Harmonic slip losses in induction motors 1

[Marke]

I have been looking at information on the effects of harmonics on the losses in an induction motor.
It is commonly accepted that negative sequence currents due to phase voltage imbalance creates a very high slip loss in the rotor of the induction motor. This leads to a derating curve that requires a derating to 75% for a voltage imbalance of 5%.
If I look at the effect of say significant 5th harmonic voltage on an induction motor, then if the 5th harmonic field was rotating in the same direction as the fundamental, then the torque produced by the 5th harmonic field is going to produce a torque at that speed. This will result in a slip loss of the 5th harmonic torque times a slip of 400% which, like the negative sequence torque, could cause significant rotor heating.
When I look at papers on harmonics and induction motors, the slip losses do not appear to be taken into account? I have found references to iron and copper losses, but would the slip losses be significant?
I am loooking at installations with a THDv of up to 16% and am concerned that there should be derating applied to the motors to offset the additional harmonic slip losses.
Best regards,
Mark.

Mark Empson
Advanced Motor Control Ltd

 

[electricpete]

I haven't particularly heard of "slip loss".

I would say you have copper and iron losses in rotor just like stator.

My two cents fwiw:

The slip of harmonic compared to rotor speed (which is roughly the fundamental field speed) is important as relates to determing the frequency which affects both copper and iron losses. i.e. for a given current and associated field, higher frequency causes more resitive losses (due to skin skin effect) and more core losses (by Steinmetz approximation), and possibly more stray losses as well.

In general, 1st is forward/positive seq, 2nd would be reverse/neg sequence, 3rd is 0 sequence, then it starts again, 4th would be forward sequence, 5th will be reverse sequence, 6th zero, 7th forward, etc. Of course we know the 1st, 5th, 7th, 11, 13th etc are the important ones. 5th is neqative seq 7th is positive.

The polarity (forward or backward) has two effects. First as you mentioned it affects the slip (5th harmonmic has slip of roughly (-5)-1=-6.... 7th has slip of roughly 7-1 =6) which affects the losses. Second there is a torque which can reinforce the main torque (7th harmonic) or oppose the main torque (5th harmonics).

At least that's the way I look at it based on the analysis provided in "Power Quality in Power Systems and Electrical Machines". That analysis is perhaps relatively simplistic, but intuitively appealing to me based on tie to first principles and perhaps leaving out some of the more complicated effects.

The Induction Machine Handbook by Boldea provides a far more detailed analysis, involving some empirical approach and generally higher complexity. I suspect it is more accurate, but I like the first one better.

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

 

[electricpete]

Second there is a torque which can reinforce the main torque (7th harmonic) or oppose the main torque (5th harmonics).

Just to clarify, a harmonic present on stator (for example 5th) typically creates two types of torques. The first type is a steady torque by interacting with the 5th harmonic induced into the rotor.....That is the one which opposes (5th harmonic) or reiinforces (7th harmonic). The second type of torque would be an oscillating torque produced by interaction of 5th harmonic with 1st harmonic.... due to oscilating nature it averages to zero.



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

 

[electricpete]

Also, I did not mention, the slip associated with a a given harmonic also affects the impedance to that harmonic and so determines the amount of harmonic current that will flow for given harmonic voltage, as predicted to a first approximation by the equivalent circuit. The power quality book mentioned above gives an equivalent circuit for harmonic voltages. Results from all frequencies are supposed to be added by superposition.

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

 

[Marke]

Hi Pete
Thank you for your inputs.
The slip losses do occur as iron and copper losses, not in addition to them, but where there are high slip values with toorque produced, there is an associated slip loss.

I am particularly interested in the effects of harmonic voltages on submersible motors that have a very long thin rotor.
These motors produce a much higher torque an high slip than to standard motors due to a higher rotor resistance PU.
This would suggest to me that the losses due to say a significant level of 5th harmonic would be higher than with a standard motor.

I have seen a number of these motors that apppear to have had significant rotor heating. These are from areas of high 5th and 7th harmonic voltages, THDv around 12 - 16%.
I have not been able to find figures for the increase in losses for this style of motor under high harmonic voltage conditions, but I expect that the losses would be significantly higher than a standard motoor under similar conditions.

I would like to determine a realistic derating figure for these motors in this high THDv region.

Best regards,
Mark.

Mark Empson
Advanced Motor Control Ltd

 

[KllrWolf]

I work on fluid filled motors such as you describe. The problem with a derating figure is it depends on frame size and power. We have been working with several customers that have been trying to just derate the motor to make up for issues with the power supply, and the only consistent answer we have found to work so far is to clean up the power. Running a motor below its design point hurts its efficiency. The cost of a good filter can usually be made back in just a few years versus the cost of operating a derated motor.

We had a customer killing a motor every 6 months or so, even with derating. Then his people tried a filter for the motor, and 3 years later, not even a hint of an issue. We have even told a few of the more stubborn ones that we would pay the cost of the filter (we do not deal in filters, just motors) and a new motor if it did not last the full warrantee period.

 

[Marke]

Hello KllrWolf

Yes I agree, get rid of the problem is the best solution, but it is going to take many years to achieve this.
I am looking for two things, 1) an indication of just how much the harmonics are increasing the losses in the motor so that I can use this to illustrate why the problem needs to be solved sooner rather than later, and 2) short term solutions if there are any.
De-rating the motor will help up to a limit, but an additional filter may be another option.
All the filters that I have seen, are designed for use on the input of a VFD rather than with a motor. Do you have a link to motor filters?
Bsest regards,
Mark.

Mark Empson
Advanced Motor Control Ltd

 

[KllrWolf]

Harmonics are just one of the issues with VFD's on motors. Unfortunatly there is not a set amount of losses caused. The harmonics depend on many variables, including the VFD construction. As to importance to correct, it is bad in that losses, and additional heating, occurs, which can shorten the motor life. It is the same importance as the voltage spikes and voltage reflections. Fortunalty, there is an easy way to take care of them - filters. These are both the short and long term solution. I do not have links to specific filters, but they should be easy to find talking to filter companies (I saw information on both MTE and . The most basic is a load reactor or harmonics filter. The next step up is a dv/dt filter. The ideal is a true sine wave filter.

One part of our guidelines for VFD operation is about the carrier frequency. We recommned no higher than 4kHz and prefer it to be as low as possible. This helps with the voltage spikes. Of course with a good sine wave filter, it is not an issue.

 

[Marke]

Hello KllrWolf

Thank you for your reply.
The problem that I have is not with motors on the output of VFDs, rather it is motor directly connected to the supply.
The number of VFDs connected to the supply is resulting in very high harmonic voltages on the supply, in some cases higher than 12% THDv. I have seen premature failure on submersible pump motors where it appears there has been excess heating in the rotor. I believe that this could be due to harmonic slip losses in the rotor.
Best regards,
Mark

Mark Empson
Advanced Motor Control Ltd

 

[KllrWolf]

With the excess heating in the rotor, is the motor needing to be rewound due to overheating?

Each VFD should have a filter between it and the supply to prevent it from affecting the supply. These harmonics can affect the total power factor of the location. Heck, the power company might even charge a penalty if the power factor or harmonics are too bad.

If I remember correctly, it depends on the VFD, but that the 5th, 11th, and 17th harmonics are the worst problems since they are reverse rotating, which increases the losses and slip.

Derating a submersible to handle the harmonics is usually not a good idea. Submersibles are designed to run at or near full load and once the load drops, efficiency really starts to drop. This is usually around the 75%-80% load point on the smaller motors. The larger motors have a flatter curve due to the fact they are not as affected by the internal drag. The other problem is making sure you still have enough flow past the motor for cooling.

I still believe that a filter before the motor is the best solution, and less costly in the long run. Talk with your filter people, but I believe a simple load reactor should be able to filter the harmonics from the supply so you get the clean power to the motor. There are other solutions, but not sure if in this situation that the higher cost would be worth it.