Old GE powermaster low voltage switchgear Arc flash incident 1

[adityalagudu]

One of Old GE Powermaster low voltage switchgear experienced a major arc flash incident a few days back. We are looking for a potential root cause for this incident.

Please see the single line for reference, there is no overcurrent relay protection on the gear. We are looking into several options of the cause of the incident, like dust in cubical, moisture in cubical, insulation failure, etc, but couldn't able to come to conclusion to one incident.

Based on previous experience, it would be helpful if anybody provides more insight into this incident. And we are also looking for further options to back this gear online, like refurbishing the breakers or upgrading the gear and replacing the gear with a new one, etc.

 

[adityalagudu]

see the single line for reference

 

[thermionic1]

The breaker is a GE Type AK (very very common years ago). Nearly every AK breaker I've seen has an integral trip unit with LT / Inst elements at a minimum.

It's difficult to tell from the photo what went wrong. The gear looks dirty, hard to tell was that existing or part of the fault. When was the last time this switchgear and breakers were maintained? When were the breakers last tested? I would think your insurance company would be asking these questions as well. The breaker with the 038 sticker appears to have a testing tag on it.

Is there any evidence in the (deenergized) cubicle of tracking or moisture ingress? It would be difficult on the picture alone to determine exactly what happened.

 

[wbd]

What was going on at the time? Was a breaker being racked in? Was it being closed? Was it being opened? Was there a fault on the circuit it feeds and it was tripping? What is the maintenance history of the breaker? What kind of trip units are on the breaker? Was it isolated to one breaker or the whole switchgear?

 

[adityalagudu]

The Breakers doesn't have any LT/Inst trip units on it, been though these breakers are feeding cooling tower pump motors.

Testing was done on the Breakers 2 years back and they have a testing sticker on it. The heaters are not working on the switchgear. The incident started at one cubicle and affected nearby cubicles because of the intensity of arc flash.

Breaker was racked in and was closed when the incident happened.

 

[Kiribanda]

From the photos, I would say that there had been a very high resistance arcing fault inside the switcgear.
Because it was totally black. The SLD shows a solidly grounded 480V system. Is it 100% correct?
Therefore, I have few questions to raise.
1) Was there adequate and a solid ground return path(s) from the swgr back to the the transformer neutral?
2) Per the SLD, the motor & other feeders are provided with 2000A current limiting (CL) fuses.
Were the bonding conductors of 3C/500MCM motor & other feeder cables properly bonded & grounded so that
there was adequate ground fault currents so that CL fuses will not operate for a long time below its
MINIMUM melting current to generate arcing faults?
3) Is it possible to check & verify the above points?

 

[wbd]

The Breakers doesn't have any LT/Inst trip units on it, been though these breakers are feeding cooling tower pump motors.

Has to. They all came with a trip unit. It may be the original EC device.

 

[RRaghunath]

"The Breakers doesn't have any LT/Inst trip units on it, been though these breakers are feeding cooling tower pump motors."
In that case, these breakers are meant to trip on receiving command from motor protection relays. Is that true!
If there is no protection built-in the breaker and presuming the motor protection and power contactors are near to the motor, ho is the fault in power cable from the switchgear to the pump control panel is expected to be cleared??
If there is nothing to clear this fault at the feeder level, the switchgear incomer is expected to clear the fault as a backup breaker and would have been time delayed. Is that right??
I think here lies the root cause if you look in to carefully.

 

[adityalagudu]

@ Kiribanda

Thank you for your input. Yes, the Transformer is solidly grounded. The fuses for the feeder are 1200A. We observed that one of the feeder breaker fuses were blown.

As far as the cable grounding between feeders and the Motor, these are 3/C 500 MCM cables. There is no grounding conductor between the feeder and motor. The motor is equipment grounded separately. But the cable doesn't have separately grounded. Is that what you are asking?

 

[adityalagudu]

@Rahunath

There is a Very OLD gear, there is no overcurrent protection with trip settings and delay time on the feeder breakers. In that case, there is no Incoming Main Breaker to isolate the Switchgear from Transformer. You can see it on SLD.

 

[Kiribanda]

Yes, this is the point I am referring to. If the feeder cables donot have a bonding conductor or if it is
not properly connected or the cables are not provided with a seperate grounding conductor,
then the zero seq. impedance for a GF is very high. As a result, the fault current is very limited
and contnued but still the current is not enough to blow the CL fuse. These CL fuses will operate fast
only when it is pushed to the CL region. Therefore, you cannot expect ground fault protection
by overcurrent protection (CL fuses) if adequate ground fault return path is not provided.
The alternative method is to install a ZCT for all cables and trip the bkr to provide GF protection.

1) Are they CSA TECK cables? If yes, what happened to their bonding conductors running inside the cable?
2) Are they TRAY cables? If yes, you should have provided a seperate grounding conductor along with the phase conductors.
3) Could you please upload the make/type/ data sheet of 2000A CL fuse?

 

[adityalagudu]

These cables are routed through Cable Tray. I will check for the grounding, I am confident that we don't have a grounding on the cables.

I don't have an exact 1200A Fuse photo, but I have 800A which is similar. See attached for reference.

 

[unclebob]

Drawings show that these breakers are withdrawable. It is quite possible that the resistance between the breaker stabs and the finger cluster became too high, causing much heat and an explosion. When servicing a breaker, many clean the main contacts inside but rarely the main disconnect behind. The breaker have to be put someplace else before accessing the stabs, hence cleaning these parts isn't easy.

Any picture of the faulty breaker?

I'd go with a new switchgear with a main breaker. That old one needs lots of cleaning before going back to service.

 

[Kiribanda]

From peak-let through curves of Class L/ CL fuses, you will find that, a 1200A fuse needs at least
16,000A let through current to initiate current limitation. During phase fault this
amount of current is possible whereas during a ground fault it is only possible
unless there is a "comfortable" ground return path is established. This effect is applicable for CL
circuit breakers too. There had been instances where these CL fuses had not operated but exploded.
Therefore, if overcurrent protection is used for ground fault protection, then one
to make sure that there is a solid ground return path provided. Otherwise, a ZCT has to be
used connected to a GF relay. So you have to perform a very thorough inspection to verify whether there had been a
arcing fault(s) (black marks) on entire 480V distribution.

After looking into your SLD I have few more questions to raise.
1) How does overload protection for the motors is provided? Is it by the 600A bkr shown?
2) What is the use of the 2x400/5 CTs shown in the SLD? Are they just kept short circuited?