Nuisance trip of CB ground fault protection

[zxfabb]

A power circuit breaker with integrated ground fault protection directly feeds a motor, the GF protection is performed by measuring the current of phase A, B and C, then calculating the vector sum in the electronic trip unit. If the vector sum exceeds the preset value, the breaker trips.

Sometimes nuisance trip occurs when the motor is starting. However, the insulation of cable and motor is verified good.

Current rating of the breaker is 630A, set value of GF protection is 0.2In=126A, time delay=0.2s. The trip event recorded in the datalog is I1:1838A, I2:1893A, I3:1885A, Ig:139A. (I1~I3: starting current, Ig:ground fault current).

The end user is doubting that the calculation accuracy of the trip unit is bad. I know the inrush current can cause spurious trip (because of DC component?). someone says it could be harmonics. But I don't think a motor itself is a big harmonic source. where does the residual current comes from and how to deal with it? Be grateful if you have any ideas.

 

[ron]

Check out this document for possibilities

http://www.littelfuse.com/~/media/p...ing-the-limits-for-ground-fault-detection.pdf

 

[zog]

A power circuit breaker with integrated ground fault protection directly feeds a motor, the GF protection is performed by measuring the current of phase A, B and C, then calculating the vector sum in the electronic trip unit. If the vector sum exceeds the preset value, the breaker trips.

Sometimes nuisance trip occurs when the motor is starting. However, the insulation of cable and motor is verified good.

Current rating of the breaker is 630A, set value of GF protection is 0.2In=126A, time delay=0.2s. The trip event recorded in the datalog is I1:1838A, I2:1893A, I3:1885A, Ig:139A. (I1~I3: starting current, Ig:ground fault current).

The end user is doubting that the calculation accuracy of the trip unit is bad. I know the inrush current can cause spurious trip (because of DC component?). someone says it could be harmonics. But I don't think a motor itself is a big harmonic source. where does the residual current comes from and how to deal with it? Be grateful if you have any ideas.

What type of breaker and trip unit? Harmonics can be a problem with certian types. I assume you are not in the US based on the ratings you provided?

 

[zxfabb]

What type of breaker and trip unit? Harmonics can be a problem with certian types. I assume you are not in the US based on the ratings you provided?

Yes, you bet. It is an IEC breaker: ABB Emax with PR122 trip unit.

I'd like to figure out more:
1. I suppose the DC components in three phases cannot offset against each other, since there are no phase angles.
2. What would happen for harmonics? Can they be offset? I know triplen harmonics cannot, which are in-phase and the sum is three times the individual magnitude. But what if a 3-phase/3-wire load like a motor? How does the sum current goes to the source?

Really confused, appreciate if any one can help me out.

 

[GoldDigger]

Since you describe a delta connection, the triplen harmonic current has no neutral to return on and must either be resolved as circulating current in the delta or actually be forced to zero.
But that does not by itself prevent capacitive current at the harmonic frequencies from returning through the PE connection to the neutral point of the service. That will be greater than the normal capacitive earth current which will both cancel out and be lower because of the lower frequency.

If the service/supply is truly ungrounded then the analysis will be different.

 

[zog]

Yes, you bet. It is an IEC breaker: ABB Emax with PR122 trip unit.

I'd like to figure out more:
1. I suppose the DC components in three phases cannot offset against each other, since there are no phase angles.
2. What would happen for harmonics? Can they be offset? I know triplen harmonics cannot, which are in-phase and the sum is three times the individual magnitude. But what if a 3-phase/3-wire load like a motor? How does the sum current goes to the source?

Really confused, appreciate if any one can help me out.

Well I don't think harmonics are the problem, that trip unit still uses analog circuitry and shouldn't be affected, however I have seen several of those trip units fail (Friday in fact) due to "drift" in some of the compoenents. Just aging issues from a 30+ year old unit.

Have the breaker tested per ANSI/NETA specifications, I suspect you will find a trip unit out of tolerance.

 

[mayanees]

I suspect it's a CT issue associated with saturation and non-linear performance during the high-current motor starting period and would increase the gf pickup to 33% of the phase pickup.

 

[zxfabb]

Well I don't think harmonics are the problem, that trip unit still uses analog circuitry and shouldn't be affected, however I have seen several of those trip units fail (Friday in fact) due to "drift" in some of the compoenents. Just aging issues from a 30+ year old unit.

Have the breaker tested per ANSI/NETA specifications, I suspect you will find a trip unit out of tolerance.

It is a totally new installation, we found a couple of breakers with this problem.

 

[zxfabb]

I suspect it's a CT issue associated with saturation and non-linear performance during the high-current motor starting period and would increase the gf pickup to 33% of the phase pickup.

The current measuring device built in the breaker is a Rogowski coil, it has no iron core to saturate thus highly linear even when subjected to large currents.

 

[zog]

It is a totally new installation, we found a couple of breakers with this problem.

My bad, I read PR112, which is a 30 year old obsolete trip unit. The PR122 is the new version (You would think they would change the name a little more than that)

Still, even new these are cheap Itallian breakers known for failures and you should have the breakers tested

 

[mayanees]

The current measuring device built in the breaker is a Rogowski coil, it has no iron core to saturate thus highly linear even when subjected to large currents.

I see that now, ~ 0.1% accuracy with those sensors.
I still like a 1/3 of phase pickup to account for inaccuracies, but if it's a new installation, and the breakers are suspect, certainly go that route.

 

[ghostbuster]

Instantaneous (milliseconds) high capacitive inrush currents have caused many GF circuits to trip during our investigations.(note:need a high speed scope to see them)