Abstract

In this work a simplified procedure for the formulation of distribution transformer model for studying its response to lightning caused transients is presented. Simplification is achieved by the way with which the model formulation is realised. That is, by consolidating various steps for model formulation that is based on terminal measurements of driving point and transfer short circuit admittance parameters. Sequence of steps in the model formulation procedure begins with the determination of nodal admittance matrix of the transformer by network analyser measurements at the transformer terminals. Thereafter, the elements of nodal admittance matrix are simultaneously approximated in the form of rational functions consisting of real as well as complex conjugate poles and zeros, for realisation of admittance functions in the form of RLCG networks. Finally, the equivalent terminal model of the transformer is created as a π-network consisting of the above RLCG networks for each of its branches. The model can be used in electromagnetic transient or circuit simulation programs in either time or frequency domain for estimating the transfer of common mode transients, such as that caused by lightning, across distribution class transformer. The validity of the model is verified by comparing the model predictions with experimentally measured outputs for different types of common-mode surge waveform as inputs, including a chopped waveform that simulate the operation of surge arresters. Besides it has been verified that the directly measured admittance functions by the network analyser closely matches the derived admittance functions from the time domain impulse measurements up to 3 MHz, higher than achieved in previous models, which improves the resulting model capability of simulating fast transients. The model can be used in power quality studies, to estimate the transient voltages appearing at the low voltage customer installation due to the induced lightning surges on the high voltage side of the transformer. The procedure is general enough to be adapted for any two-port devices that behaves linearly in the frequency range of interest.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call