Research on transformer condition evaluation based on oil chromatography differentiation threshold

Abstract

As one of the important power transmission and transformation equipment, the reliability of transformer is directly related to the safe operation of power system. At present, the on-line monitoring technology based on dissolved gas analysis in oil is widely valued by researchers all over the world because it can continuously monitor the operation status of transformers and timely and effectively find the early faults of transformers. In this paper, the on-site oil chromatography data are differentiated and classified according to the voltage level of the transformer. The curve fitting method is used for statistical analysis and to build a distribution model. The cumulative probability is associated with the defect rate and failure rate of the equipment, and the corresponding threshold value is obtained by using the inverse cumulative distribution function as the attention value and warning value. By using this method, the attention values of the volume fraction of dissolved gas in oil and the gas production rate of dissolved gas in the existing national standards are improved, and the corresponding warning values are calculated to make the warning mechanism more flexible and reliable.