. Having a larger input hose ensures that water is always feeding the pump at least as fast the the pump can push it out. If your pump can't push the water as high as you like, just reduce the diameter of the hose. I simplified it a bit, but it's close enough.

I only point this out becuase the discussion has been entirely about resistance, which is caused by the plumbing. Gravity is a factor too and must not neglected.

Gravity is one place that the "fluids is like electricity" analogy does not work.
 

·
Registered
Joined
·
4,333 Posts
Discussion Starter · #15 ·
JERP said:
One big reason the outlet hose is smaller is that water is heavy. Reducing the hose diameter just a little bit relieves the pump of a lot of work. Having a larger input hose ensures that water is always feeding the pump at least as fast the the pump can push it out. If your pump can't push the water as high as you like, just reduce the diameter of the hose. I simplified it a bit, but it's close enough.

I only point this out becuase the discussion has been entirely about resistance, which is caused by the plumbing. Gravity is a factor too and must not neglected.

Gravity is one place that the "fluids is like electricity" analogy does not work.
At first glance, your logic seems correct, but in fact, your statements are in error.

You've fallen into a common trap that fools a lot of people. Assume that you have a two tubes that are each 10 feet high. One is 100" in diameter and the other is 1" in diameter. The water pressure at the bottom of the tubes is exactly the same in both. The height of the water column in the tube is important, but the diameter of the tube makes no difference in a static condition. Assuming that the pump is at the bottom, given enough time, the Eheim 1260 pump has sufficient power to fill up a million gallon reservoir to a depth of 4 meters (its shutoff head).

Now, if you take the more common application of an aquarium return line, we want the water to be in motion. The pump has to overcome two forces to be able to move water. The first is the height of the water column. If you want to pump water 2 meters uphill (out of a sump for instance), the pump will do no better than the outflow shown on the pump curve for a 2 meter head. The second force to be overcome is caused by the resistance to flow. Resistance to flow picks up dramatically with narrow diameter tubing because high fluid velocities create turbulence, which robs the pump's energy. This produces additional head that is additive to the first. For a given condition this might create a total of 3 meters of head, dropping the Eheim 1260's output to roughly 10 liters/minute (from the graph). Exact numbers depend, of course, on the particular setup.

Most canister filters don't have to pump uphill since the intake and return come from the same body of water. The height of the water column in the intake line is exactly the same as the height of the water column in the return line. The only force to be overcome is the resistance created by the filter media, tubing, connectors, strainers, and such. The most important variable, by far, in the flow equation is the radius of the water conduit, since the resistance is proportional to the velocity squared, and the velocity is proportional to the radius squared.
 

·
Premium Member
Joined
·
2,359 Posts
guaiac_boy said:
...

Most canister filters don't have to pump uphill since the intake and return come from the same body of water. The height of the water column in the intake line is exactly the same as the height of the water column in the return line. The only force to be overcome is the resistance created by the filter media, tubing, connectors, strainers, and such. The most important variable, by far, in the flow equation is the radius of the water conduit, since the resistance is proportional to the velocity squared, and the velocity is proportional to the radius squared.
I'm reviving this thread because it has a direct application to a new tank that I'm setting up. The tank will be a 1024l (270g) tank. The "filter room" will be about 5 meters away (same floor) with piping running under the floor to and from the aquarium. The current plan is to use piping that has a diameter of 32mm (1/8"). Questions:

1. Given the statement quoted above, does this mean that I don't need to take into account the 5 meter (16.4') distance when deciding on the pump (if I use a sump) or cannister?

2. What will the effect be of having two outlets from the tank, merging into the 32mm piping and then splitting again in the filter room to let's say two cannisters. The output of the cannisters would then merge in one 32mm return pipe under the floor and then split again into two returns into the tank?

And no, I'm not complicating my life because I want to ;) ; the location of the tank is the problem. It will be sitting on a base that's only about 20cm (7.9") high. The water depth will be 75cm (29.5") so the water level will be 95cm (about 37.4") from the floor. The only place to put filtration is as described.

And secondly, what would you guys recommend: a sump or a couple of cannisters (I was actually considered 2x Eheim 2128 and a 2080 but this seems really complicated, especially as the 2080 has two inlet tubes and one outlet). It is not yet decided whether the tank will be an African Cichlid tank or a planted tank...
 

·
Registered
Joined
·
4,333 Posts
Discussion Starter · #17 · (Edited)
An interesting series of questions. If you use large enough tubing, the length is almost negligible, within reason. 5 meters isn't too far provided that the velocities in the tubing are low. You frightened me for a minute when I read 1/8" tubing. I assume that 32mm is correct which would be 1.25" tubing, a much better choice. That size should be plenty large for a 270g tank.

My Eheim 1260 pump has an effective flow rate of about 300 gallons/hour. In my closed-loop I use 3/4" (19mm) PVC pipe which gives a flow velocity of about 3.8 feet/second (1.2 m/s). Even though it doesn't seem like it, my loop actually uses about 20 feet (6 meters) of pipe and passes through a myriad of tees, elbows, and valves, each of which adds resistance. Even with all this, if anything, the flow is perhaps a bit excessive.

In your application, the Eheim 2080 has a rated flow of 1,700 liters/hour and the Eheim 2128s are rated at 1,050 liters/hour each. Combined into a single inlet or outlet, the total is 3,800 liters/hour. In actual practice, with strainers, media, tubing, valves, etc. the the true flow will be much less - I'lll assume 2,000 liters/hour to be generous. From the equation Velocity = flow / cross-sectional area, expected velocity in the tubing will be 0.7 m/s. I *think* (gut feeling) that this will be low enough that you should be just fine.

Second question - what is the effect of using a single inlet & outlet in a manifold-type setup? This one worries me a little because the pump curves for the two different canister filters are different. Using a single tube for the pump intake side should be fine. The height of the water column in the full tank should be high enough to avoid a negative-pressure situation. Combining the pump discharges worries me because the delivery head of the 2128s is 2.0 meters while the 2080 has a delivery head of 2.6 meters. What this means is that the 2080 can develop more pressure than the 2128's. If the pressure in the return line is greater than 2.0 meters of water, the flow out of the 2128s will be zero, or they might even flow backwards, having been overpowered by the more powerful 2080. In actual practice I'd guess that the pressure in your return line would be much less than this unless you had a valve partially closed or had an obstruction. Still, the Pro-II's (2128s) will be working against the pressure generated by the Pro III (2080), and their flow will be less than you'd see if they had their own return lines.

Is a single 32mm (1.25") intake and three individual 19mm (3/4") return lines an option? Sized like this, velocities in each tube would be roughly equal.
 

·
Premium Member
Joined
·
2,359 Posts
Great response... thanks!

Yes, 32mm is correct, not sure how I got 1/8" from that!

So if I understand you correctly:

- The 5 meter distance shouldn't have that big of an impact on flow rate.
- Water from the tank to the filter system (two outlets from the tank joining into one 32mm pipe to the filter room and then splitting into several filters, if using cannister filters) should work in this setup.
- The return of the filtered water may be more problematic from multiple cannisters. I'm not sure that three return lines is possible but will check.

Did I get it right?

I think I'll just go with a sump, what does everyone think? In that scenario: dual lines from the tank joining into a 32mm pipe that leads 5 meters directly to the sump. Then a pump (Eheim 1262??) returning the filtered water through a 32mm pipe which then splits into two for two returns into the tank. Should the two lines going from and to the tank be about half the diameter of the 32mm pipe?

And will an Eheim 1262 give me an adequate flowrate in that type of setup? (it's rated at 3400l/hr and a max head of 3.6m).
 

·
Premium Member
Joined
·
2,359 Posts
hhhmmm...

I just realized (if I've understood all this stuff correctly) that in the sump configuration, the 5 meter distance back to the tank *does* matter as you won't have the input pressure from the tank... right?

So in this case, would an Eheim 1262 be enough to keep up with the water flow coming into the sump?

I think I need to get a book on this stuff! :), But in the meantime... help!
 

·
Registered
Joined
·
2,253 Posts
Yes, in a sump the distance will matter a little. But I don't think it's significant at that point. Static head will be the killer here.

Then again, your flow into the sump should be directly proportional to the output of the pump. If the pump puts out less water, there's less water to overflow into the sump. This is my theory and someone with more experience on sumps will need to correct me if I'm wrong.
 
1 - 20 of 37 Posts
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top