Non-30° Phase Shfit on Transformer Differential Circuit of a Delta-Delta Transformer

[constantlylearning]

We are currently trying to trouble-shoot a differential circuit on a delta-delta power transformer and are experiencing a strange phenomenon. We are consistently seeing a 10° to 15° phase shift between the HV and the LV windings of the transformer (measuring from the HV and LV CT's). To the best of my knowledge, there should be no phase shift between the HV side and the LV side of the transformer. For reference, the CT's are wye-connected and are made up outside the delta on the HV and LV windings. The HV CT's are located on the HV bushings of the transformer. The LV CT's are located in an adjacent switchgear cabinet about 20 feet away from the transformer.

While trouble-shooting a previous differential trip (which we do not believe is related), we noticed a small amount of differential current on our differential relay. At first, we suspected that we had made an error in the magnitude of the correction factor between the HV and LV side of the transformer. Eventually, we wound up installing a digital fault recorder and discovered that the differential current is not due to a magnitiude error. Instead, it seems to be due to a phase shift between the HV winding and the LV winding. In my expereience, you might expect to see up to ~1° phase shift due to CT error, but I think 15° would be unheard of.

If memory serves, current in the HV and LV windings of each phase of a power transforer should always be in phase or 180° out of phase, depending on the winding configuration. If that is the case, the winding currents on the HV and LV winding should sum in an identical manner at the delta corners in each of their respective windings. So, I believe the phases currents between the HV and LV of a delta-delta should always be in phase, unless there is some kind of problem within the transformer. However, this phases angle shift we are seeing is balanced across all three phases, so I don't think this is a fault.

Incidentally, when installing the fault recorder, we also picked up an additional set of CT's on the LV bushings of the transformer (in addition to those in the adjacent switchgear) to create a second differential zone. Both sets of LV CT's agree.

Also, we believe the differential trip mentioned earlier was due to a contamination problem in the LV bus duct. That trip was a phases-phase fault and was not exhibiting the same type of symptoms we see here.

 

[waross]

Have you considered the contribution of the out of phase magnetizing current on a lightly loaded transformer?

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[constantlylearning]

We did briefly consider that, but the fault recorder trace on the source winding looks pretty clean. The transformer is only loaded to about 21% of it's base rating. So, I guess it's possible there is not enough load current to mask the exciting currently completely. Let me subtract the secondary current from the primary current and see if it is non-sinusoidal.

 

[DTR2011]

Do you have a voltage reference (Va, / Va-Vb) to compare your measurements to? How are the CT's connected(delta, wye?)

 

[dpc]

I'd check to make sure all of the CT secondary connections are tight and the lugs are properly crimped. Also, if there are multi-ratio CTs, the tap is something to check.

 

[arrehman1]

Are the HV and LV CTs connected directly to the differential relay? Perhaps there is some other device in the circuit adding considerable amount of reactance which may cause some phase shift?

I also agree with @dpc, I've heard more than once that untightened CT secondary connections can cause quite a bit of mysterious problems...

http://GeneralPAC.com

- Making power systems intuitive with awesome video tutorials!
The Wye Wye Transformer Connection video:

http://gpac.link/1GdGRSQ

 

[constantlylearning]

Unfortunately, we don't have a voltage reference tied into the circuit right now. Also, CT secondaries do a appear to be tight. We've made another observation. Viewing the mathematically derived differential currents on the fault recorders shows that the differential currents have a slight DC offset and asymmetry, along with a small amount of 3rd and 5th harmonic noise. (Not a lot, but it isn't a perfectly smooth sine wave.) However, the DC offset and the asymmetry leads me to believe that the core might be magnetized. I believe this offset and asymmetry might be translating to the differential current.

Just a theory at this point.

 

[DTR2011]

If you are able to take the system out of service, a CT excitation test will remove the residual magnetism on each individual CT. There is another method of doing it on line with resistors, but by todays standards, it would not be considered safe.

 

[constantlylearning]

For what it is worth, the differential current has gone away, which leads to me to believe that this was the result of residual magnetism in the core.

Thank you all for your responses.