Research On DC Bias Detection And Suppression Measures Of Transformer Based On Vibration Characteristics

In order to improve the trans-regional power transmission capacity and promote the renewable energy consumption,China’s "13th Five-Year Plan" proposed a large number of UHVDC transmission projects.While bringing power transmission convenience,it also causes a series of engineering problems.When the High Voltage Direct Current System works in Ground Return Status,a huge DC current is injected into the earth and invades the neutral point of the transformer,causing the transformer to generate DC bias.This phenomenon will cause the transformer core to be saturated,the excitation current is distorted,the harmonic content is increased,the loss is increased,the vibration is intensified,and local overheating and insulation aging are caused.at the same time,which also affects the safe operation of the AC grid.Therefore,this paper has carried out research on the detection method of DC bias and its restraining measures.The main research work is as follows:Firstly,this paper theoretically analyzes the DC bias generation mechanism of transformers and the vibration characteristics of transformers.Introducing the networked modeling method,the DC magnetic bias simulation model is established based on the 220 kV and 500 kV grids near the DC grounding pole in a certain area.The distribution of DC current in the neutral point of each transformer under different operating modes of the DC system is calculated.Taking one of the transformers as an example,the variation law of the excitation current and the variation of each harmonic are analyzed under different DC offsets.Then,a vibration information acquisition and testing platform under DC bias is built to collect the vibration acceleration of a single-phase transformer in no-load operation under different DC bias.The influence of DC bias on core vibration signal and its spectrum is analyzed.The under determined blind source separation algorithm based on EMD is studied,and the vibration signal of the iron core under load operation is separated,and the vibration signals of simulation and actual measurement are used for verification.At the same time,the wavelet packet energy entropy algorithm is used to extract the characteristic vector of the transformer core vibration signal under DC bias.The energy entropy analysis of the characteristic frequency band is used to judge whether the transformer has DC bias phenomenon,which lays a foundation for online detection of DC bias magnetism.Finally,the existing DC bias restrain measures are studied.Based on the neutral line series capacitors,a double protection measure of parallel spark gap and arrester is proposed.Among them,the zinc oxide surge arrester is used as backup protection for the capacitor,and the spark gap is used as backup protection for the lightning arrester.The effectiveness of this restraining measure was verified by simulation analysis.