Parallel work of fans vs. pumps

[mechWatt]

So, I have a quick question and would be grateful if someone could explain me the difference.

When you have 2 centrifugal pumps working in parallel (with same head and flow), resulting operating flow is not two times the flow of a single pumps, but somewhere in between (with head staying the same?). Example: Q1=100 m3/h, Q2=100 m3/h, resulting Q is 100<Q<200.

When you have 2 ventilation fans (with same pressure and flow), resulting operating flow IS two times the flow of a single fan. Example: Example: Q1=10 000 m3/h, Q2=10 000 m3/h, resulting Q is 20 000 m3/h.

At least, that is what I figured out so far from the literature.

Could someone explain me why this is the case? As far as I get it, both pumps and fans have the same operating principle, it's just the fluid that is different.

Does this happen because pressure loss in an air system is much lower than the pressure loss in water pipes, and so, the system characteristics (curve) for air is a lot less steeper than the one for water, so the resulting operating point is close to the double flow?

 

[LittleInch]

Well without knowing the system you're looking at or how you're able to make those statements it is not easy to say.

Dual centrifugal units will jointly operate at a slightly higher head than a single unit because they are further left on the pump curve in order to create more flow than a single pump as system frictional increase.

If the system curve is very flat for a fan driven system then it might appear as if 1+1=2, but it won't be. It might be 1.9 or 1.95, but it won't be 2.

What "literature" do you refer to - that might help us understand how someone is seemingly able to bypass the laws of physics.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[GT-EGR]

Fans and pumps follow the same rules when working in parallel. I can't tell if you came to this conclusion by actual measurement of two parallel pumps/fans, or through some textbook.

If I had to guess, in your pump scenario the reason why you don't see flow double with the second pump coming on is because adding that additional flow also increased your system head. With a single pump on, and the system seeing a lower total flow, the single pump encounters less head since your velocity through all your valves and fittings is lower than your worst case system head. So you end up operating on a different part of your single pump curve - further to the right. If you plotted your system on the single pump curve, you would see that it intersects at a point that is more than 50% of your GPM.

Once you bring the second pump on, then both individual pumps see the maximum head (because your velocity through all your distribution is now higher), which likely corresponds with the flows that were selected for.

As for your comparison fan system, you may be observing this if the ductwork is very oversized, and doubling the flow adds a negligible amount of static pressure loss. But otherwise it should follow the same rules mentioned above.

 

[mechWatt]

Basically, it's not really a conclusion that I came to, it's more of what I've been told.

I just wanted to make sure that there is actually no differrence between pumps and fans when it comes to parallel work mode.

My problem is that I have a smoke exhaust system that has a workflow of 50 000 m3/h and it has to be comprised out of two fans with 50% flow each, which means 25 000 m3/h each.

When they are both working (100%), considering the above stated for fans in parallel work regime, it will not add up to 50 000 m3/h, right? That means that, even tho my fans are supposed to be 50% each, I need to make them have more than 25 000 m3/h each so when they are operating together, they will make sure to extract 50 000 m3/h.

 

[LittleInch]

Correct.

You need to find out the required pressure for 50,000 m3 then that becomes the duty point for your fans @ 25,000 each.

It means that if you run one fan you'd probably get 30,000 out of it, but if the design is 2 x 50% then just design it that way.

If these are exhaust fans (i.e. at the end of the ducts before the open air) then the amount of vacuum they can draw it normally quite low. It might not be the same thing as if they are at the start of the duct.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[mechWatt]

@LittleInch Just for final clarification:

Design calculations determined that at 50k m3/h, system has a pressure loss of 400 Pa.

So, if I pick two fans with 25k m3/h @ 400 Pa each, it means that in total, when both fans are operating, I will be getting 50k m3/h?

And when only one of them is in operation, since air flow is lower, meaning that the air speed thru the ducts is lower (meaning that pressure drop is lower), I will be getting a new operating point, with increased flow (>25k m3/h) at reduced pressure drop (less than 400 Pa)?

I know I am repeating myself a lot know, but I just want to be 100% sure that I am not making a (costly) mistake.

 

[LittleInch]

So, if I pick two fans with 25k m3/h @ 400 Pa each, it means that in total, when both fans are operating, I will be getting 50k m3/h?

Correct

And when only one of them is in operation, since air flow is lower, meaning that the air speed thru the ducts is lower (meaning that pressure drop is lower), I will be getting a new operating point, with increased flow (>25k m3/h) at reduced pressure drop (less than 400 Pa)?

Correct - the extent will depend on the particular fan curve.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[GT-EGR]

Do your calculations as if it is one fan. And then when you come up with your pressure - 400 Pa you said, all the fans whether it is one, two, or three, have to all be sized for 400 Pa and either 100% or 50% or 33% of the flow.

The pressure of all fans have to be selected equal and at the max/design pressure, the flows are additive - this is how parallel pumps and fans work

If you plan on operating a single fan at times, you also have to check to make sure your system curve falls on a stable point on the curve of a single fan operating. Depending on what manufacturer you are using some have good functionality that allows you to do this check very easily.

And a follow up to my original post applied to this situation - when you plot your system curve on the curve of a single fan running, you will find that you will get more than 50% of your design flow - because your system resistance isn’t 400 Pa because you are pushing less flow through the same duct system

 

[LittleInch]

The only thing to add is that it is rare for two items to be truly identical. Choose fans with a fairly steep curve at the duty point to avoid one running harder than the other if you've got fixed speed motors.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[capliard]

The rule of thumb : a single fan will produce about 66% of flow VS twin same 2 fans in parallel.

Rule ot thumb no2 : A loss of a fan in a 3 fan parallel arrangement will leave you with about 90% of full flow.

To have real answer, look at fan curves vs system curve.

CAPL

Ingenieur Minier. QuTbec, Canada.

 

[EnergyProfessional]

you still need to consider system pressure drop.
The pump curves add. In parallel they add along the flow axis, in series they add along the pressure axis.

 

[ChasBean1]

One pump might produce 500 gpm at 100’ head. When you run two pumps in parallel in the same system, you might see 800 gpm at 115’ head. Pumps in parallel do not generally produce a doubling of flow.

One fan might produce 1,000 cfm at 3” total static pressure (TSP). Two fans in parallel might produce 1,500 cfm at the same TSP.

The academic concept is that fans or pumps in parallel double the flow; fans or pumps in series double the pressure. These are academic rules; neither is entirely accurate but represents the general idea—parallel devices have a larger effect on flow and series devices have a larger effect on pressure.

 

[KesTheHammer]

Fans and pumps follow the same rules.

The question of how much you "lose" by adding fans/pumps in parallel is a direct result of fan/pump curves being superimposed over one another - the flow gets added onto the X axis, but the pressure not on the y-axis.

If you are operating near the top of the curve, the amount of flow you get extra is a lot less than if you were operating near the bottom right.

Image on pg 27



Kestell Laurie
South Africa

 

[LittleInch]

Errr, Care to explain what the coloured lines are?

assuming the purple line is the system curve why does it change so dramatically? It should be the same otherwise you're not comparing like with like.

Why is the black pump curve line the same in both charts?

This is the opposite of what you normally get where the second identical parallel pump increase flow, but doesn't double it.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[LittleInch]

Your purple line is wrong / different. All of a sudden it becomes a different shape and at no flow goes from 2 ft resistance to 48 ft.

It's more normal to keep the system curve the same and elongate the 2 pump curve to show what 2 pumps will do ( identical just double the flow for the same head, if different pumps find the same head and add the flows together.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[LittleInch]

engr paper

This is supposed to be serious engineering forum.

You can't just go round posting data and curves which are incorrect, don't confirm to reality but just happen to be close to the answer.

Engineering doesn't work that way.

As to being correct, you are probably not far off, but my extrapolation of the first system curve puts 72ft at less than 200 (gpm?), but your second graph would be 210 assuming they are two identical pumps.

I suggest you get the two curves to use the same system data, are correct and then repost and delete your earlier incorrect versions.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[lilliput1]

The performance curve of parallel pumps and fans would show double the flow at each pressure however actual parallel fan or pump operation would be at the intersection of the combined pump head and the system head. You have to draw the system flow/pressure curve to find the intersection. You can draw the system head at various flow by using the principle that the pressure varies as the square of the flow.