Technology And Application Of High Impedance Fault Diagnosis At Distribution Networks Based On Wideband Synchronous Measurements

The distribution system plays an important role in achieving the integration of distributed generations(DGs)and ensuring the power supply’s reliability.The distribution system has wider coverage,more complicated topology,and more variable operation statuses,so it is more difficult to continuously guarantee normal operation and timely maintenance.According to statistics,faults in the distribution system and the caused outages account for 80%of the entire power grid.Besides,the weaker protection technologies,automation level,and management compared to developed countries make the effectiveness and reliability of fault diagnosis much far from the requirements.As a result,accelerating the construction of an efficient and intelligent modern distribution network has become an urgent issue to be solved.The high impedance fault(HIF)is a common fault type in the distribution system,which usually happens in the form of a line coming into contact with a high impedance ground surface or tree and will bring significant potential risks to public security.In recent years,fires,electric shocks,and facility damage caused by HIFs have been frequently reported worldwide,causing huge casualties and property loss.However,effective industrial solutions still haven’t emerged.As the weak features of HIFs,some simplifications used in traditional fault analyses will no longer be valid.The relationship between the system behavior and fault characteristics also becomes more complex.Furthermore,the nonlinearities caused by the arcing process of HIF and some nonlinear elements like the electronic-based DGs make the original description approach of fault features unreliable.In this thesis,the construction of a novel monitoring system in the smart distribution network is taken as the background.Then,the key is to address the problems dividing into three aspects:the mathematic modeling of fault characteristics,the detection of differences between normal and faulty states,and the exploration of the spatial distribution of wideband signals in the system for the fault location.Major researches and contributions are listed below:1)To realize the mathematical modeling of the variable nonlinearity of the high impedance arcing process,a distortion controllable(DIST-C)model is proposed based on the heat balance theory.First,the physical process of the air-solid dielectric breakdown during fault and the changing characteristic of nonlinearity are discussed.By fitting the features of the dynamic power during the dielectric breakdown,the independent control of the major dimensions of nonlinearity characteristic has been realized.The applicability of model parameters,the accuracy of nonlinearity simulation,and the convenience of use have been significantly improved.Besides,considering the collaborative numerical simulation of the fault and the system,a high-efficiency parameter determination procedure is proposed based on the specific configuration strategy of combined pulses.The time to determine the parameters of the HIF model is reduced by several tens of times.Verifications are also carried out.2)The analysis of kernel fault characteristics related to the system operation status in the distribution system with different neutrals has been implemented,and a comprehensive HIF detection algorithm based on the synchronous phasors and waveforms is proposed.First,the zero-sequence equivalent circuit is established to describe the nonlinear features when arcing HIF happens in a system containing nonlinear elements.On this basis,we decouple the wideband signals and analyze their characteristics under the variation of system operation status.For the distribution system with non-solidly-earthed neutral,an online calculation approach for the kernel system parameters,including damping ratio,detuning index,unbalanced degree,capacitive current,is proposed inspired by the geometry theories.It solves the calculation problem when the unknown parameters are less than the known conditions,making the threshold adaptive to the system parameters can detect HIFs over 5.3 kΩ.For the system with solidly-earthed neutral,the description of interval slope of zero-sequence current is proposed by combining the Grubbs-based modified robust local regression smoothing(Grubbs-RLRS)and the least linear square fitting(LLSF).The method realizes the uniform description of differentiated waveform distortion characteristics when HIFs take place under different fault conditions.Case studies and comparisons verify the effectiveness and the superiorities of the methods.3)By utilizing the synchronous measurements of wideband signals,a HIF location method is proposed based on the nonlinearity of dielectric breakdown.First,the coupling of nonlinear HIF and the nonlinear system is modeled.It is combined with the decomposition theory of wideband signals to derive the mathematical expressions of signal in each frequency band for faulty and healthy sections,respectively.Furthermore,the spatial distribution law of the wideband signal characteristics is systematically discussed when HIF happens in a system with different operation statuses,including the damping ratio,detuning index,and lengths of all transmission lines.This chapter mainly focuses on the resonant system.It proposes the HIF location algorithm based on the wideband zero-sequence inductive current,which enhances the robustness to system nonlinearity and fault intermittency,and is not dependent on the features in a short window.Take the 3rd harmonic component as an example,simulation and field HIFs are used to verify the effectiveness and comprehensive advantages of the algorithm.A software module of HIF diagnosis applied in the distribution network,which is integrated with the above algorithms,has been designed and developed,including the functions of interaction with synchronous measurement units,fault detection,and fault location.The module has been integrated into the dispatching center of the demonstration system for a National Key R&D Program project.Artificial fault tests have verified the effectiveness of the software and the correctness of the algorithms.