Hi Friends,

I have a question regarding the overall efficiency of VFD Driven motors. Normally the VFD efficiency is around 96% and motor efficiency around 94% (around 30kW range).

When we made a measurement for a motor used in pumping system, the overall efficiency fell much below the expected, and as a result the contractor is held responsible for the lower efficiecy assuming the lower pump efficiency. The pumps are KSB, manufactured in Germany and factory tested and confirmed to operate at declared efficiency.

Therefore i feel the lower efficiency is due to the VFD / or measurement method

Test arrangement is as indicated below,

* Pump flow is measured with calibrated electromagnetic flow meter at the pump delivery.
* Pressure is measured using calibrated pressure gauges across the pump.
* Power analyzer connected to the input of the VFD.

Inut power is measured and the VFD efficiency of 96% and motor efficiency of 94% is used for the calculation of overall efficiency.

As the input current to the VFD is not sinusoidal, can the current transformer detect the correct current for calculating the power? Is there any other way of doing this test?

Has anyone of you come across a similar situation?



Paliha

Try checking the efficiency without the VFD in the circuit. Two reasons;
1> This will avoid distortion power factor issues which may not be properly handled by your test equipment.
2> The stated pump efficiency is at rated speed and the efficiency may vary at other than rated speed.

Bill
--------------------
"Why not the best?"
Jimmy Carter

I'll amplify waross' second statement: Pump efficiency varies a lot with speed. See attached from DOE.

Your VFD / motor combination assumed efficiencies are pretty good. Make sure your watt meter is a true rms type, or do waross's suggestion and run the pump across the line. You'll need the pump's curves and an accurate speed too.

Good on ya,

Goober Dave

Haven't see the forum policies? Do so now: Forum Policies

• http://files.engineering.com/getfile.aspx?folder=42e633fc-7671-4f51-9238-cd

In answer to your specific question, yes, harmonics created by the VFD can make conventional metering methods inaccurate for purposes such as this. Many VFDs will display the output kW of the motor, that tends to be very accurate. Then all you have to do is accept the VFD efficiency at face value.

"Dear future generations: Please accept our apologies. We were rolling drunk on petroleum."
— Kilgore Trout (via Kurt Vonnegut)
For the best use of Eng-Tips, please click here -> FAQ731-376: Eng-Tips.com Forum Policies

Can you install a torque meter on the pump drive shaft and determine how much power the pump is actually using?

http://www.binsfeld.com/index.php/products/

waross (Electrical) 21 Jun 12 5:54
Try checking the efficiency without the VFD in the circuit.

There is no bypass contactor- The contractor is not ready to do a temporary wiring. So it is difficult. .

DRWeig (Electrical) 21 Jun 12 8:50
I'll amplify waross' second statement: Pump efficiency varies a lot with speed. See attached from DOE

Onething i forgot to mention- the test was carried out at 50Hz. Therefore the pump speed is as rated. Thank you for the link.

jraef (Electrical) 21 Jun 12 15:41
In answer to your specific question, yes, harmonics created by the VFD can make conventional metering methods inaccurate for purposes such as this. Many VFDs will display the output kW of the motor, that tends to be very accurate. Then all you have to do is accept the VFD efficiency at face value.

We used VFD display (percentage of motor kW) but the efficiency is still low.

TD2K (Chemical) 21 Jun 12 21:44
Can you install a torque meter on the pump drive shaft and determine how much power the pump is actually using?

No there is no space and the shaft is rotating.

You never mention the speed at which the efficiency is tested. Pump, VFD and motor efficiency all drop with load, and load in a centrifugal pump varies at the cube of the speed. So even a slight reduction in speed will move you into the lower reaches of your eff curves.

"Dear future generations: Please accept our apologies. We were rolling drunk on petroleum."
— Kilgore Trout (via Kurt Vonnegut)
For the best use of Eng-Tips, please click here -> FAQ731-376: Eng-Tips.com Forum Policies

Did you mean a slight increase in speed, Jeff?
Such as 3000 RPM rather than 2940 RPM.

Bill
--------------------
"Why not the best?"
Jimmy Carter

You should be able to get the pump's efficiency curve, too. Pressure and flow (system curve) affect the operating point. MEs try to spec near the best efficiency point on the pump curve, but the final operating point is a toss of the dice... While the pump speed may be as-rated, variations in the system are what set the operating point.

Pardon me if you've already considered that --

Good on ya,

Goober Dave

Haven't see the forum policies? Do so now: Forum Policies

What was the operating speed during the testing and what is the rated speed? I expect you were off of the BEP of the pnump and the efficiency won't be as good as expected. I would also suspect the VFD/motor combination might have a little lower efficiency than you are estimating.

There is a good paper on the subject called "Motors With Adjustable Speed Drives: Testing Protocol and Efficiency Standard" authored by Anibal T. de Almeida, Pierre Angers, Conrad U. Brunner and Martin Doppelbauer. It describes VFD and motor efficiency testing at different speeds and loads. I did a quick search and didn't find a good, but try Googling it.

The paper Lionel cites is available at DocStoc or GoogleBooks. It doesn't appear to be a public domain thing. It was published in:

Proceedings of the 6th International Conference eemods '09:
Energy Efficiency in Motor Driven Systems
Nantes, FRANCE, 14-17 September 2009
Volume 1
Editors: Paolo BERTOLDI, Bogdan ATANASIU
EUR 24142

Attached is another relevant paper from US DOE.

Good on ya,

Goober Dave

Haven't see the forum policies? Do so now: Forum Policies

• http://files.engineering.com/getfile.aspx?folder=cead7835-9c91-430a-be71-8b