Induction Motor Efficiency at Reduced RPM

[condesinc]

Hello all:

Bit of intro: I am a gyro gearloose (mechancal) as opposed to a wireloose, but I have a lot of past experience in controls/electrical design/analysis etc, so if the reply is not too technical, I can probably follow.

A source of great frustration for me for several years has been the inability to get estimates of induction motor efficiency and power factor for motor loads less than nameplate and especially at less than full rpm. Replys from salemen to design guru's are always silence, audible head scratching, or someone saying that someone looked at that years ago and they don't remember what the results were.

Motor manufacturers always give eff and PF numbers for (usually) half to full motor loading, but it is always at full rpm. However, the whole point of an inverter drive system is to run at something different than nameplate line frequency/speed.

We have an application in the chemical process industry where knowledge of the process torque demand is extremely valuable for design optimization. Motors that drive the process (through gear reducers...process is typically from 2-4 rpm final) run from 30 hp all the way to 450 hp, and I am designing a system now requiring 750 hp. I have data for past installations, but a lot of it is only motor nameplate, and process volts, amps, and frequency. We also have rpm, but the method of measureing is probably not as good as recording freq and estimating slip. Recently, through site testing, we have discovered that some makes of inverters will give an accurate indication of electrical torque, and I have some limited data on this for two larger systems, which solves the PF problem for these, but still leaves the efficiency issue. Motor manufacturer's are willing to test at a (hefty) price, and there are devices that can measure torque (of which I may have succeeded in getting a client to pay for on a new job). But,all this does not solve digesting the past data where some real value lies.

I realize that motors vary all over the map, but is there someway to at least estimate, with range of uncertainty known, the change in effeciency and PF at less that full motor rpm's? Are there rules of thumb? Are there models...and if so, what needs to be known to trim them?

thanks in advance!
John

 

[dpc]

Are you talking about motors running on VFDs? Operation on a non-sine wave also reduces the efficiency of the motor somewhat.

Power Factor will not be an issue on a VFD.

David Castor

http://www.cvoes.com

 

[LionelHutz]

Is this the type of info you're looking for?

thread237-273230

The paper link posted by BigInch has some good data.

 

[jraef]

The more I read that paper the more I like it.


"If I had eight hours to chop down a tree, I'd spend six sharpening my axe." -- Abraham Lincoln
For the best use of Eng-Tips, please click here -> faq731-376

 

[condesinc]

Thanks all, the paper link is by far the best I have ever found on efficiency. I would be very grateful if any more data like this is found, especially at the higher horsepowers, and also treatises on PF. I wish they had published the 100 hp case...perhaps I can contact the author.

David:

For my purpose, PF curves are very important. The old data that I have are (VFD outputed)volts, amps, and frequency. Without motor PF, one cannot separate real power from apparent, and without eff, one cannot get the deliverd mechanical power from the recieved real elecrical power.

 

[dpc]

My point is that when viewed from the source side of the VFD, the motor pf is not much of factor. The input power factor to the VFD is not dependent on the motor power factor.

If you are trying to use volt and amp data measured at the VFD output, I would be extremely suspicious of the accuracy of the data. Most measurement equipment is not capable of providing much useful data from the output of a PWM drive. The VFD can do a much better job of computing motor kW load than can be done with external measurements on the drive output.

David Castor

http://www.cvoes.com

 

[condesinc]

Yes, it takes some sophisticated equipment for measuring this given the highly distorted waveform. However, vendors (more than one) have told me that what the VFD itself reports as output for amps and volts, at the cabinet display, is pretty accurate. Never thought to question that further, nor have I dug into how it actually determines these values.