Strange behavior while energizing a transformer

[thermionic1]

We recently retrofitted new SEL relays into an existing 15kV Metal Clad Switchgear line up at an industrial facility. The gear was first installed in 1984 and built by an unknown switchgear vendor that did not supply a lot of details with the drawings. The gear has (3) bays that each feed a 1500 kVA D/Wye (solidly grounded) transformer. At the time of the event, it was only the MV cables and the transformer energized.

The original static relays were fed with the same CT's, however there was also a "Ground Sensor" zero sequence type CT that fed a separate ground relay. It was decided that the residual connection of the existing CT's would feed the separate IN input of the new relay, rather than the Zero sequence CT.

What we found was a bit of a random occurrence of ground current that tripped the relay. The secondary wiring was thoroughly checked prior to installation, including single point ground, megger @500V & secondary current injection from CT shorting block in both balanced and unbalanced 3 phase conditions all the way to the relay.

The plot on the left shows the currents, triggered by the breaker closing. The red trace is IN, which only starts to develop about 1.5 cycles after the breaker closes. It increases and trips according to the settings (IN PU=30A Pri, TD=2cyc) that were utilized in the previous relay. Not shown in the plot, but all Voltages and Currents look correct in the phasor diagram. Unfortunately, this is the filtered event and not the raw, unfiltered version. The plot on the right is the voltage, which doesn't show anything abnormal.

We have no idea exactly what the CT's are other than 300:5. There is a plan to test the CT's for all the normal criteria (ratio, excitation, burden, etc). At that time, the raw event reports will also be extracted from the relays.

This event occurred on 3 identical bays with identical CT ratios, relays and downstream transformers. In each case, the tripping is somewhat random, but in all cases the IN comes in 1-2 cycles after energizing and trips the breaker.

The next picture is the same transformer being energized and not tripping, with no IN (Yellow).

We are thinking it has to be the CT's or there is some kind of existing low grade fault. Any other ideas?

 

[unclebob]

If ground current is the red trace,isn't the frequency to low? Should not be 180 Hz?
2 cycles delay is too fast. On protection curves, inrush point of transformers is defined at 0,1 sec, at a multiplier of the FLA (8 - 12 x).

 

[davidbeach]

Get the raw event files from the relay and look at the CT saturation.

I’ll see your silver lining and raise you two black clouds. - Protection Operations

 

[stevenal]

Ic looks low in both cases.