Arcing in MCC

[gordonl]

I've recently had an MCC tub arcing at the bolts which fasten the tub to the frame. We found that the tub grounding system was quite poor so we added grounding jumpers to each tub. We couldn't find anywhere on the tub, front or back, where there had been a phase to ground fault, and the upstream ground fault protection never tripped.

My question is: Could induced voltage in the tub have been enough to cause the arcing. The arcing was quite severe at one of the two bolts, causing welding to the panel and black burn marks, along with black burn marks at approximately 10 other bolts between sections of the tub, and the tub and the frame and rails. The tub contains a 600A switch and fuses which feed the MCC. The max load on the MCC is two 200HP motors @ 600V, simultaneous start. (Simultaneous start is now prevented) The motors are on hydraulic pumps, and the starts are relatively tame.

Thank You,
Gord

 

[DanDel]

gordonl, did you say the the MCC bucket where the burn marks are was the main feed for the MCC?
Typically, these pull-out style buckets have a ground strap which connects first upon pushing in and disconnects last when it is pulled out. Who is the manufacturer?
I do not believe that induction caused your problem. The marks you describe are usually seen after a ground fault when there is not a solid ground connection between components, as you have surmised.

 

[GusD]

Hi GordonL

DanDel may be right if the burnt marks are as extensive as you mention.
The following thread 238-50453 had some interesting ideas on a possible similar problem.If induction is the problem
give the thread a read.


GusD

 

[gordonl]

Thanks for the comments, I'll follow the thread. It's a CH MCC and has a ground strap that shorts to a metal plate attached to the rail, which in turn is attached by one or two screws to the frame. All surfaces were pre-tapped and painted with no ground washers or paint removal.

A little added description of the burned bolt & system:
- There was arcing and splashing from the bolt over an area of about 1/4" by 1/4" on the tub, with black burn marks radiating out for approx 1".
- The other bolts didn't show arcing, only black burn marks for 1/4 - 1/2".
- The system is solidly grounded, about 500' away with two 3C teck cables with bonding conductor. (Connected at both ends, I checked)
- The MCC also has a connection to the building ground.

 

[DanDel]

gordonl, did you say the the MCC bucket where the burn marks are was the main feed for the MCC?

 

[Shortstub]

You experienced the classical "arcing-groud fault" phenomenon!

Presumably the system neutral is solidly grounded.
can you supply additional info, i.e., voltage? Did fault occur during start? Ground-fault protection parameters?

The phenomena is not unusual, and often occurs in switchgear systems having, as a sole ground connection, the drawer wheel/rail arrangement!

Several faults I investigated the fault occured where there was insufficient clearance between the MCC bolts, and the "tub" backup.

 

[jbartos]

Question: Is upstream a ground fault protection device?
Also, an arc protection could be considered.

 

[gordonl]

The upstream ground fault device is mechanical 51 relay, recently cleaned, on a ZSCT. It's a small MCC, I'm not sure arc protection is relevant.

 

[jbartos]

Suggestion: Single pole arc protection circuit breakers are available to 20A.

 

[Shortstub]

Sorry, I should have said the "tub's" incoming breaker terminal bolts (not its mounting bolts) and the backplane of the "tub". The arcing was not obvious without removing the breaker.

Regarding the 51. Where is it installed? What are its pickup and delay settings? Did it operate?

In arcing-ground faults, the resultant rms current is too low to be "seen" by the 51.

 

[gordonl]

The ground relay is a CO-8 set at 200A primary pickup, 2TD. It never picked up. It is installed at the feeder in the sub approx 500' away. Bonding connectors running in 3C Teck connected at both ends. When the incident occured we were starting two 200HP motors. The motors started fine. If it was an arcing ground fault it self extinguished. I suspect the motors may have still been spinning during the start.

I don't have breakers, I'm using disconnect switches and fuses.

MCC is 600V, solidly grounded.

Would you get arcing ground faults without going phase - phase?

I couldn't find any arcing around switches or fuses, but I was looking through a clear plastic guard. I would expect to find fairly obvious damage.

Thanks for your attention,
Gord

 

[jbartos]

Suggestion: If the arcing is caused by a harmonic content, especially by the higher harmonics, than they will start heating and damaging insulation in locations where the conductor is closest to the ground through the conductor insulation

 

[Shortstub]

Gordonl,
You have a serious problem. This is the classical case of an "ineffective" ground-return path!

You "tub" bonding fix is a step in the right direction. Now you must correct the inability of the relaying to detect, and then operate quickly to clear the fault. What is the "grounding" circuit path, between the MCC structure, and the transformer neutral?

Also, the 51 is a phase relay, and its setting is such that it won't detect the rms value of an arcing-ground fault. The rms value could be as low as 5-10% of the theoretical phase-to-ground value.

 

[Modula2]

JBartos, what do you mean by harmonics causing ground fault. I would think a ground fault would be caused by a breakdown between phase and ground.

 

[busbar]

Conductor movement/mechanical stress can be significant during motor starting. In particular, the rattling of insulated conductors in conduits during inrush is sometimes quite memorable.

Is it possible that the ground conductor in the TECK cable has been incorrectly routed along with the feeder phase conductors through the zero-sequence CT feeding the CO-8?

Is there likelihood of near-simultaneous ø-ø and ø-ground faults, with part of the arc energy not seen by the 51 relay?

 

[Modula2]

Busbar, you seem to be making three points. The rattling of conduits during motor starting is new to me and unexpected in my applications, but I am sure it exists in yours. But what are you thinking this has to do with the situation?

In the second, I interpret as possible ways a ground fault not being detected.

Can you clarify your third point of near-simultaneous ø-ø and ø-ground faults as I am not sure what you mean. A phase to phase fault would be more likely to be sustaining, but arcing ground fault may not be.

 

[busbar]

The point was to offer suggestions to gordonl’s situation that had not already been mentioned. In an unenergized state, conductors positions and insulation may certainly have shifted, causing a fault that may not be immediately evident afterwards. 2L-G faults, among others, are considered in coordination studies, and have limited zero-sequence current; id est, arc energy may not have been 100% devoted to a ground-return path, in turn limiting 51 operation.

These certainly would not rate as top-of-the-list concerns, but shouldn’t be discounted if other possibilities had been ruled out. [It’s the best I could do a thousand miles away from the incident.]

 

[jbartos]

Suggestion to Modula2 (Electrical) Apr 19, 2003 marked ///\\JBartos, what do you mean by harmonics causing ground fault. I would think a ground fault would be caused by a breakdown between phase and ground.
///The higher harmonics are passing through conductor insulation that acts as a capacitive dielectricum. They heat the conductor insulation and may cause the insulation break thus forming arcing and ultimately ground fault.
Notice that the capacitive current Ic leaking through insulation is:
Ic=jx2xPIxfxV
where
PI is 3.14
f is frequency in Hz
V is voltage in volts from conductor to the ground
Therefore, the higher harmonic frequency f is, the higher Ic is, and higher cable insulation is.\\
\\

 

[jbartos]

Suggestion: Sometimes, damaged conductor insulation may cause arcing. If conductors are installed, a good care must be taken not to damage their insulation over sharp edges, sharp bends by tools, etc.

 

[gordonl]

Shortstub

The grounding return path is quite good. Approx #3awg. bonding conductor in teck cable solidly connected at both ends. (Visually confirmed)

In my inspection I had a close look at the areas where the cables pass in and out of the bucket, and found no damage.

I had a CH Appl. Eng. come in and have a look, and his opinion was that the damage wasn't due to a 600V L-G fault due to a lack of damage. I realize this assumes a good ground return path, but that seems to be the case, other than the connection from the MCC bucket to frame.

I haven't suspected a protection failure because the problem seemed to be transient. When the event was witnessed, the motors completed their start up, and ran normally.

We have experirenced a severe arcing fault on this sub in the past. The main incomming to the sub (which feeds this MCC) was busduct which had no ground conductor run in the bus duct enclosure. The bus duct arced phase-enclosure for a signifigant period of time before going ph-ph and clearing the upstream protection. (The transformer neutral ground protection failed to operate.) I haven't visually checked the transformer neutral grounding to the sub ground, I'll do that my next time out.