A New Method Of On-line Monitoring For Power Cable Insulation

Because of the increasingly widespread use of XLPE cables, more and more accidents have been caused by cable breakdown. After incipient failures such as bad terminals and joints and some mechanical damages are ruled out, water-tree degradation is considered the main cause of XLPE cable breakdown. Since the 1980s, a variety of online monitoring methods have been developed to monitor insulation aging of the cables. However, to a certain extent, these methods have their limitations. Thus, they are not conducive to the large-scale popularization. On the basis of a comprehensive analysis of various online monitoring methods, this paper aims to propose a new online monitoring method for insulation, the monitor signals of which can be easily separated from those undesired signals for the convenience of extraction. Moreover, this monitoring method shows high accuracy.Based on the establishment of this 10kV cable aging model and simulation analysis, it is concluded that low frequency leakage current and low frequency dielectric loss factor values can be more sensitive to reflection of the water tree deterioration than power frequency signal of the cable. In order to verify this conclusion, this paper has artificially cultured a few different aging water tree cables and then verified the above conclusion by comparing the curves of low-frequency signals and frequency signals. Meanwhile this paper has proposed two promising sources of low-frequency signals: the inherent low frequency signals of the power system and the low-frequency signals generated by AC Superposition Method. The former still needs detailed study for application because of the uncertainty of the signals. Though theoretical analysis and simulation analysis of the latter one, this paper has concluded that the method can generate the required low frequency signals and the frequency has a fixed value, therefore, they are easily separated from power frequency signals and high frequency harmonic signals.According to the requirements of online monitoring of low-frequency signals, this paper has carried out the software and hardware design of this online monitoring system. In the hardware, taking into account the weak low-frequency signals, this paper has respectively designed the current and voltage signal extraction, filter amplifier and data acquisition and transmission. In the software, this paper has adopted the modular design method. The entire system has been divided into several modules in accordance with the tasks. This modular software design is conducive to extended follow-up. First design each module respectively, and then integrate each module together to complete the design, finally we can form a complete set of online maintenance system of XLPE cables.