Sequence Impedance estimate for a Wye-Delta-Wye transformer 2

[bacon4life]

Is there any way to estimate the sequence impedances for the Wye-Delta-Wye 12.47 kV/480V transformer in the attached image? I am trying to calculate how high I should set the ground fault relaying at the substation to avoid tripping on a feeder that has several similar Wye-Delta-Wye distribution transformers. On transmission transformers with tertiaries the impedance varies over a large range, but am hoping maybe the impedances are more consistent for distribution transformers. I tried contacting the manufacturer, but they do not have the sequence impedances available.

 

[davidbeach]

The scientific approach to getting the impedance of that transformer is to ..... guess. Or, determine it by test. Worst case would be to apply that 5.49% all the way around. It’s the P-S value, the P-T and S-T values will be higher, so using the lower limit gives the worst case. Since arc-flash is all 3-phase and the delta has little to do with that you should be ok there.

 

[bacon4life]

Ah, glad to know that P-T and S-T will always be higher. The transmission banks I have access to seem to have the zero sequence impedances about 90% of the positive sequence impedances. Will that ratio be similar for this size bank? Getting an outage to test is complicated. If the worst case assumption interferes too much with the substation relay settings, the outage for testing will be easier to justify.

 

[davidbeach]

Without a zero sequence test, for a three-phase transformer I assume R0 = R1 and Z0 = about 0.85 * Z1. That's based on a number of rule of thumb values, including St. Pierre. Seems to work reasonably well.

A problem in knowing the P-T and S-T impedances is that delta is buried and there are no 'Y' bushings to connect to or to leave open circuited. Not much point scheduling that outage; just wait for the first through fault. ;-)

 

[cuky2000]

In absence of manufacturer value If no zero sequence is known, and not specific tests data is available, 80% to 85% of the positive impedance could be used satisfactorily for estimating purposes. The negative sequence could be considered same as the positive sequence impedance.
Test of the specific transformer may not be available. However, it is hard to believe that the manufacturer do not has record of a test o a prototype unit.

 

[bacon4life]

Thanks, I think these estimates good enough to set the substation relaying high enough to avoid tripping in the future (and to confirm the cause of a past sympathetic trip).
Cuky-Good point, I only inquired about tests on specific serial numbers.