Nollföljdsmodellering av transformator-Beräkningar av följdproblem vid ökad kablifiering på mellanspänningsnivå
Historically transformers with system earth equipment have been modeled as open circuits in zero sequence. This approximation has been satisfying, but lately it has been discovered that equipment on one side of the transformer have responded to faults that occurred at the other side. Most of these problems took place when the majority of the system on the secondary side of the transformer contained cables. These findings have led to questioning whether or not the former model of the transformer in zero sequence is correct. This report investigates if it is necessary to represent the transformer in zero sequence in another way than it is usually done today. Zero sequence models of transformers that represent the reality more correctly will be presented. In the simulation program PSS/E the found models are applied and calculations are made for faults both at the primary and secondary side of the transformer. A power system has been created where the secondary side can be varied between containing only lines or cables.
The results show that the zero sequence voltages over the system earth equipment, at the low voltage side of the transformer, are very high when a fault occurs at the high voltage side. These voltages exceed the voltage for which the system earth protection trips. Another observation from the results is that the phase voltages on the secondary side exceed the construction voltage, when the fault occurs at the primary side. Both observations mentioned above happened when the secondary side of the transformer consisted of cables. They also occurred when the secondary system contained lines with the system earth resistance disconnected. When the fault were applied at the secondary side, the zero sequence current on the primary side exceeded the boundary for which the earth current protection trips.
A conclusion is that the former system model of the transformer, which assumes an open circuit in zero sequence, does not correctly demonstrate what happens during a fault. The model, which is presented in this report, is a more accurate representation and the results speak for themselves. No longer is it possible to disregard what happens at one side of the transformer when the fault occurs at the other side. Using a stabilizing winding resulted in a smoothing effect, which is why it is recommended to use one if it is possible. It is recommended to keep the system earth resistance connected at all times when the system contains mostly cable. A disconnected resistance result in a worse condition. The report illustrate that it is important to represent the transformer model in zero sequence correctly and complement the secondary system in zero sequence, which is not always done today.
A solution could be to run the transformer ungrounded primarily, or leave the secondary side ungrounded and put the system earth equipment in a grounding transformer. These solutions create an interruption in zero sequence.
The report is written in Swedish.