Power transformers are critical components in a power system network, and they are subjected to multifactorial stresses. Regular performance assessments are conducted, and the Health Index (HI) technique is sometimes adopted. HI integrates field test results into a single index representing the overall health. Existing Transformer HI (THI) models do not consider electrical and environmental factors unique to oil and gas facilities. Exclusion of operating history and failure data analysis affects the accuracy of the models. A pilot study was conducted using 4 existing THI models to investigate the accuracy of the THI scores in depicting the actual condition of a failed transformer installed in a gas processing plant. 2 models with 24 and 15 input parameters successfully indicated a significant deterioration within the transformer. Nevertheless, a final 'Fair' score would not trigger urgent intervention. With evidences that similar failure has occurred repeatedly across the facilities, the relevant field tests could have been included as dedicated input parameters to produce a more precise THI score. Another 2 models with less than 4 input parameters, placed the failing transformer in 'Very Good' band, which was totally erroneous. To design an accurate THI model, it is essential to consider specific load characteristics and environment factors by applying actual failure probability as basis. © 2021 IEEE.