Acceptability of three transformer hottest-spot temperature models

The acceptability of three transformer hottest-spot temperature (HST) models for use in dynamic loading are investigated and discussed in this paper. The hottest-spot models are the modified HST model, Swift's model, and Susa's model. The investigation fitted the models to measured field data obtained from utility transformers using numerical and statistical methods. Acceptability metrics to evaluate the models' accuracy, adequacy, and consistency in predicting HST were evaluated using results from the fitting procedures. These metrics serve as absolute measures to reveal any potential deficiencies in the models when they are used to build hottest-spot models for transformers using field data. The metrics also serve as relative measures to compare the performance of the models to one another. It is shown that the modified HST model and Susa's model are acceptable for implementing models from field data. And it is concluded that the Swift's model is not acceptable because it is structurally inaccurate.