Study On The Dynamic Characteristics And Suppressing Methods Of Metal Particle Dust In DC GIS/GIL

With the rapid development of the DC transmission system,the DC gas insulated switchgears and transmission lines(GIS/GIL)have broad application prospects in the future power system.The metal particle contamination is a major factor threatening the safe and stable operation of DC GIS/GIL,in which the insulation faults caused by metal particle dust(MPD)occur repeatedly.However,the current research of MPD movement and distribution characteristics in DC GIS/GIL is insufficient.Therefore,it is difficult to give a targeted suppression method.Based on that,this paper firstly carried out the research on the dynamic characteristics of MPD in horizontal GIS/GIL,including its movement and distribution characteristics.Meanwhile,considering the special properties of MPD,the adhesion mechanism between MPD and rough walls was studied.Then,the new suppression method was proposed.Finally,a comprehensive suppression method of metal particles was given.Specific studies include:1)To achieve semi-quantitative analysis of the impact of surface charge on MPD movement and distribution characteristics,the surface charge inversion algorithm was studied.To address the shortcomings of slow calculation speed and low accuracy in current inversion algorithms.Based on integral form of the Laplace equation,the integral equation method(IEM)was proposed to calculate the transformation matrix.Compared with the traditional analytical method,IEM has the advantage of simple computation process and low computation cost.To suppress the numerical divergence in IEM,the Tikhonov regularization method was introduced and Tikhonov-IEM was proposed.The simulation results show that when the amplitude of the noise signal is the same,compared to IEM,the peak mean square error of Tikhonov-IEM results can be reduced by 40%.However,the simulation results show that Tikhonov-IEM is not suitable for the complex geometry,such as the basin insulator.Based on that,least square method(LSM)was introduced and LSM-IEM was proposed.The simulation results show that when the amplitude of the noise signal is the same,compared to IEM,the peak mean square error of LSM-IEM results can be reduced by 30%.Finally,the accuracy of the algorithm was verified by physical experiments.2)A test platform for the movement and distribution characteristics of MPD was built.The distribution characteristics of MPD in horizontal GIS/GIL was obtained.Based on the boundary element method(BEM),the model of MPD movement was established.And the influence of insulator surface charge on MPD movement and distribution characteristics was explored.The research results show that under DC voltage,the distribution of MPD in the axial direction of the chamber is wide,and the metal particle contamination at convex side of basin insulator is more serious.When there is no surface charge at convex side,the insulator will show the ’repulsion’ to the particles in multiple reciprocating motions.The homopolar surface charge will strengthen such ’repulsion’.While the heteropolar surface charge will weaken the "repulsion".3)Based on the atomic force microscope(AFM),the van der Waals adhesion between micron-sized metal particle and 1)epoxy resin(ER)insulators and 2)metal parts were measured.And the mechanism of influence of surface roughness on van der Waals adhesion was analyzed.A rough adhesion model suitable for GIS/GIL was established.The AFM tests results show that the magnitude of van der Waals adhesion force to micron metal particles in GIS/GIL is closed to that of the electric field force.The adhesion force will decrease rapidly with the increase of RMS roughness.When the RMS roughness value is greater than 40 nm,the van der Waals adhesion force tends to be a constant.4)Through simulation research,the influence of particle trap parameters on electric field and particle motion characteristics was obtained.The capture mechanism of particle trap in horizontal DC GIS/GIL was analyzed,and the criterion of selecting particle trap parameters was given.In view of the shortcomings of the current particle suppression methods,an online dust removal method for insulator surface based on gas flow was proposed.The influence of the inlet/outlet position on the dust removal effect was obtained.Based on 220kV GIL test platform,the feasibility of this method was verified.The research shows that after colliding with the high-voltage electrode,the electric field force on particle toward to the trap is the main driving force for trapping particles.The greater the trap thickness,the greater the axial electric field strength,also the better the particle trapping effect.The smaller the width of the trap slot,the smaller the electric field at the bottom.The greater the thickness and height of slots,the smaller the electric field at the bottom.During the dust removal,when the inlet/outlet is close to the insulator,the fluid will flow around the insulator surface,which will lead to the re-adhesion of metal particle dust.Based on the above research,a comprehensive suppression method for metal particles in horizontal DC GIS/GIL was proposed.Firstly,particle traps are installed at both sides of the insulator,also in the middle of the chamber.The parameters of traps are fixed according to the criterion of ’high slot width ratio’,’low field strength at bottom’and ’large thickness’.The online dust removal of the insulator surface is realized based on the gas flow during operation.