Answers are grounded in this paper and cite their sources.
Mechanical damages in transformer windings during its nascent stages are usually confined to small portions of the winding and can therefore be assumed as localized and discrete in nature. Early detection of the first such event followed by its localization and severity assessment is desirable. Previous researchers have attempted this task by trying to capture correlations that exist between the deviations in measured frequency responses (using indices) and the known types of damage. But, experience reveals that this method lacks generalizability, thus limiting its use. This, perhaps, can be attributed to the lack of a theoretical basis to adequately capture the complex interactions between damages and the frequency response. With this motivation, this paper develops an analytical method (which is obviously generic) for localizing a discrete damage and estimating its severity. During the initial stages of development, the damage can be treated as a localized shunt capacitance change occurring at any arbitrary position in the winding. The analytical formulation is based on equating the coefficients of the driving-point impedance of the ladder network with one of its series/shunt capacitances arbitrarily changed to the respective coefficients of the rational function constructed from measured frequency response of the winding in the presence of a discrete damage. Experiments on an actual continuous-disk winding show that the results are promising.
No takes yet. Share an insight, caveat, or question.
Maulik et al. (2016) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: