Flux coupling among primary secondary and teritary in three winding transformer 4

[menatpgande]

Can someone explain core and coil construction of a three-winding transformer and its magnetic flux coupling among the three windings. How can some of the equivalent T-models have a very high % value of ZT (e.g. 58%) while others have a very normal value (of single digit %, e.g. 5% to 7%).

 

[Marmite]

There are basically two alternatives for the transformer. It can have either close-coupled or loosely-coupled secondaries. Both LV windings have to be disposed symmetrically with respect to the HV winding so that they have identical impedances. There is a crossover between the two LV windings half way up the limb. There are two ways to perform the cross. Inner LV upper crossing to outer upper, and inner lower to outer lower gives loosely coupled windings. Upper inner crossing to lower outer and upper outer to lower inner gives closely coupled windings. To limit fault levels you would need to use a loosely coupled arrangement, but a loosely-coupled transformer will have a high degree of magnetic imbalance, so from that point of view you would need to use a close-coupled arrangement. This will be particularly relevant in the event of faults on the secondaries, and for a situation of low load on one of the secondaries, a closely coupled arrangement would be almost an imperative. If the transformer is represented by a T-section network then for the close-coupled arrangement all the impedance is in the upright part of the T, i.e. the primary. With the loosely-coupled arrangement all the impedance is in the secondary, which is represented by the cross piece.
One other significant disadvantage of the close-coupled arrangement is that with high load on one of the secondaries, both will experience high regulation.

Regards
Marmite

 

[menatpgande]

Thank you very much, sir, for your expertise and explanation. Most of them go over my head (ha ha). The crossing between the two LV windings (secondary and tertiary) kind of gives me some idea that there are two kinds of construct it hence two results. You mentioned imbalance magnetic flux on the loosely-coupled configuration. Is that (the imbalance magnetic flux) the one people said it would take the tank as a return path? Either way, could you help explain about this as well? When you say "all the impedance" is in the upright part of the "T" for close-coupled, and in the cross piece of the "T" for the loosely-coupled, what did you mean by "all the impedance"? I thought each branch of the "T" represents impedance of each winding (converting from the measurement between any 2 of the 3 windings). I am assuming that we are discussing about zero sequence. I am not sure if I am on the same page with you on this one.

 

[prc]

menat, let us talk in common language. What marmite explained is for the two LVs of equal MVA rating. I don't know whether you intended the same or one LV and a TV of a lesser rating.

Let us take the case explained by marmite. Loosely coupled or de-coupled LVs= There is relatively high impedance for the two LVs in T circuit and you will see negligible impedance in the HV limb of T circuit. Then LV1-LV2 impedance ( adding up the Z of LV1 &LV2 in T circuit) become very high The two LVs will be one above the other and HV also will be in two parallel halves one above the other. In effect, these are two, double wound transformers axially placed, but outside only three circuits brought out -H,LV1 &LV2. You can put the LVs radially also, on both sides of HV and then the same "loosely coupled" feature will be obtained but bit more complex in design esp when HV has tappings for voltage control. When one LV is loaded, HV limb with less impedance, the voltage drop through H will be less and hence the joint of T circuit will see almost same H voltage ie the other unloaded LV terminal voltage ( H+ Lv2 drop) will not change (ie voltage regulation in one LV is not affecting the other LV)

With closely coupled LVs, winding arrangement can be as explained by marmite. But it can also be like Core-LV1-LV2-HV, all radially arranged on core limb. In this case you will see LV1-LV2 impedance is very low, being near to each other( ie closely coupled magnetically),but the HV limb of T will be of high impedance.