How do you calculate total Head Loss in Parallel Loops?

[Donl]

Can anyone explain how to calculate the total head loss in parallel loops?

For example I have 3 X 250 foot loops of 1/2" pex=al=pex. If X = the head loss for each individual loop, what is the total head loss if they are connected in parallel? I have been looking for a formula but can't seem to find one.

Don

 

[penfrydd]

I teach irrigation, which includes friction loss in pipes. When figuring normal friction losses, you take the loss per foot of pipe for the amount of flow in gpm times the number of feet of that size pipe.

For loops, you utilize the same friction loss tables, but you use 1/2 the length of the pipe and 1/2 the total gpm. I know it doesn't sound totally logical, but it certainly works.

As soon as flow begins, even though it may appear that the water flow is shorter in one direction, increased speed in that leg causes increased friction, so the water comes from the other direction.

It does get a bit more complicated when telescoping pipe sizes up or down, but the idea is basically the same.

 

[Donl]

I teach irrigation, which includes friction loss in pipes. When figuring normal friction losses, you take the loss per foot of pipe for the amount of flow in gpm times the number of feet of that size pipe.

For loops, you utilize the same friction loss tables, but you use 1/2 the length of the pipe and 1/2 the total gpm. I know it doesn't sound totally logical, but it certainly works.

As soon as flow begins, even though it may appear that the water flow is shorter in one direction, increased speed in that leg causes increased friction, so the water comes from the other direction.

It does get a bit more complicated when telescoping pipe sizes up or down, but the idea is basically the same.

Thanks for your response, but I did not understand your explanation. Maybe I didn't phrase my question well enough. I will try again with an example.


In my radiant floor I have 3 loops of 1/2" pex=al-pex each 250 feet long. I want to connect each 250 foot loop to a manifold so that the three loops are in parallel.

I have calculated that the head loss off each 250 foot loop at 2 gpm is 33 feet of HL. I need to know what the total HL is when the loops are in parallel so I can select the appropriate pump. I know that placing pipes in parallel will reduce the amount of HL, I just don't know by how much. I have a feeling it is very much like putting resisters in parallel in an electronic circuit. Is there a formula?

Don

 

[kabbott]

If the flow through each loop remains at 2gpm for a total of 6gpm then the head would stay the same I would think.
If the total flow is 2gpm divided between the three loops then re figure head at .66gpm.(2gpm /3 loops = .66)

 

[Donl]

If the flow through each loop remains at 2gpm for a total of 6gpm then the head would stay the same I would think.
If the total flow is 2gpm divided between the three loops then re figure head at .66gpm.(2gpm /3 loops = .66)

I did a head loss calculation for a single pipe having approx the same area as the three 1/2" pipes in parallel @ 6gpm and the result was 19 feet of head instead of 33 for the 3 in parallel. Unless I am missing something here I don't think your assumptions are correct here.

 

[kabbott]

There is more friction loss in three smaller pipes than one larger pipe of the same area.

 

[BrownianHeatingTech]

Parallel loops allow you to reduce the flow rate, which in turn reduces the head, but once you figure the flow through each loop, the head loss for the system will be equal to the head loss of the highest-head loop.

Joe

 

[BrownianHeatingTech]

There is more friction loss in three smaller pipes than one larger pipe of the same area.

Yup. Friction occurs based upon the interior surface area of the pipe. A single large pipe has less surface area compared to its cross-sectional area than a group of small pipes of similar total cross-sectional area.

Joe

 

[Donl]

Thanks all for your explanations. I had a hard time grasping this concept, but I do now understand. I have also come to the conclusion rightly or wrongly that when choosing the appropriate pump for a given system most just try and use a taco 007 or grundfos 15-58 even though the graphs would suggest a different pump. With the example I gave .66 gpm through each loop would give a HL of 4.7 From the Taco graphs either a Taco 003 or Taco 006 should be selected. Most similar installations I have seen just use a Taco 007.

Don

 

[BrownianHeatingTech]

Thanks all for your explanations. I had a hard time grasping this concept, but I do now understand. I have also come to the conclusion rightly or wrongly that when choosing the appropriate pump for a given system most just try and use a taco 007 or grundfos 15-58 even though the graphs would suggest a different pump. With the example I gave .66 gpm through each loop would give a HL of 4.7 From the Taco graphs either a Taco 003 or Taco 006 should be selected. Most similar installations I have seen just use a Taco 007.

Using the correct pump will save energy, and also extend pump life. Running a too-large pump can wear the pump out, as well as reducing the performance of the system, in some cases.

Joe

 

[in hot water]

Select a circ that operates in the middle third of it's curve for best results and efficiency. That's the beauty of 3 speed circs available today, you can dial them in, run them at the sweet spot and use the least amount of energy.

hr