800 HP DC Motor Failure - Severe Brush Wear

[jktwn]

We lost a GE 800hp motor on an extrusion line this week when the commutator flashed over. The motor brushes were completely worn away on the A- racks and the A+ brushes were almost gone.
We measure brush lengths every 90 days so the failure caught us by suprise.The National Electrical Carbon engineer (very good resource) suggested I look for changes to the environment in the vicinity of this motor, especially ventilation work involving silicon sealants. Checking this, there was a major ductwork installation at that end of the building in the last couple of months and silicon was used extensively to seal the joints.
I am not yet convinced that this is the root cause of the failure.
Has anyone had experience with silicon and excessive brush wear or know of articles on the subject?

Thanks,

 

[busbar]

Anecdotally, silicones and brush/commutator interaction seem to have gotten some reasonable bandwidth lately.

 

[electricuwe]

See thread237-9155

 

[jktwn]

Thanks for the pointer to the thread. Good information. I'm still not clear on exactly what the mechanism is for accelerated wear. It sounds like there are some theories but has a study actually been done to understand the chemistry and physics of the Cu/Si interaction?

 

[peebee]

Can I ask a stupid question -- why would you use DC for such a large motor rather than AC?

 

[godsaxeman]

Strange as it may sound, softer brushes may last longer due to graphite content.

 

[37pw56gy]

Difficulty has been experienced with silver-impregnated carbon contacts exposed to silicon compounds outgassing from RTV sealent inside a relay's dust tight enclosure. The failure mechanism works like this: Minute arcing occurs each time the contact opens. Silicon present in this atmosphere converts to silicon dioxide (quartz) on the face of the AgC contact at the site of the latest arc. Given sufficient time and relay operations, contacts that normally have a resistance of 0.05 to 0.2 ohms develop a resistance in excess of 10 ohms. The solution is to eliminate use of silicon-bearing sealants.

The connection to the brush wear problem, if any, lies in the chemical changes that occur when airborne silicon atoms are exposed to an electric arc. Perhaps examination of the failed brushes would reveal the presence of silicon dioxide deposits, and thus explaining the unusual wear mechanism?

 

[jktwn]

I can see how that happens.

I have sent the failed brushes to National Electrical Carbon for failure analysis. I will ask for a check of SiO2 on the brush faces. This would be direct evidence of a silicon vapor induced failure. I'm afraid the extreme heating that occurred during the flashover may have vapoized anything on the brush faces.

By the way, neither myself nor the project engineer that had the ventilation system installed was aware of the silicon/dc motor incompatability issue. It seems to be an obscure phenomenon.

Also, the motor repair shop has found a short to ground of a compensator winding that apparently is in series with the armature circuit. Not sure if this was a cause or effect of the flashover. Motor was built in 1979.

 

[UKpete]

I'm not necessarily pouring cold water on the silicone theory, but some years ago there was concern over the manufacturing process on DC traction motors, specifically the commutator building where it was discovered a silicone mold release was being used on them. We stopped the practise though it was never shown to cause a problem.

Another thing I have noticed is that often pigtail insulating sleeves have silicone impregnation, and I've seen an RTV used on brusharms, again with no apparent ill effect. I'm no chemist but isn't it a solvent that is given off initially by silicone, then it's inert.

If the problem is that serious, maybe there have been tests carried out by carbon brush manufacturers because it should be pretty easy to prove, rather than being anecdotal.

Finally, I see no mention of commutator profile in this or the other thread referred to. Poor profile kills brushes -it's a bit of a runaway situation once sparking starts at a particular position on the commutator. Incidentally measuring the profile both on and off the brush track is recommended, it distinguishes electrical problems from mechanical ones.

 

[UKpete]

One other point wxashcra, I doubt that the compensating winding caused the problem. If it had previously had an earth fault on it, it would have caused the supply to trip well before the brushes had time to wear down; it was the brushes wearing down that finally triggered the flashover. The compensating windings are in the armature circuit and there would have been a LARGE current surge at the moment of failure that may well have caused it to go down to earth.

 

[jbartos]

Suggestion: Any hard substance dust, which may have been present during ductwork installation, could affect the wear of brushes, if the dust can get under brushes. If you like to convince yourself and exaggerate the situation, try to poor some fine sand on the commutator and watch the clock how soon the brushes will be worn out due to the abrasion.

 

[peebee]

Really -- can anyone please give me some idea why you'd use DC for such a large motor?

 

[jktwn]

A new extrusion line being installed today would use mainly ac motors and drives. When this line was installed 20 years ago the technology was not available to exploit the lower cost and more reliable ac induction motor.

It is not cost effective to convert dc motors and drives to ac in most cases. The dc motor does have the advantage of higher efficency in converting electrical energy to mechanical however.

Hope this helps.

 

[peebee]

Ahhh -- thanks.

 

[jbartos]

Suggestion: Major manufacturers of larger motor drives claim the the Mean-Time-Between-Failure (MTBF) of a larger motor drive is over 50000hours. This is a noticeable amount of time for so much of power electronics in them.

 

[purplepete]

peebee,
the main reason DC drives are still used is they can generate full torque right down to standstill. This is useful in drives on, for example, paper machines. Only recently has the technology come available to do this with the AC induction motor, even then I'm not sure it can match the performance of the DC drive
Cheers

 

[don01]

HI
2 small inputs to the thread.
1) re silicone atack. I've heard of this but the discussions usually centre around sealing on the motor and the presence of a "rust" type of attack on some of the metal work around terminations.
are you sure there isn't shorted turns in the interpole?
2) Hi power dc drives. a big advantage in the bad old days was using a much smaller field current for control or excitation current in ward-lenard type systems (how do you spell that) this kept control gear size & cost down heaps. Eg steel mill & rolling plant.

 

[jktwn]

The report from the motor shop confirmed a short to ground in the compensating winding which is in series with the armature windings. They also reported degraded insulation throughout the motor which leads me to conclude that the motor failed from old age. It was placed in service in 1979 and runs 24/7. We are now scheduling insulation testing on some of the older motors that have seen similar service.

The accelerated brush wear was most likely not the cause of the flashover. The severe wear was caused by a combination of clogged filters in the cooling air system and silicon vapors from a new ductwork installation.

Following up on the DC versus AC motor question - We still use dc motors in new designs if the motor is say 800 HP or larger.

 

[jbartos]

Suggestion to the previous posting. Please, would you elaborate on the "silicon vapors from a new ductwork installation."
It appears that the silicon vapors would need about 1100 deg C.
See:

http://www.seas.ucla.edu/~goorsky/pres/pfabsmrsfall99.htm

for reference on the silicon vapor temperature.

 

[sksingh]

I am also facing the similar problem in my plant. A DC motor of 100kw is running with certain grade of carbon brushes. Its carbon brushes get rubbed entirely within 5 to 6 days. So after every 5 to 6 days we need to change carbon brushes with new one. Other DC motors of same rating is running healthy with similar carbon brushes.
we are not able to diagnose the problem. Now I will look into the matter from silicon theory angle.