Requesting Advice for DC Motor Flashover / Failure (GE 1500HP)

[RyanER]

Hi Everyone,

This is my first post here in regards to some motor/drive issues that we experienced yesterday. The motor is a GE #5CD695WA802C800, 1500 HP, 600/942 RPM, 630V ARMATURE, 1865 AMPS, SHUNT WOUND.

-The first failure we experienced was in the GE drive cabinet, where the drive faulted (assuming self-protection) and the 2400A contactor, which caused a nice loud bang and an arc flash which blew the plastic covers off surrounding the contactor. [This happened as the system/motor was slowing down to a low speed.] We were not able to capture the fault codes from the drive at the time because of someone prematurely disconnecting the drive logic power. Initially, we speculated that maybe the contactor "chattered" causing the arc flash, or simply, the field on the contactor coil has weakened and caused it to behave erratically. I proposed that there may be a hardware issue with the drive itself causing the digital output controlling the contactor to fail, the drive opened the contactor as a means of self-protection, or a few other possibilities. Regardless, they wanted to replace the contactor - so they did.

-We were able to retrieve some logs off our computer monitoring system later, indicating that the armature current was abnormally high when the failure occurred. This leads me to believe again, the drive is protecting itself and opening the contactor due to the excessive armature current. We removed the side panels on the motor and the commutator looked ok and the brushes seem to be well within tolerance. There are approximated 297 running hours on the motor. Myself being tired of all the speculating, recommended that we have GE come in for an emergency service, inspect the drive, the motor, and even Megger it, however, this suggestion was ignored, and later that evening a crew decided to continue tinkering with the system.

-After they ran the system last night a few times, of course, there was another failure, but not at the contactor. The drive fault code was pretty generic (will follow-up - I don't have it in front of my right now), but whatever happened did some damage to the brush bars in the motor indicating there was a decent arc/flashover last night. See attached picture. [When this failure occurred, it was during onset - increased RPM - acceleration.]

-At this point this morning, I am very upset with how mgmt/engineering has handled this situation, so I blew the whistle, and disabled the device so there is NO chance anyone can start the device, let alone power up the drive.

Now, my questions for this forum are the following:

> In the initial incident, while the motor was slowing down, why would the armature current be so high?
> Is it at all normal for the contactor to violently arc-flash and almost self destruct?
> During the second failure, what would cause such a violent flashover that melted the end of the brush bar?

I am adamant about getting GE in here to evaluate the system, but I personally want to grasp a better understanding of how this systematically failed. Additionally, I am using this incident as a learning tool for myself and to develop some better safety procedures and incident response guidelines.

I sincerely appreciate any information, suggestions, and tips from all in this forum. Please let me know if you have any questions or would like pictures. I have many more pictures I will post somewhere and add the link, since it seems I can only upload one file at time here. Correct me if I am wrong - this is my first time posting here

Thank you!!!

 

[RyanER]

Another picture attached of the commutator and brush bar with flashover markings.
Again, this marking was not present during the first failure yesterday, but after the second failure.

 

[nic00b]

Good afternoon Ryan,

Do you happen to know the protection arrangement for the drive i.e. any bender/overcurrent devices that control the contactor (possibly DCCB)? What is the motor driving and what is the overload capacity of said motor?

On the second fault, was the motor changing direction when the fault occurred? Have any checks been carried out on motor to eliminate any potential winding faults?

 

[RyanER]

Hi Nic,

Thanks for your reply. The motor (custom) is rated for "300% torque (316% Current) Peaks @ 3 second durations, frequently repeated." The motor is connected to a right angle Flenders gearbox, which spins a large, long, centrifuge arm. We only spin in ONE direction, never reverse.

As far as I am aware, the GE Drive is controlling that contactor.

There have not been any checks to determine winding faults. GE service is weeks away apparantly, so I am currently looking for a 3rd party to evaluate the motor and megger it to see if there is any issues with the windings.

 

[RyanER]

More pictures of the motor commutator:

 

[waross]

Flashover or "Ringing the 'Comm" is generally associated with excess armature current/voltage.
We were commissioning 20 DC motors. A mix of 1050 HP and 1300 HP. The GE technician was not quite ready for prime time. We had a lot of control issues. Almost every control failure involved over-excitation and resulted in a "Ring of Fire" around the generator armatures.
If the fault is in the motor, it may be a failing armature. I have seen smaller armatures checkout good when they were cold or when they were heated in an oven but fail when they heated up during normal operation.
If the fault is in the controller, there may be a problem in the ramp circuit. That is the control may be ramping up and/or down more rapidly than normal.

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

 

[nic00b]

Hi Ryan,

Do you have any field protection on the drive or have any data acquisition/monitoring for possible analysis?

Are you able to carry out some checks on the motor to eliminate any faults?

 

[djs]

Being a relatively new system, is it still under warranty? If less than a year old then it probably is. If so get GE out for warranty repair. Could be much quicker.

 

[RyanER]

So I have gathered logs from our Fault Log Manager as well as some information from the drive itself.

Here is the 1st run where the drive contactor opened during de-accelerationThis picture is zoomed/cropped to show detail)

Same run, just zoomed out for a glance at the complete profile. You can see in the beginning it was operating as normal, high onsets, high current, all normal. Then, during the second half of the profile (that is zoomed in above) it is a much slower commanded onset, therefore we do not require as much torque/current from the motor. Regardless, the commanded speed matched the actual speed until about 238 seconds into it (X-axis), current drops, E-Stop, etc ...

These were the faults that were pulled from the drive:

 

[RyanER]

Later that day, after changing the contactor, 3 profiles were successfully run (in a row).

First:

Second:

Third:

The next run caused the large flashover inside the motor (see next post, I have to get the data together)

 

[RyanER]

Here is the last run performed which is when the flashover inside the motor was experienced. This first picture is zoomed in to show the ESTOP detail. See next image for complete profile.

Complete Profile - Zoomed out.

Now what I find interesting in this final chart is the fact that AFTER our computer system initiated an ESTOP and set the Drive Enable Status to LOW, the drive did not close the contactor until after the crazy armature current spikes. From the 1st fault, the drive caught itself before our computer system did, and in this fault, our computer system caught the fault, initiated ESTOP and drive kill commands, but the drive did not self protect itself by opening the contactor.

Is it possible the contactor got stuck from the high current?

 

[RyanER]

In regards to the chart above showing the time delay between the computer ESTOP/DRV_EN and the actual drive contactor, I found out that somebody modified the contactor by removing one of the 2 springs installed.

Is it possible the delay was a result of the contactor NOT being able to open because a spring is missing!!!??????

This is starting to make more sense to me as now, I will post a log of the first run (before anyone touched the contactor) and it is very clear that the Computer ESTOP/DRV_EN & the actual Drive Contactor (a monitored DI) all went LOW simultaneously.

Middle Zoom so you have an idea of whats going on:

Full Zoom showing all going low:

 

[nic00b]

Good Morning Ryan,

Are you able to attach those images as i am currently having trouble reading the signal names. Do you have contactor feedback into the drive enable circuit?

 

[panter]

Ryan,
I think you have a problem with the controller and its current limitations and barriers.
The switch has to be extremely fast and every defect like missing spring can be a problem.
Make sure you have a resistor in rotor dynamic braking circuit , which should pick up voltage changes at a rapid disconnection circumstances .
Good luck

 

[waross]

Have you checked all the drive settings? Someone may have tampered with the settings.

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

 

[RyanER]

Good Morning All,

Let me get myself together and I will upload all of my charts to where they can be viewed more clearly.

To answer a few questions:

-Yes, the contactor feedback is a Digital Input (Interlock) on our computer control software. In the 3rd to last chart posted, it shows that DI going LOW (after) the computer system issued an ESTOP and cut off the drive enable signal. There should not have been ANY delay here and should have went low at exactly the same time as the last two images posted

-The drive settings have not changed and were not tampered with. This system has been in place and in-use (un-tampered) for the past 5 years with a total of almost 300hrs on the Drive/Motor. Validated by a mechanical timer that only records the time when the motor is actually spinning.

-Someone removed one of the two springs on the contactor thinking that it would help in case the contactor became stuck.

-Yes, this system has a regen circuit in place.

Thanks!

 

[RyanER]

I have uploaded the picture here, seperated into two folders representing both failures.

https://ryaner15.imgur.com/

 

[dArsonval]

It has been many months since signing on to the forum here.
Life happen-stances are my poor excuse.

Because this thread is now over a month old in contributing to a reply,
it’s possible the D.C. motor and its application have been diagnosed
and a solution was found. Never-the-less, there’s always room for further
armchair root cause analysis, especially since nothing has been posted explaining
what was determined to be the installation problem.

The photos posted of the motor commutator are excellent examples of what many
here on the forum usually appreciate. Clear, up close, great lighting, and more so.
Thank yOU for the great pictures!

If there was no commentary posted with the commutator photos, and one was relying
completely upon the pictures only as a means of “guessing” what the motor problem
may have [had] or has, I would submit there’s likely nothing electrically wrong with the motor.

With the internal motor environment being very clean and the commutator wear pattern
(for the most part) showing no obvious sign of "in-operability", I’ll toss out the following observations
based solely on the photographs.

The commutator bars show a slight evidence of edge burning. A factor leading to flashover.
The dark color streaks near the end portions of the commutator indicate possible material drag.
Deposits of copper material from the “edge burning” would likely be the source.
Again, a situation resulting in flashover.
Light brush pressure, or the wrong brush grade could be a cause, yet with such a new piece of
equipment, I’d lean more toward some kind of mechanical vibration as the root culprit.
(A kind of shaking, and rattling of the brushes in their boxes.)

A maximum speed of 942 RPMs (if the motor even runs that fast at any given time) seems too
slow to create very much vibration.
The graph screen shot, looks like it was running 150ish RPM? (I can not tell for sure.)

It’s possible the back bearing pocket of the motor is out of tolerance. Yet that’s a stretch given
the age and physical appearance of the motor.
No info relating to how the motor is mechanically coupled to the gearbox, the history of the gearbox,
or how long the installation has been in service other than (less than 300 hours) is known.

If this is/was a completely new machine arrangement, what about coupling alignment?
Is there an aspect to the gearbox that was investigated?
“Asking” the motor to drive a mechanical load that has issues would certainly cause high current.
Hence, the weakest electrical link failing first.

One can see a scenario where the motor is removed from service, shown to be perfectly all okay;
the brush holders cleaned up, commutator skimmed, undercut and re-chamfered, new brushes installed,
contactor and drive all remedied, etc.
And then, upon re-energizing the system, the motor runs for a little while and the same failure re-occurs.

Very curious about any new updates on this failure.

John