415V motor failure

[neps3]

Hi

We had a stacker boom conveyor motor rated 415V 150HP, 1485rpm failed. On inspection the failure was observed to be in the overhang area at the NDE. The failure looks like turn to turn and then to earth. Two phases are down to earth. The motor is controlled by soft starter for starting. The failure occurred after heavy rain, but the inspection concludes as if moisture had no impact. The rewinders are of the same opinion. One phase is measuring more than 100Meg.

Any technical/analytical comments please.

 

[electricpete]

At our plant we have 80% of our motors indoors, but it seems to me that 80% of our failures occur on our outdoor motors, and 80% of those in the rain (very rough guess based on limited info). Odds are rain has something to do with it.

Most likely moisture, less likely thunderstorm causing voltage surges.

Regarding moisture entry, was a TEFC or "open" (weather protected 2 etc)? I'm guessinG open.

I have a hard time imagining how someone can rule moisture out of the failure scenario that you describe, unless they know exactly what caused the initial turn-to-turn short. There may be various weak points in the insulation of each conductor in the end turn, some caused by nicks, mechanical stress, manufacturing imperfections, whatever. If no two weak spots are near each other than fine, no problem. When you add water to the equation (particularly in presence of any conducting contaminants... and some contaminats are alwasy expected on an open outdoor motor), then there can be conducting path which will allow the electric stress to reach farther out of each defect, as if they are now closer ser together.I can't imagine how you rule out water as a contributor unless you have identified some other clearcut root cause which initiated the turn-to-turn.

Regarding the possibility of lightning surges.... was it a lightning storm? Most important, was the failure in a line-end coil (nearest the terminal)? Lightning surges hit the first coil the hardest.

 

[neps3]

electricpete

Thank you for the good response.

Yes personally I will put moisture into the equation. The motor is a totally enclosed fan cooled motor. It is on a stacket reclaimer for shifting Coal in the Coal yard. It is open to the weather and rain pours on the motor. Although it is a TEFC motor I agree with you on the moisture because of breathing effect.

I have been advised by the rewinder that the failure occurred in the middle of a coil and not near the line end. Also the news is that the particular type of motor fails in the overhang. I will not disclose the manufacturer. It is an assumption that the insulation weak point might have gone between phases and then to earth. It is just an assumption. Will the soft starter have any effect on the dielectric. I know the VFDs but how about the soft starter. OR can we assume that the motor had its life. Its a 25 year old motor overhauled five years ago. I was not here that time anyway.

I am getting the motor to be VPI and to class "F" insulation.

But whatever it is, if the motor can breathe moisture then can we stop the failure. An IP55 motor, how good is it with rain.

Thanks and regards

 

[electricpete]

IP55 is a pretty good spec for outdoor motors.
VPI is a must-have for outdoor motors.... that is a very good choice in my opinion.
Class F or higher is now standard for all rewind shops to my knowledge.

I understand your scenario that the failure appeared to initiate in the "overhang" (end turns) which means it appeared to initiate as a turn-to-turn failure. In my mind that doesn't rule out moisture as a contributor at all, and I have given my logic above.

It seems like Soft starter could create similar voltage spikes to a vfd, since the waveform is electronically shaped (anyone else want to comment on that?).

With the failure being deep in the winding I would have thought the power spikes from soft start have a good opportunity to decay by the time they get there, so it seems to make the scenario of power spikes less likely. Maybe for random wound motors such as yours that decay is not as rapid as for form wound motors.

 

[neps3]

electripete

Thanks

Yes you are correct as you said before, the spikes would not have caused the failure as the failure was in the middle and not line end. The motor is been rewound now to class F, VPI and also a core flux test on the stator core to determine interlaminar failure. Excitation will be low for the test.

Regards

 

[GGOSS]

Hello Kantor and Electricpete,

Both of you appear to be very well versed in diagnising motor failure modes, however in the interest of all members and visitors viewing this thread I would like to suggest the following link for additional information.

Go to:

http://www.easa.com

Click on 'Industry Info'

Click on 'Failures in three phase stator windings'

What you'll find when you get there is a bunch of photographic images of stator failures together with good notations re the fault cause.

Hope everyone finds it interesting.

Regards,
GGOSS

 

[jbartos]

Suggestion: In some places, the power supply quality deteriorates during the rains. It may be a good idea to analyze the power supply during rainy time periods for voltage sags and voltage dips (due to shorts), larger harmonic content, etc.

 

[neps3]

GGoss

Thanks that was a worthwhile excellent site.

 

[neps3]

electripete, GGoss, Jbartos

Thanks the failure is turn to turn. I agree with your equations. I would like to add some data. The TEFC motor has a double shaft. On one side there is a brake drum and the other side is the drive coupling. The brake drum was covered to comply with the safety standards. There was only about two inches clearance between this and thats where the air intake mesh is also located. I observed Coal sized about one inch lying between the space and in the cooling fins. Therefore I am adding cooling constrain into the equation which caused rapid degradation. The overhang failure is on the drum side. The operators are suppose to clean the Coal dusts anyway. But??

I intend to redesign the whole drive system if time is available.??

Regards

 

[Marke]

Hello kantor.
I would not expect the soft starter, (provided that it is working correctly) to cause any transients that would lead to insulation failure.
Modern VSDs use IGBT switches to switch the voltage applied to the motor with a very high dv/dt and a high switching frequency. The current flow is constantly being interrupted by the transistors. In a soft starter, the voltage applied to the motor is phase controlled, only during start and soft stop. The voltage is switched on part way into each half cycle by the SCR which have a relatively slow switching speed compared to the IGBT resulting in a low dv/dt. The SCRs self comutate OFF when the current drops to zero, so the current is not swithed OFF by the SCRs. Switching only occurs once per phase per half cycle rather than at the 10kHz rate with the IGBTs.
I would suggest that, all winding wire has pin holes in the insulation. Good winding wire has less pin holes per unit length than poor winding wire. The presence of contamination can provide a leakage path when damp and this can lead to a full breakdown. If it is raining, then I would expec the humidity inside the motor to match the humidity outside because the motor breaths, and so the windings will get damp. This is particularly a problem the first time a motor is run after a shut down period.

Best regards, Mark Empson

http://www.lmphotonics.com

 

[neps3]

marke

Thank you for the info.

Regards

 

[CB2]

Since you have the motor open for the rewind, I would suggest you install a space heater. Keeps the motor warm during "off" time, helps prevent condensation when a motor cools to ambient. Your rewind shop (EASA I hope) can handle as standard practice.

 

[electricpete]

I'm not all that familiar with the variety of soft starters and vfd's out there. I am a little surprised that we can rule out a soft starter producing voltage spikes simply by the fact that it is a soft starter (for the sake of argument, ignoring the location of fault deep within winding which also seems to rule out voltage spikes)

Won't it depend upon the type of electronic switch used, along with surge supression, and system (cable) filtering/resonance characteristics?

Is it being argued that fewer potential spikes per unit time (due to lower switching rate) make the soft starter less dangerous? OR is it being argued that the type of electronic switch used on soft-starters is a slower-switching device?

 

[gpedens]

Sounds like you might have "mush" wound or random wound motors with not enough bracing in the end turns. Also the winding may ony partially fill the slots. These are 2 of the cons of mush wound motors. Even with soft start there is some current inrush and the winding can move, especially in the end turns. The VIP you suggest will help, but its not a 100% cure. The extra cost for more slot filler and extra taping and epoxying of the end turns may out weigh the benefits. More handling of the winding to do this may cause more nicks. Form wound motors may be a better option if they are available. I would think moisture condensation would be a factor if the TEFC motors are only ran occassionaly. If they are ran about 3 hours a day this would tend to push out the condensation. In some cases if there is enough water and the motor is running while it is raining, the shaft may tend to "pump" the water inside the motor. I think (memory not so good) this has something to do with the bearing design. But even with water inside the motor there would still need to be nicks, cracks or weak places in the winding insulation. We even had some 4160 volt (that is 4 thousand volts plus) motors full of water and running one time. I wouldnt suggest trying this. Some OEMs used to test motor windings at twice rated plus one thousand volts under water. Make no mistake, moisture will wreck a motor. Look for flakey rust spots where water has stood or condensed. Thank You,Radar. Duke Power

 

[jbartos]

Suggestion: Visit

http://powerqualityanddrives.com/EMI-Figure1.html

for: paths of flow of EMI/RFI

http://powerqualityanddrives.com/emi.html

for: EMI/RFI mitigation, including dv/dt filtering

http://www.mgitech.com/access_old_020322.htm

for: LRC dv/dt filter sample
etc. for more info

 

[jraef]

Kantor, electricpete et al,

Marke is right, and to reiterate, it is the slower switching speed of the SCRs PLUS the lack of forced commutation that precludes the creation of "spikes". I am repeating what Marke said, but it warrants clarity.

The phenomenion referred to as "spikes" with regards to VFDs is a result of several factors. 1) the switching speed of the transistors, i.e. 10kHZ+, 2) the fact that they are being switched on AND off, and 3) a capacitive coupling effect of the wires. Those 3 factors create "standing waves" in the conductors to the motor which can be cumulative and create short bursts of low energy high voltage pulses at the motor. The failures are typically in the first turns, causing a microscopic burns through the insulation that gets worse and worse over time. Since the latter is dependent upon there being enough wire to generate the effect, the phenom is somewhat dependent upon the distance from the drive to the motor. Without going into more detail about this, I only bring it up to compare against SCRs.

SCRS fire ONCE per cycle, not 10,000 times, and the voltage rise time through them is many times slower than a transistor, so any capacitive coupling effect on the wires is practically non-existant. Therefore, the SCR firing cannot "generate" spikes of any magnitude greater than the available line potential (plus a little noisey fuzz from gate pulses). The SCRs are also on-line such a short amount of time that they can hardly be responsible for anything detrimental to the motor. For instance, if a motor operates for 8 hours per daycumulative, and is started 4 times per hour, the SCRs were in the circuit 3.3% of the time. Hardly capable of doing much Quando Omni Flunkus Moritati

 

[jraef]

Back to the original problem. I cast my vote for moisture as well. gpedens post says it well, but I want to add that contrary to popular belief, all motor winding insulation material is hygroscopic, meaning that it wll absorb moisture. Better quality material is less so, but none are totally impervious. That is why gpedens mentioned the issue of duty cycle. If these motors are off for long periods of time in a wet environment, then it is quite common for the isulation to break down. I have seen lumber drying kiln fan motors in British Columbia fail in a matter of months because they open the doors of the kilns in winter, and the hot steam envelopes the motors. Since it takes them 2-3 hours to unload and reload the kiln, the motors get cold quickly, especially in winter. When they turn them back on, that moisture has infiltrated the windings and begins the breakdown process. Strip heaters or motor winding heaters can prevent this very effectively. Quando Omni Flunkus Moritati

 

[jbartos]

Suggestion: Also, any moisture help various chemical reactions to materialize and become more detrimental by seeping the chemical liquid into various cavities, thus causing them bigger and potentially fatal to the motor operation; especially, when the insulation impedance/resistance is affected.