5kV motor starter failure 2

[rockman7892]

We had an interesting series of events happen when one of our motor starters failed that has left me scratching my head. I wanted to see what others thought of these events and see if they had any input.

This morning we lost power to the plant when the main utility circuit switcher opened. Upon investigating loads in the plant we came to realize that we had a fault on a 4.16kV starter when trying to start a 1500hp motor.

The contactor used to start this motor is a 4.16kV Siemens vaccum contactor used in their 8100 series switchgear. After pulling the starter out we found that the fault was signifiganct enough to blow apart both of the vaccum contactor bottles on tha A & B phases of the contactor. Both of the R type motor fuses on the contactor were also blown on the A & B phases as well.

After going to our main switchgear lineup we found that we had an 87 differential trip on the differential relay protecting the cable betwwen the utility transformer and main switchgear lineup. I am puzzled as to why a fault downstream of the main swtichgear lineup would cause a differential trip on a zone which was upstream?

After isolating the faulted starter we restored power to the plant. Going back to the starter and motor we found that the source of the fault appeared to be powerfactor capacitors that were located on the load side of the motor contactor. The cap bank had two 200kVar caps in parallel and had the A&C phases blown on each cap. After isolating the cap it appeared that the cap was shorted as well.

So in trying to piece this all together I am trying to answer a couple of questions.

1) If there was indeed a short caused by the cap bank why would the contactor bottles explode? Shouldn't the R fuses have blown and protected damgage to any equipment? These fuses did indeed blow as well so if they blew i'm thinking they blew as a result of the arcing falult caused when the bottles blew

2) Why would this fault on a bus downstream of the main bus cause a differential relay to pickup on an upstream zone?

 

[davidbeach]

Differential operated on CT saturation. One end saturated and the other end didn't. Comparing current waveforms from the two sets of CTs the relay saw current in that didn't match the current out. Without a coordination study it would be hard to say why the fuses didn't protect the vacuum bottles.

 

[Elmir]

It hard to pinpoint the exact cause. All we can due is provide senarios without being on site first hand to investigate.

The fault current seen by the motor starter may have exceed the interrupting capacity of the motor starter.

Are you able to interrogate your differential relay for fault event data? This may provide clues. Possbilities for the differential relay tripping are numerous.

 

[LionelHutz]

The contactor probably began to open the fault or chatter on the fault causing it to fail. It needs a solid closure and hold during the fault.

CT saturation is a likely reason for the differential trip.

 

[rockman7892]

The differential relay is an older Siemens 70UT relay which I am not familiar with. I'm looking for the manual now to become more familiar with it and possibly try to find event information.

The CT saturation seems to make sense. I guess every CT is different and therfore has a differnet saturation point. It sounds like in this case the two CT's may have had slightly different saturation points and therefore one saturated and the other didn't. The CT's in reference are 3000:5

I am looking through the coordination study now to try to find what the avaliable fault current is at this starter as well as find coordination of the OCPD's. I am looking for the contactor rating as well.

Are most PF capacitors connected in wye or delta between the phases? My electricans are telling me that these capacitors failed open and are now currently reading open. I want to make sure that they are testing the capacitors correctly with their meters by letting them charge and discharge and looking at resistance values. Are there any red herrings to watch out fore when testing these caps depending on how they are connected?

 

[magoo2]

400 kvar per phase might be beyond the cap switching capability of the vacuum contactor.

As LionelHutz poined out, if the contactor doesn't make up solid, you rapidly build up voltage on the capacitors until something fails. Although vacuum switches are often used for capacitor switching, a vacuum contactor is not the same as a definite purpose capacitor switch.

On motor starting, before the shaft starts turning, the circuit looks mostly like a capacitor switch.

 

[rockman7892]

Attached is what I have found from the coordination study. The fuse used on the contactor is an 18R fuse. It appears that the avaliable fault current at the contactor is about 27kA.

I went and looked at the interrupting ratings of the contactor itself. At 4.0kV it had an unfused interrupting capacity of 5kA. With fuses it had a symetrical MVA of 350MVA. I'm not quite sure how to interperet these ratings however I would think that the contactor alone without fuses only has an interrupting rating of 5kA, and by adding the fuses this interrupting rating is brought up to 350MVA = 48kA.

When looking through the motor relay I noticed that the Multilin had a short circuit trip setting used. This setting was set for 8 x FLC (205A)= 1640A with a delay of .5s. I have heard and have read in our power study that these Siemens 97H35 contactors were not capable of interrupting short circuit faults and thefore should not have a short circuit setting in the relay. Tha attached TCC does not show an instantaneous setting on the relay however in the field someone must have set this setting.
Also the dropout time for the contactor I found to be listed as >340ms.

Is it possible from this information that a fault occured somewhere (possibly at the cap) and the multilin saw this fault current as an instantaneous value and tried to clear the fault with the contactor. Because this fault may have exceeded the 5kA capacity of the contactor alone it caused the mentioned damage to the contactor which in turn caused another fault which lead to the two 18R fuses being blown? I do not have the cap fuse curve to see where these fit into the coordination of everything.

I'm not sure I understand LionelHutz's comment about the contactor needing to hold during the fault. Is this basically saying what I mentioned above about the contactor trying to interrupt the fualt via the multilin short circuit setting. Should the contactor stay closed during the fault and rely on the fuses for interruption since the contactor itself is not rated high enough?

Where is a good place to look for the contactors capacitor rated switching capability. I could not seem to find this in the application guide.

 

[dpc]

Generally should not use short circuit trip function on MV motor starters due to the interrupting rating issue you mentioned. Short circuits are to be cleared by the fuse, not the contactor.

Some newer contactors have a built-in time delay on dropout to give the fuses time to clear.

 

[davidbeach]

If the delay of 0.5s wasn't enough to coordinate with the fuses, the current was probably below 5kA (I have not looked at the fuse curves); the short circuit setting is probably a red herring. I'm not particularly familiar with these starters, are they electrically held (similar to LV starters)? If electrically held, what is the source of the control voltage? Was it a CPT fed from one or both of the phases that had a fuse blow? Did the control voltage drop below the hold voltage during the fault? If so, the contactor may have opened under fault conditions, not from the protective relay but because there was no means of keeping it closed.

 

[rockman7892]

Because the 3000:5 CT's used on the differental relay most likely saturated is it safe to say that the fault that occured was a magnitude of at least 3000A or greater. I'm trying to find at what percent over their rated value CT's saturate at. With this said, I'm assuming that its possible that the initial fault could have been greater than 3000A, and since the S/C on the contactor is set for 1640A it would have tried to interrupt it and possbily caused further damage. The one thing that is puzzling me about the CT saturation is that if the 3000:5 CT's saturated why didn't the 1200:5 CT's on downstream breakers saturate and cause them to trip instantaneously.

We have had multiple cases of failures related to capacitors recently. All of these failures occur during our plant shutdowns and occur when shutting motors down or when starting motors up after a shutdown. (None of these failures seem to occur with plant in steady state) With that said I am starting to think that we may have capacitor switching issues or back to back capacitor switching issues occuring as others have mentioned. I was abl to find the instruction guide for the Siemens contactor and it said it was rated for capacitor switching in paragraph format, but did not give any specific values.

Just so I understand the phenomenom of capacitor switching I want to summarize what I have learned through research on the issue. When switching on a capacitor since the current leads the volate by 90deg the current tries to jump instantaneously. Because the current cannot change instaneously it changes rapidly and overshoots its peak value. From there it osilates at a high frequency unitl it settles down to steady state. This initial overshoot of current causes the voltage to be pulled to zero followed by an voltage overshoot and oscillation until the voltage settles down to steady state. These voltage and current oscillations can cause damage to switching devices as well as the capacitors themselves.

Back to back switching occurs when one capacitor in proximity to the other is energized. This appears as a short circuit to the energized capacitor and therefore this capacitor discharges into the capacitor being energized causing hish current and voltage transients. These high current and voltage transients can damage equipment as well as the caps themselves.

Does my understanding sound correct?

 

[davidbeach]

You were probably typing as I was typing. If you are correct about a 0.5s delay for the 1640A setting, it moves way down the list of possibilities, at 0.5s your fuses should have blown.

 

[rockman7892]

Davidbeach

These starters are electrically held similar to LV starters. The control voltage is 120V AC which is then converted to DC for the actual contactor coil.

The control voltage does come from a CPT which is connected L-L on phases B&C on the 4.16kV side. Since phases A&B fuses blew, phase B connetcted to the CPT was one of the effected phases.

The contactor operating data lists the dropout voltage to be Hot-50% rated and Cold-40% rated. This tells me that the contactor would drop out somewhere between 50-60V.

I like this possibility that you are presenting. Are you saying that the fault possibly at the capacitor caused the voltage to drop in the circuit thus forcing the control volage low and causing the contactor to drop out?

 

[burnt2x]

Rockman,
If the 120AC was converted to DC, does the circuitry of the rectifier have caps? I know DC coils drop out slower than AC coils, especially if rectifer have caps. 0.5 seconds could be more than enough for the fault to pull down the voltage below dropout level.
Davidbeach's scenario is a great possibility. Contactors trying to break short circuit currents usually blow out and get damaged.

 

[Laplacian]

I'm struggling to understand how a faulted capacitor could cause enough voltage drop to drop out the contactor; especially a DC coil type.

Was is the capacitor fuse size?
What is the capacitor lead wire size?
What is the capacitor configuration; 3 individual cans, or one 3 phase unit?
Was the fault phase-phase or phase-ground?
What is your maximum ground fault current; resistor limited?
What is your source transformer impedance and size?
What is the maximum phase-phase fault current on its primary?

27kA is a decent amount of fault current; how much fault current would have to flow to cause your contactor to drop out and for how long? I would think the capacitor lead wire would limit the current below this value. Maybe a fault somewhere else and the capacitor fuse was already blown.

Connect a variac to a similar starter test circuit and drop the voltage down to where the contactor chatters. Does it actually chatter, or does it seal itself out? Or does your contactor have two wire start/stop control where an external relay seals the contactor in?

I've had a contactor chatter problem with a different manufacturer and discovered the seal-in contact needed adjustment as it was breaking too long after the main contacts had separated.

 

[LionelHutz]

The interrupt rating of 5kA means it can open as long as the current is less than 5kA. The fused interrupt rating of 350MVA means it can withstand 48kA of current while the fuses clear the fault. In other words, if the contactor is closed and held closed it can withstand 48kA of current long enough for the fuses to clear.

So, as I hinted at before - if the fault current was higher than 5kA the contactor needs to stay closed. There are 2 main things that could cause it to open. The protection relay on the motor could trip and begin to open it. The control voltage could drop and allow it to open. Considering you had a short that may have pulled L-L voltage down close to 0V, the protection relay and the control voltage could have both been working to open the contactor.

FYI, it is common to see the protection relays running on UPS or DC power to avoid keep them powered during a fault. However, it is not common for the contactor to be on UPS power. As you have possibly seen, this can still cause problems.

 

[rockman7892]

The one thing that I can not seem to make sense of, is why a deferential relay would saturate before a downstream 1200A breaker would open?

As I mentioned the differential realays use 3000:5 CT's for the differential scheme. Downstream of this main incoming breaker there is a 1200A breaker with 1200:5 CT's feeder breaker which serves the bus which the faulted starter is located.

So if we had a fault that was large enough to saturate one of the 3000:5 CT's wouldn't it have saturated the 1200:5 CT's as well? I believe the instantaneous setting on the 1200A breaker is aprox 24,000A. I would think that this breaker should have opened if the 1200:5 relay saturated or saw a fault current high enough to saturate the 3000:5 CT's?

 

[climb1]

CT's have a power factor rating, most are 2. This means that the CT can handle 2 times the amount of current in the primary. Ex. if the 1200 is rated at 3, it would not saturate until 3600A.

FYI... the PFR is dependant on the gage of wire used.

 

[davidbeach]

What kind of CT are you talking about climb1? With no DC offset in the primary current, a C rated CT can supply 20 times rated secondary into a standard burden with less than 10% error. There may be a small amount of saturation in that 10% but certainly not much.

 

[climb1]

davidbeach, pfr is for continuous current ratings - my bad. you are right with no DC offset.

 

[PolyMike]

Maybe the problem is the vacuum contactor and capacitor bank combination. Because of the short interrupting time of vaccum switches/contactors, they can induce overvoltages (current chopping phenomenon), and the capacitors might amplify these overvoltages. Surge arresters are recommended(See EPRI Considerations for Conversion or Replacement of Medium-Voltage Air-Magnetic Circuit Breakers Using Vacuum Technology)
Maybe the fault sequece was:
1- During the motor start, a short-circuit at the capacitors is detected by the Multilin.
2- The Multilin orders the contactor to open, but owing to the high short-circuit current (not enough yet to blow out the fuses), during the open maneuver the contactor suffer the current chopping phenomenon (increased by the capacitor bank) and causing the bottles blown.
3- As the contactor couldn't clear the fault, the motor fuses blow; meanwhile the upstream differential protection fires beause of the CT's saturation.

The current chopping phenomenon of vacuum breakers could explain your problems with the capacitor banks.