Research On On-line Insulation Monitoring Method For Single-core Power Cables Using Power Disturbances

With the improvement of insulating material technology and the limitation of space resources,power cables,replace overhead lines,occupy an increasing proportion in power transmission and distribution systems at wider voltage levels.Because cable faults are more concealed,the fault detection is difficult,and the replacement needs a long period,the realization of on-line monitoring of cable insulation is of great significance to eliminate hidden troubles in time,prevent unplanned blackouts and ensure power supply reliability.The traditional on-line monitoring methods of cable insulation are mainly based on the analysis of power frequency monitoring signals or the diagnosis of artificial disturbance response signals.However,the information of the cable insulation condition contained in the power frequency monitoring signal is extremely limited leading the monitoring effect is not good,and the artificial disturbance signals will interfere with the normal operation of the power system.Thus,there is still an urgent need to develop a new on-line cable insulation monitoring method.Power disturbances are the transient processes different from power frequency steady states in power systems.They can be generated by various switching operations,transient faults,lightning strikes and so on.These disturbance signals not only exist naturally in the system,but also contain abundant frequency components and sufficient amplitude,which can make the cable produce transient responses which are easier to extract the change information of the insulation state.This paper makes full use of the natural power disturbances in the system,and takes single-core power cables as the object to study a new method to realize on-line monitoring of cable insulation status.The change of cable insulation state includes overall insulation aging and partial insulation defects.This paper,uses the external power disturbance of cable as high frequency excitation source,combines with the propagation analysis of the disturbance signal in the two cable metallic sheath bonding types including single-point bonding and cross bonding,studies the on-line monitoring method to realize cable overall aging.Based on the waveform analysis of power disturbance caused by sub-cycle arc fault,the method of on-line location of partial insulation defect of cable is studied.The main achievements of this paper are as follows:1)Distinguishing method for cable internal and external disturbance source using phase angle differences of disturbance signalsPower disturbance may originate from internal and external of the cable.The disturbances from different cable parts carry different information of cable insulation status.Distinguishing the internal and external cable disturbance sources is the precondition for correct use of power disturbances.Utilizing the character that the disturbance currents produced by the internal and external disturbances present different flow directions at the ends of the cable,this paper proposes to quantify the disturbance current direction by the phase angle difference of the voltage and current monitored at the ends of the cable.Based on this,the discriminant index of the disturbance source is established.The simulation and experimental results verify the validity of the method.2)On-line insulation aging monitoring method for single-point bonded cables using external power disturbances.For single-point bonded cables,this paper uses the high frequency excitation source generated by the external power disturbance of cable,analyzes the transient responses at both ends of cables,and proposes the on-line aging monitoring methods including identifying cable insulation parameters and extracting extremum frequency of insulation characteristic function.The on-line insulation parameter identification method is based on the frequency-dependent parameter model of the single-point bonded cable.The objective function relating to the insulation parameters is constructed by the amplitude spectra of the currents at both ends of the cable.The sensitive frequency range for parameter identification is selected by the frequency-domain sensitivity analysis method,and the parameters are optimized by least squares algorithm.The method of extremum frequency of insulation characteristic function utilizes the phenomenon that the extreme frequency at high frequency range deviates when the cable is ageing.A method for calculating the extreme frequency of the insulation characteristic function on-line is proposed by the equivalent insulation function which is obtained by measuring the grounding current and the cable core voltage at the head end.3)On-line insulation aging monitoring method for cross-bonded cables using external disturbances.Based on the analysis of the special structure of the cross-bonded cables,and using the high frequency transient excitation generated by the cable external power disturbance,the on-line aging monitoring methods by identifying the insulation parameters of three-phase cables and extracting the insulation admittance spectrum separately are proposed.The three-phase insulation parameters identification method is based on the solution equation of the cross-bonded cable model.Sensitivity analysis is used to select the appropriate monitoring signals and frequency range.The objective function of three-phase insulation parameters is constructed,and the differential evolution algorithm is used to solve the function.The insulation admittance spectrum method decouples the three-phase of the cross-bonded cable by the currents on the grounding lines and the cross-link lines.Based on the equivalent model of the decoupled cable,the loop equations are constructed to eliminate the irrelevant variables,and the phase-separated on-line extraction of the insulation admittance spectrum of the cross-bonded cable is realized.4)Sub-cycle arc model for power cables.Aiming at the problem of lacking sub-cycle(less than 1 power frequency period)arc fault model at present,this paper carries out cable arc experiments to build the model.The cables with different thickness of insulating layer were tested,and a large number of measured voltage waveform data of sub-cycle arc were collected.By analyzing and inducing the shape of sub-cycle arc voltage waveform obtained from the experiment,the CSC function sub-cycle arc voltage model with arc duration,maintenance voltage and recovery voltage as model parameters is proposed.In addition,based on the experimental data points of cable with different thickness of insulation layer,the reference values of arc model parameters of cable with different thickness of insulation layer are given by parameter fitting method.5)Cable partial insulation defect location method based on sub-cycle arc disturbance analysisUsing the sub-cycle arc fault caused by the cable partial insulation defect,a method for locating the partial insulation defect is proposed by utilizing the special shape of sub-cycle arc voltage waveform.In this method,the differential equation is constructed by the voltage difference between cable core and metallic sheath and the core current at the head end of the cable.The time domain waveform of the sub-cycle arc voltage is calculated along the length of cable.Then the calculated arc voltage waveform is fitted with the proposed sub-cycle arc model,and the fault location is pointed by searching the minimum fitting error point.Accordingly,the partial insulation defect location can be found.In addition,a variable scale searching method is proposed to improve the location efficiency.