VFD efficiency pump application

[KRAkid]

For a given flow rate is it more efficient to run one VFD controlled pump at 90 percent or two VFD controlled pumps at 45 percent?

 

[jraef]

For a given flow rate is it more efficient to run one VFD controlled pump at 90 percent or two VFD controlled pumps at 45 percent?

90% or 45% of what? Flow? Speed? Power?

Assuming that you most likely mean speed, then probably run one motor at 90% speed. To really know the answer you would have to plot the pump efficiency curves against the motor / drive efficiency curves, but my but feeling says that single pump will win. The pump likely was sized for optimum efficiency closer to the top of it's curve rather than the bottom (most pump curves dont go below some point anyway, so 45% is closer to the bottom, not the middle). Then the added motor losses from having to energize a second machine, plus the added mechanical friction, windage and copper losses in that machine will push it way over the edge of being lower efficiency than just one pump.

But to this, you have to factor in redundancy. One pump may be more efficient, but if it fails and it was all by itself, your efficiency goes to hell in a hand basket doesn't it?!

"Will work for (the memory of) salami"

 

[KRAkid]

I will check the pump curves. Thanks

 

[squeeky]

Dear Jraef, it must be flow as 90% of a 110kW pump will be far more than 2 x 45% of 55kW pumps at the same suction head. Smaller pipe sizes add to the loses. Large motors are more efficent and have better power factors. As you say the key is in the pump curves. I have tried for many years to get a pump supplier to give me a pump curve with a fixed suction head, variable speed against volume. Never suceeded yet. There must be some clever people out there that can do it.

 

[quark]

jraef is right from a very simple fact that pump head is not additive in a parallel operation. This means, for example, if the head requirement is 30 meters at 90% flowrate, two pumps at 45% flow rate should deliver a head of 30 meters each. Here, the power consumption will not follow affinity laws. With constant head, whether 90% or 2x45% is one and same. The other losses are already mentioned above.

squeeky, with variable flow pumping, the operating point generally follows system curve and not pump curve. So, that data is redundant in most cases.

 

[squeeky]

I have just found this. It is specific for filing requirements and not like a manifold that requires a constant pressure like a multiple pivot installation. It states that in a filling application on long to medium feeds where friction head is the major concern the stop start method of filling is more efficent than VSD. Can some please let me know if the up load works. Thnaks.

 

[KRAkid]

Squeeky,
The link worked for me. We are not using VFDs for filling applications, but good to know why.
KRAkid

 

[jraef]

Interesting paper, only SLIGHTLY biased in that Aucom does not make VFDs...

The crux of their argument lies in this statement:

However, there is increasing evidence that for many pumping applications where real-time flow control is not necessary, VSDs are not the most energy efficient solution.

VFDs are for flow control ergo their premise, as applied to the smaller subset of pumping applications that do NOT involve real-time flow control, is accurate. But "increasing evidence"? What the heck does that mean?

Also, this line of "marketing rationalization" is like that which decries how much more energy efficient trains are compared to trucks, as measured by energy consumed per unit of payload. Perfectly good argument, but only as long as your destination has train tracks going to it...


"Will work for (the memory of) salami"

 

[LionelHutz]

I high doubt that 90% and 45% refer to the pump speed.

Squeeky - the pumps would be the same size, not 1 large pump vs 2 half-sized pumps.

I would also expect that running 1 pump near full flow capacity is the more efficient way to operate.

As for the paper. There are many VFD manufacturers who give the example of changing from cyclic filling to continuous filling as a way to save energy.