The effect of THDv on the THDi of an induction motor. 1

[Marke]

A recent onsite test carried out to establish the THDi drawn by 6 pumps controlled by VFDs and soft starters yielded unexpected results.
We have a requirement that the THDi at the point of supply with all pumps running is less than 8%.
There are four pumps controlled by soft starters and two pumps controlled by VFDs. One VFD is fed via a Mirus Lineator filter to reduce the THDi.
On paper, it was expected that the THDi with all pumps running would be less than 8%, but this was not the case.
The background voltage distortion, THDv was around 4.5% and the THDi with one pump running on a soft starter was around 9%.
It would appear that the motor, fed by a distorted voltage waveform, was drawing current with twice the distortion.
Does anyone have any information on what the mechanism for this is?

Best regards,
Mark.

Mark Empson
Advanced Motor Control Ltd

 

[electricpete]

It is pretty normal for higher frequency components to have lower impedance than the "fundamental impedance" at full load.

Look at the harmonic equivalent circuit. Neglecting leakage reactance, due to the very high slip at harmonic frequencies, the impedance for the harmonics can approach locked rotor impedance of approx 0.2pu, in which (at full load), your harmonics would increase by a factor of approx 5 above the voltage harmonics. However the harmonics have higher leakage reactance, so the harmonic impedance is typically higher and increase somewhat lower.

=====================================
(2B)+(2B)' ?

 

[electricpete]

Clarification in bold:

...so the harmonic impedance is typically higher and increase somewhat lower. ..

should have been

...so the harmonic impedance is typically higher and current increase somewhat lower. ..

=====================================
(2B)+(2B)' ?

 

[electricpete]

Section 4.4 here describes induction motor equivalent circuits for harmonics:

http://books.google.com/books?id=4-...lent circuits for variable frequency"&f=false

=====================================
(2B)+(2B)' ?

 

[sibeen]

Mark, what were the dominant voltage and current harmonics?

 

[electricpete]

Good question. I think THDi>THDv is more expected for low order harmonics 5, 7, 11, 3. THDv>THDi is more expected for very high order harmonics such as from PWM where the high frequency effects on impedance (leakage reactance, rotor bar resistance at high freqs) really kick in.

There was a similar discussion, including some posted examples here:

http://maintenanceforums.com/eve/forums/a/tpc/f/7161085912/m/554102767

=====================================
(2B)+(2B)' ?

 

[electricpete]

Also the load of the machine is a minor variable to be considered since the THDi compares the harmonics against the fundamental. Lower load machine should generally exhibit higher THDi than higher load machine with same THDv.

=====================================
(2B)+(2B)' ?

 

[MAGTiger]

More info, motor hp, and running load. did the other drive have a line reactor in front. any bus magnetics on either drive. Source impedance value? Power frequency (50hz?). Is lineator designed for that frequency?

Neil

 

[Marke]

Hello MAGTiger

Motor size 15KW, all the same.
Only one pump running at rated load when the THDi measurement was made, so other drive/filter info irrelevant at this point.
I was expecting the THDi to be close to the THDv, not double.
The primary harmonics present were 5th and 7th.

Pete, your comment about the slip impedance of the harmonics is very valid. I had not thought about the obvious!! I guess that would explain it. This throws quite a slant on things as that suggests considerable slip losses and extra heating in the motor!!

Best regards,
Mark.

Mark Empson
Advanced Motor Control Ltd

 

[MAGTiger]

So, the one pump that is running, is that the one with the lineator harmonic filter. And you are wondering why it isn;t filtering or is the unfiltered drive the one you are measuring harmonics from? Understand that the THDv and THDi will not necessarily be matching numbers. In fact on an unfiltered VFD, the thdi can easily be 60% or more if there is little or no bus magnetics and a stiff source. When filtered the thdi should be in the 7 to 9% range depnding on load and source impedance. Where are you measuring thdv? It should be below 5% and the filter input terminals and well below 5% at the transformer terms.

Neil

 

[Marke]

Hello MAGTiger

The pump that was running was running on a soft starter. - no filter. The filter was on a VFD.
The requirement is to get the THDi contribution of the whole site to less than 8%.
The expectation was that by using four pumps on soft starters, - no contribution, one pump on a VFD with a filter, - approx 4.5% contribution, and one pump with no filter, 35% contribution, that the total harmonic current contribution would be less than 8%.
The problem was that the contribution from a single pump running effectively across the line was 9%, so there is no way that the installation willl comply due to the effect of the externally generated THDv.

Best regards,
Mark.


Mark Empson
Advanced Motor Control Ltd

 

[MAGTiger]

I'm a little surprised that the thdi for the line running motor is that high. Makes me think you may have a preexisting thdv problem. Do you know what the voltage distortion is without the motor running? Also, have you measure the voltage to insure it isn't so high that the motor is saturationg? That could cause the current distortion to get up some.

Neil

 

[Marke]

Hello MAGTiger

Makes me think you may have a preexisting thdv problem.

Yes, that is the cause, refer back to the priginal post.

The background voltage distortion, THDv was around 4.5%

The question was why the THDi was 9% when the background distortion THDv was 4.5%. This has been explained by Pete with the low slip impedance at the harmonic frequencies.

Best regards,
Mark.

Mark Empson
Advanced Motor Control Ltd

 

[MAGTiger]

Sorry, I see now, but when you ask, what is the mechanism for this, what do you mean? Are you referring to the point that thdi is greater than thdv? If so then you need to understand that the motor responds like a transformer to voltage. The magnetic core's response is nonlinear, and much more so the higher the excitation goes. 3rds can become very significant if the motor is driven at too high flux density but other harmoncis will be present. too. Even if the original voltage has no harmonics, the magnetic core can respond with distorted current. Starting with distorted voltage only makes the problem worse

Neil