Fluid dynamics question

[zapf]

Designing my return pump feed system and not sure the best approach for the size of the lines.

Return pump is 1200 GMP at a 3/4" output line. Back of the tank has four ½ return nozzles at the top edge of tank.

Here is the question: What gives me more flow?


1. Keep the ¾ line all the way from sump to each tank return nozzle keeping the size transition as close as possible to the nozzles and just do a ¾ water manifold horizontal at the bottom.?
2. From the pump go into a bigger water manifold say 1 1/2 , run that horizontal then feed straight up to each return nozzle with ½ lines?
3. Run a 1 ½ line up to a manifold running towards the top edge of the tank then transition into each ½ nozzle keeping it as close a possible.

 

[ThisGuy12]

3/4" line will run faster, think of it like squeezing a tube of toothpaste. The harder you push the faster it sprays out of the smaller opening.

Stick with the 3/4" all the way.

 

[zapf]

Thank you. was just over thinking it. I was thinking the head pressure might be less if I keep the bigger line all the way to the nozzle point instead of four 3/4 lines up to nozzles.

 

[ThisGuy12]

Run 2 - 3/4" lines and t them to 2 returns. That will create the most pressure.

Is your end goal to maximize the flow in the tank utilizing the return water or just get the water back to the tank? I'm using 2 - 1" lines stepped down to 4 - 3/4" loc line nozzles into my 280gal tank.

 

[zapf]

The goal is to just get the water back to the main tank from the sump. Planning to run 4 wave makers for the main tank flow but I want the 10 cycles per hour..

The Danner 12 really does not like an head resistance. Thinking I might need to step up to an Danner 18 so I can get the 10 cycles per hour.

My first tests indicated I could get 10 cycles per hour at tank and sump at the same height but then when I moved the tank above the sump " basically just above it" my flow drops to 6 cycles per hour. I really want that 10 cycle per hour.

 

[Gorgok]

To get the most flow use the biggest pipe you can (that is reasonable, on a 3/4 output 1.5" pipe is a almost silly, but 1" or 1.25" would be good). It will be slower (velocity) but more volume (which is what you are after). Use as few bends as you can and don't make it smaller at the end and you are good to go.

Many pump instruction papers say to upsize (by a size or two) plumbing to get the rated flow out of them, it might even be one of the Danner ones that say to use 1.5" on a 3/4" outlet, but i forget.

 

[rjjr1963]

To get the most flow use the biggest pipe you can (that is reasonable, on a 3/4 output 1.5" pipe is a almost silly, but 1" or 1.25" would be good). It will be slower (velocity) but more volume (which is what you are after). Use as few bends as you can and don't make it smaller at the end and you are good to go.

Many pump instruction papers say to upsize (by a size or two) plumbing to get the rated flow out of them, it might even be one of the Danner ones that say to use 1.5" on a 3/4" outlet, but i forget.

Would it be better to use a Y rather than a T to split the return line? Does it matter where you put the split, right at the pump up up higher in the plumbing. My pump has a 1" outlet but the holes drilled in my tank are 3/4".

 

[Gorgok]

It would be better in total volume to not split it at all. One big pipe to just dump the water back into the the tank. Powerheads are way better at moving water around than a return pump, let each do the job they are designed for.

Going over the top is always an option, and capping off (or using as cable pass-throughs) existing holes.

If not, then i would run 1.25" or 1.5" pipe on a 1" outlet up to the level of the holes and split it at that size, reducing it only at the bulkheads. A Y is probably a little better, but plumbing it is more annoying. I would split it up high as well just to make it cleaner. If you are upsizing the pipe having 2 run the whole length is a little silly...

 

[DSMreefer]

I definitely wouldn't go any less than 1.25" pvc. To achieve a flow rate of 1200gh through 3/4" pvc (I assume you meant gph not gpm) you would have to have a pretty high pressure pump. Something to the tune of 20 psi. Not sure what the max ft of head rating is on your pump but most pumps are less than this. You can calculate your pump psi by taking the shutoff feet of head rating and multiplying it by .433. I would run this as far as you can then branch off to your returns. Depending on how many fittings you plan on using this will also affect flow. You may have to go up to 1.5" pvc. There are lots of charts online that will help you calculate friction loss.

 

[rjjr1963]

I definitely wouldn't go any less than 1.25" pvc. To achieve a flow rate of 1200gh through 3/4" pvc (I assume you meant gph not gpm) you would have to have a pretty high pressure pump. Something to the tune of 20 psi. Not sure what the max ft of head rating is on your pump but most pumps are less than this. You can calculate your pump psi by taking the shutoff feet of head rating and multiplying it by .433. I would run this as far as you can then branch off to your returns. Depending on how many fittings you plan on using this will also affect flow. You may have to go up to 1.5" pvc. There are lots of charts online that will help you calculate friction loss.

The Vectra has a head of 21.5' at .433 that's 9.3 psi???

 

[Gorgok]

Those pumps are built for flow not pressure. That is why reducing any extra head ix good.

Same with the other dc pumps out.

 

[DSMreefer]

I agree that you don't want to much head pressure on a pump but you still need pressure to overcome the friction loss. I think the friction loss on 3/4" pipe is 31psi/100ft at 20gpm and 1 1/4" is 2.5psi/100ft. This would also cause high head pressure on the pump. I agree that 1.5" may be overkill. I just threw that out there in case lines were being run, say from a basement up to another floor. I didn't see any measurements on the previous posts. Yes 9.3 psi would be correct.

 

[DSMreefer]

I'm not real experienced with high flow low pressure pumps. Im basing most of my info off of my work in hydronics.

 

[BeanAnimal]

I agree that you don't want to much head pressure on a pump but you still need pressure to overcome the friction loss.

I think you are confused here...

"head pressure" is a combination of vertical head, friction loss, and other losses. The pressure the pump outputs - the head pressure = the flow of the pump at that given ratio...

In other words the pump generates a pressure differential between the intake (suction side) and output (discharge side). Clearly the discharge pressure must be greater than the intake pressure AND the head pressure, for fluid to flow.


Given the above, one must also consider that centrifugal pumps have an operational "sweet spot" where the output per watt is maximized. This can be found by looking at the pump curve for the pump in question. In the real world, one would determine the required output rate, calculate the total head loss and then consult pumps curves to find the pump that is most efficient at the given ratio of total head to required flow. Reefing is not the real world, we just buy whatever is available and concern ourselves with the total needed flow, not hitting the sweet spot (BEP, Best Efficiency Point) on the pump curve. In general, that means we buy pumps, attach big pipes and hope for the best flow we can get, regardless of where it falls on the efficiency curve. So, rule of thumb, use the largest diameter discharge pipe you can accommodate in your setup. If terminate it in a fully closed loop, then the pressure/flow at EACH discharge nozzle will be the same. If you terminate it in dead-end branches, then the closer nozzles will flow higher than the distant nozzles.

Happy Reefing

 

[Reefstarter2]

Head pressure is the same with one line or 4 it depends on height not pipe size I.E. A 4 inch pipe at say 10 feet is about 5 lbs a half inch pipe at same height is still , yep you guessed it 5 lbs

 

[DSMreefer]

I may be a little here but correct me if I'm wrong. The smaller the pipe the more pressure it will take to reach the targeted gph? Of course there are other factors involved that limit flow but I didn't see any length of pipe and a fittings list to calculate it all in. Definitely not trying to start an argument and certainly don't want to being giving out bad info

 

[Gorgok]

I may be a little here but correct me if I'm wrong. The smaller the pipe the more pressure it will take to reach the targeted gph? Of course there are other factors involved that limit flow but I didn't see any length of pipe and a fittings list to calculate it all in. Definitely not trying to start an argument and certainly don't want to being giving out bad info

For the same flow it takes more pressure in a small pipe than it does in a big pipe. That also makes the flow faster in smaller pipes. This part is key to the rest of the issues of small pipes.

Friction is based on length of pipe and speed of fluid, longer pipe or higher speed means more friction. Since you can't really change the length of pipe (except with fittings) you can only reduce friction with slower velocity (ie bigger pipe).

Fittings are equivalent to a length of straight pipe, there are charts that show it. http://www.engineeringtoolbox.com/pvc-pipes-equivalent-length-fittings-d_801.html
Basically sharp 90s are really bad, 1" 90 degree fittings are equivalent to another 5 feet of straight pipe.

Here is a chart that has velocity and friction head for various sizes of pvc listed. http://www.engineeringtoolbox.com/pvc-pipes-friction-loss-d_802.html
You can quickly compare the same flow in multiple pipe sizes to see how drastically it can increase head by using pipe too small.

For example, if you were getting 1200 gph with 3/4 pipe you are going to add 72.3 of head for 100 feet of pipe, or 7.2 feet of head at 10 feet of pipe (which is reasonable to consider with a few 45s and ~6 feet of straight pipe).

If instead you used 1" pipe the additional head would be 21.8 feet at 100 feet of pipe, which is still pretty bad. At 1.25" you are down to 5.6 feet of head at 100 feet, or 1/2 a foot of head at 10 feet of pipe. Much more reasonable.

Now those aren't directly relatable like that, since a pump capable of that much head in the larger pipe could flow more, or a pump incapable of that much head would not flow 1200 gph in such a small pipe. But for quick comparison sake its handy...

 

[DSMreefer]

That is what I was trying to say I guess but may have my terminology mixed up, I apologize. Correct me if I'm wrong but you should be able to convert feet of head to pressure by multiplying it by .433? Wouldn't that be the same as calculating the friction head?

 

[Gorgok]

For our purposes having head loss, friction or static, in feet seems better suited to normal use though. If you want it in other units convert away, but i don't see a point. The conversion is p = 0.433 * feet of head * specific gravity if you want to be really accurate.

 

[BeanAnimal]

Head pressure is the same with one line or 4 it depends on height not pipe size I.E. A 4 inch pipe at say 10 feet is about 5 lbs a half inch pipe at same height is still , yep you guessed it 5 lbs

You have to be careful with your terminology, but in general you are not correct, as you are missing the key concept of total head.

You are referring to "vertical head" or "static head", the pressure, noted in feet of column, that is exerted on the pump. That pressure for seawater is .44 PSI per VERTICAL FOOT at standard pressure/temperature.

The "static head" or "vertical head" is only one component of the "total head" that the pump has to work against. The friction, turbulence and other losses caused by the suction and discharge plumbing also must be calculated to determine flow. So the size and shape of the plumbing MOST CERTAINLY have an effect on the rate of flow.

When you are looking at a pump curve or specification, the "head" is usually listed on the Y axis and is expressed in TOTAL HEAD, the combination of all resistance to flow for the pump, not just the static (vertical) head that you mention in your example above.