Motor Stator failure 5

[koshyeng]

A 700 Hp motor failed during running with smoke and burnt smell very apparent. Upon inspection, found the stator windings burnt in the slots going to ground. I am attaching couple of pictures here to help indentify possible cause of failure. This motor was in service for 8 years and the previous motor had stator windings tutn to turn failure due to dirt accumulation. We put enclosure around the motor with AC to address the dirt issues. This time there was no dirt observed.

Voltage unbalance is 0.8% and current imbalance is less than 1%.

Motor was operating at 250 amps and spiked to 800 amps for few secs prior to failure.

Any help to identify the cause is highly appreciated.

 

[koshyeng]

I couldn't upload files at this time as it shows run time error. I will try again after few hrs.

 

[waross]

Is this a lightly loaded 460 Volt rated Motor?
What is the load?
Is this motor on a VFD?
How long are the conductors from the controller to the motor?
Have you monitored the motor temperature to see if the AC is adequate?
Are you fed from an overhead line? Have you had a recent electrical storm?
Welcome to 20 questions.


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

 

[electricpete]

The symptom of current increasing for a few seconds prior to trip seems unusual to me.
At first glance, it suggests to me the failure did not begin as a ground fault.
(Can your system even supply that much current to ground. And if it can, would it last that long without tripping. I doubt that both these questions can be answered "yes", and if not then the high current was not ground current. Then again I don't know much about your protection...).

Please describe the motor protection and which elements tripped and which did not trip. Also describe system ground configuration.
Was that current increase observed on panel meters or recorded?
Would certainly like to see more data about what currents were recorded if available.
(How did the three phases behave... is any indication of residual/ground current recorded?)

Was this motor rewound following the failure 8 years ago, or is it a different motor?
If rewound, then I'd be looking at the rewind records to see if anything unusual was recorded. One thing always of interest is the core loss test and associated thermography images after winding burnout.

As you get further into the inspection, you'll want to try to see by electrical measurement if more than one coil is affected. And if possible, where are the failec coils in relation to the line end connections. Another useful thing is to plan some aggressive tests on portions of unfailed winding (surge test and dc step voltage test) to see if any overall weakness is apparent either in the turn or groundwall insulation. If you want to invest a whole lot of time, then cold strip the failed coil with minimum heat necessary to pull it out strand by strand.... there may be some clues in there.

Also of course those photos...

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

 

[koshyeng]

Waross,

This is 600 hp 480V VFD motor with average load of 520 amps. The controller is in the MCC which is roughly 150 ft from the motor leads.
Enclosure is maintained at 78 F and there are alarms and safety shutdown interlocks if the enclosure temperature reached 95 F. No alarms nor the thermostat indicated high temperatures. AC unit is running normal.
Again, there was no electrical interruptions and the power is fed through a main transformer and switch gear.

Electricpete: I don't think that huge current would go to ground. Current increase is recorded on our DCS and this is a brand new motor that has replaced the failed motor 8 years ago.

 

[iceworm]

Is your 480V system solid-grounded, or HRG?

I'm guessing solid grounded

iceworm

Harmless flakes working together can unleash an avalanche of destruction

 

[electricpete]

I don't think that huge current would go to ground.

I agree. It looks like the DCS samples/stores ever few seconds and there are at least two stored samples of increasing current. No well designed system is going to allow a large ground fault current that persists for many seconds.

So, it seems the initiator was a turn-to-turn fault, which also seems supported by inspection of the photo's:
1 - the carbon trail on the rotor looks fairly uniform as if the fault sprayed from rotor onto stator for many revolutions (not a single rapid blast)
2 - the damage pattern indicating a long area of overheating along the slot with melted copper beebees evident.

By the way it is a random wound motor, unusual for a motor this large.
In the last random wound motor I saw with turn to turn fault, the evidence of overheating of selected strands could be seen in the endwindings...don't exactly see that here but maybe it's buried beneath what we can see.

Figuring out why we had a turn to turn short to begin with will be a challenge. As suggested, see how close the failed coil is to the line end as indicator of how likely that the cause was the power supply (repetitive spikes from vfd or other power system surges). Try surge test on some non-failed portions of winding to see if turn insulation weakness is evident. Carefully pull out the strands to see if there is more evidence. Look at overall workmanship including tightness of winding in the slot. Also examine any core damage in the vicinity to try to ascertain whether it came before or after the fault. Just some ideas for things to check. Although from my experience, more often than not there is no definitive smoking gun.


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

 

[zlatkodo]

Is a 6 pole winding?
If yes, there are 18 groups of 4 coils ( if the number of slots is 72).
Number of circuits is more then 1 for sure, usually 2 or 3 (or 6 in worst case).
Each coil contains only few turns and the voltage per turn is high even if there is no any VFD.
Note that is about random wound windings and there is a possibility to first and last turn in coil can touch each other.
Looking at attached picture it seems that failure occured in the first coil of group. Am I right?
You should check whether or not this group is the first group in phase winding.
I am not an expert on VFD, but from practice, such a failure is often related to VFD, somehow.
Winding Design & Repair

 

[panter]

Interesting but I can not see wedges at stator coils.
I had seen that they are not installed on several kW power motors but 600HP motor is fairly serious .
Good luck !

 

[electricpete]

Good point. It looks like it has the soft u-shaped cap, but not hard wedge or topstick (unless it is hiding under the cap and is shorter than the cap)

I googled and found this discussed here:

http://www.mandc.co.za/pdfs/Electric_Motor_Failure_Wedges.pdf

Picture 4 shows an otherwise standard winding that clearly does not have a stator slot wedge in it at
all! Some customers have recently discovered this practice in imported motors and have reacted by
reissuing their specifications outlawing this practice. Sadly, although the practice was stopped for the
customers that complained, the rest of the industry may still receive these inferior motors without
their knowledge.

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

 

[koshyeng]

Thanks guys. Will get the vendor to do additional tests on the windings and inspect while we pull out the strands at the motor shop.

One thought overheating of the coil overtime and finally the insulation gave up after 8 years of service? This is F type insulation.

The replacement motor installed after this failure is H type with RTD's and Clixons installed. I regret not conducting meg test on the windings

as part of annual PM.

We are actively looking in to ATpro for offline tests to develop predictive maintenance. Any recommendations?

 

[electricpete]

One thought overheating of the coil overtime and finally the insulation gave up after 8 years of service? This is F type insulation.

Do you have any records that might shed light on the running temperature of the winding? (recorded winding RTD temperatures while in service, recorded currents).

In general, I would not suspect thermal failure unless I saw some visual evidence of overheating in other points of the winding to point me in that direction. Given that turn insulation initiators is suspected, I think surge test on other non-failed portions of the winding will have some value. If remainder of the winding seems to have healthy turn insulation based on surge test, I would not suspect global overheating as a cause.


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

 

[jraef]

At 150ft from drive to motor, I would not be immediately suspicious of the VFD having created standing waves with voltage pulses that exceeded the motor insulation rating, but there are numerous other details that would need to be vetted. I did see that happen 2 years ago on a good quality "inverter duty" (Reliance) 250HP motor fed from a good quality (Mitsubishi) VFD with an output reactor and had only a 125ft cable run from drive to motor. Nobody could figure out why the motor insulation kept failing. I went to the site and observed that the cables, although STARTING in conduit at the VFD cabinet and going up the wall, exited the conduit and were laid in open cable tray, right on top of the INCOMING cables, for about 50ft. After the motor was rewound the 2nd time (which is when I saw it), a scope on the motor terminals showed 2200V spikes! I had them re-route those cables away from each other by temporarily laying the output cables on the ground, the voltages became very clean, spikes were below 800V. I was amazed at how much that short distance of running on top of the fixed frequency incommers affected the standing waves, but there it was.


"You measure the size of the accomplishment by the obstacles you had to overcome to reach your goals" -- Booker T. Washington

 

[jcbvet]

That failure is an insulation failure. Starting at where the winding enters the stator stack.
Large HP windings move a lot under inrush. there should have been some glass wrapping laced through the windings.to prevent the inrush torque on the windings. Maybe there wasn't thick enough slot paper.