River tank set up, advice on flow please?

[Direwolf82]

I got a bug up my somewhere and am going to set up a 34 gallon river tank finally.
The tank will be 3' long 1.5' wide and 1' deep. Dedicated to sewellia, hopefully breeding.
I am looking at circulation pumps now and not sure how much is to much, with the manifold I'm planning I can control flow velocity to a degree. The real question is I can get from 600 gph to 1600 gph pumps. I'm leaning to something around 900 gph, to much? There will also be a HOB filter piped for uniflow, not sure how big that will be at the moment.
There will be gravel and small rocks with some driftwood and larger rocks to give a calm place to hang out in half the tank.The pump and filter will return water onto a sloped bottom on the other side of the tank to stimulate small rapids.
Any thoughts? I know I'm going a little overboard but hey, why not have fun with it. Thanks for any input!!

[Diana]

I would aim for a minimum of 20x, but I see you are thinking more. I do not know at what point the fish end up plastered to the wall. Plants will be just about impossible in something with 30x+, the flow would tear up the leaves.

[atmichaels]

I have two Aquaclear 70's on my manifold in my 40 long (4ft). They're rated at 400 gph so I'm at something less than 800 after factoring the drag from the manifold. I also have a Fluval 406 on the tank which adds another 380 gph. I find this combo & flow rate to be more than adequate while still allowing for rooted plants.

For your size tank I'd be cautious not to put too much power in there and end up with too much push back. So 900 for the pumps may be too high, depending on your manifold design. I'd be curious to see what you're planning.

Austen

[Direwolf82]

Thanks for the reply, after looking at photos of sewellia in they're native environment it appears to be mostly rocky steam bed with little to no vegetation. And fast moving water with small waterfall/rapids so I'm not going to plant this tank, just bare rock, gravel and the driftwood.
The manifold will be a practice in evolution, just tweaking it till I'm happy with it. One inch PVC and couple saws and drills will get me where I want to be, I'll attempt to put up pics when it's done. But just a warning, this is going to be an ugly tank.

[Diana]

No need to glue the PVC, cut, push the fittings on by hand, and test.

[Diana]

No need to glue the PVC, cut, push the fittings on by hand, and test.

[Direwolf82]

Thank you for your input. I am a little too paranoid about water leakage from un-glued PVC joints, I am planning on using silicone instead of PVC glue. The piping will be exterior of the tank and wood floors hate water. Plus being a plumber professionally means I have to attach it somehow. There will be slight back pressure in the piping due to building a bio filter inline with the pump, the wife would have my you knows in a vice if I dropped thirty odd gallons of water in the house due to blowout.( besides friends making fun of the plumber with leaky tanks and pipes.)
I have discovered that water speed is around one yard a second in the hillies natural environment, any smart people know how to extrapolate pump gph into speed? Given the approximate size of the tank I assume this is possible, other variable please ask.
Thanks!!

[Diana]

Do not use silicone. Does not weld PVC pipe. Use the proper PVC cement.

I use PVC with no solvent inside the tanks, I did not realize your plumbing is outside the tank. Probably not reading the post thoroughly.
Definitely glue pipe that is outside.

You can make a test set up, but with pipe outside the tank I would run a test outdoors.

[Bas Pels]

I've glued all my PVC piping with normal PVC glue - without any ill effect.

PVC glue consists of a solvent, solving the PVC on both sides, and then it evaporates, leaving only the PVC. But even if I'm wrong, there are still 15 tanks with glued PVC inside or outside the tank (water change system) which did not give any problems during the last 10 years

@ turning gph into speed - I'm used to liters per hour, but what I do is, I take the inside diameter of the pipe, and turn this into square centimeters (pi * r sqare, a 2.5 cm (1 inch) pipe is 4.9 square cm, or 0.78 square inch)

Then the lph is turned into speed: 100 lph is a column of 10 meters * 10 * 10 cm. The 25 mm pipe, 4.9 square cm goes 20.4 times in that disection, and that means the speed is 10 * 20.4 meters, of 204 meters an hour

I have pumps which pump 5000 liters through such a pipe, where the waterspeed will have to be 10 * 204 = 2040 m/hr. or 0,56 meter a second

I can't explain this in gallon, miles, yards or inches, I would have to turn them into liters, km, meters and so on, doe the maths and turn the resulst into your dimensions

[Direwolf82]

Excellent info. Thank you all who replied. If it has been proven that PVC glue doesn't affect fish and plant life I will go ahead and glue them. I know if I were to sand the fittings and pipe with 80 grit paper the silicone would hold it and not leak but I'd rather use my weld on PVC glue, that stuff is beyond strong. I will be figuring out the flow rates tonight but the project might be on hold for a little while, my temperature spiked to eighty nine in the tank when the ac went out in that room and my hillstreams payed the price, four sewellia and five gastro's died. Maybe in the fall I can order some more and try then. Maybe I'll just build the tank and start letting the algae and diatoms and what not build up before I stock it. Either way I waited this long, not killing me to wait a bit longer.

[Diana]

Bas Pels, there are sites that deal with hydraulics (the movement of water through pipes) in both metric and imperial figures. You can figure out the raw data the way you are doing it, figuring out the volume of the pipe and so on. But water does not work exactly that way. Studies have shown that water in a pipe behaves in a pretty complex way. Further studies have shown that there are certain limits to how much flow is reasonable to put through pipes of certain sizes. Beyond that limit the pump has to work harder, but the amount of actual flow is not much better. So, within a low range of flow rates, you can say that pump will move that much water through that size pipe. But outside of that range, all bets are off.

Basically, in aquariums we are asking how many times in one hour does the volume of the tank get moved (usually through the filter).

10x means 10 times the tank volume per hour. So a 10 gallon tank would have a 100 gph pump. (filter, power head... ) A 10 liter tank would have a pump moving 10 liters per hour.
10x is a generous amount, about right for most community fish, and enough to properly filter the water using a hang on back filter. (not very efficient filter)
20x is a high water flow. A 50 liter tank would have a 1000 lph pump. This is a good flow for river or stream dwelling fish. It is about the highest that is OK with the toughest plants. Of course the plants might be growing in a low flow area of such a tank.

[Bas Pels]

@ Diana

You are completely right pointing out I gave an average value

In a clean pipe, whithout any corners, I understand the water at the walls is almost standing still, and the further one goes to the middle, the fastger the waterflow. In the middle, the flow is much higher than the calculated average.

I do, however, not agree with your rule of thumb for 10 * flow or 20 * flow. Perhaps that is because I think in more sizes. I got a 4 meter tank, and several 1.25 meter ones. Obviously, if I provide a 10 * flow in all of them, the current in the 4 meter tank will be much higher than in the small tanks.

Further, fishes like gouramis originate from the swamps. I wounder whether these will enjoy even a 4 times flow. If I had them, I would not have that much flow.

[Diana]

Gouramis and similar fish (from slow moving water in nature) are generally OK with water up to about 5x, but when the tank is turning over that slowly there are often spots with the water moving even more slowly. If the fish does not like the water movement in one area (right in front of the filter outlet, for example) he can move to another area. I have kept several Gouramis and Angels in tanks with 10x, and they were OK, selecting the area of the tank they liked, but visiting other areas, especially when the food was swirling around. One Angel liked hanging out in front of the power head. I think he found a spot where the water was calmer even that close to the power head.

The definition of 'how many times does the water in the tank turn over' is a valid one for our purposes.

There are other ways to look at this, of course.
People have tried to set up a tank with all the water flowing in one direction as if the tank itself is a section of pipe. It is pretty tricky to do this, usually ending up with some approximation, and it is difficult to figure out how fast the water is moving in the tank in terms of kilometers per hour or related numbers. You are back to figuring out how many times the tank volume turns over per hour. Even that is an approximation, though you can test how efficient a pump can be using a timer and a bucket.

[Loachloach]

Are you looking at manufacturer's stated flow of the devices or the real flow? Because most recommendations about fish and tanks are based on manufacturer's flow. If its recommended 20x, it may mean manufacturer's flow which if not taking in consideration which manufacturer of the devices we are talking about, personal user setup, hose size, etc.., has a real true flow of 10x, which is what they mean in reality.

Personally I am an overfiltering person and currently I have a tank that has no fast water loving fish but has flow per manufacturer's specs, using filters only, of 20x. No one at all is blown away because the distribution is circular/laminar and my "slower water" loving fish are living to the full life spans advertised online despite that, although not ideal for them. But the flow distribution is not turbulent or running in all directions.

[Bas Pels]

Manufacturers always tell what the pump does without any work involved. That is, one could even consider it a theoreticla figure, one can never reach.

Hoses slow the water, bends even more. A canister does the same and believe me, a dirty canistar can, for all practical purposes, be water tight. Therefore a 1000 l/hr pump can, in all reality, pump between 100 and say 700 l/hr, depending on when it was cleaned for the last time.

Personally, I'm quite fond of filtration inside the tank, as this in almost al lcases results in far less wriction for the water