Research On Earth Parameter Inversion And Optimal Grounding Problems

Grounding grids serve as an important role on the safety operation of power grid. As the bulk of substations become larger, the size of grounding grid has expanded and the configuration has become more complicated. The safety issues of ground potential rise, touch voltage and inner-grounding-grid potential difference, corrosion of grounding grid, which make grounding grid be the main research focus of electromagnetic compatibility in power system. To meet the requirement of design and maintenance of grounding grid for Chinese power grid system, optimal grounding is applied on existing condition to carry out earth parameter measurement and inversion, the optimal configuration and corrosion detection of grounding grid. Optimal grounding aims to obtain better effect of grounding grid design, maintenance and function.Multilayer earth is considered as the most common inhomogeneous soil, and the existing researches on derivation of Green's function are within5layers. Considering the limitation of theoretical derivation and numerical computation of Green's function, the intelligent derivation and computation method has been put forward to accomplish the multi-resolution and error auto control of complex image method of Green's function. By compared with the commercial grounding computational software CDEGS, the intelligent derivation and computation method has been proved to be a valid and effective method.The effect of distribution of fault current on the inner-grounding-grid potential difference is neglected in the past researches, which focused on the inner-grounding-grid potential difference caused by the current flowing into the earth only. Actually, the inner-grounding-grid potential consists of scattering inner-grounding-grid potential difference and conducting inner-grounding-grid potential difference. The computational model of inner-grounding-grid potential difference has been built to obtain the variation of the inner-grounding-grid potential difference. Lower earth resistivity leads to lower inner-grounding-grid potential difference. The larger grounding grid causes the larger inner-grounding-grid potential difference. The inner-grounding-grid potential difference increases and saturates with the earth resistivity and grounding resistance. When the earth resistivity is greater than100?·m or grounding resistance is larger than0.3?, the inner-grounding-grid potential difference will get into the saturated region. Proved by the analysis in this paper, the inner-grounding-grid potential difference of cable can be set to be50%of the total inner-grounding-grid potential difference during the design of grounding grid. When the inner-grounding-grid potential difference is greater than1.4kV, a copper conductor with two grounding terminals is recommended to decrease potential difference on the shielding layer of secondary cables.The model of conductor with decreased section area and iron rust coating has been proposed to simulate the corrosion of grounding grid. The model considers the grounding impedance, earth surface potential distribution and conducting resistance under different corrosion situation, which provides a detailed model for corrosion detection of grounding grid. The grounding impedance is not evidently affected by coating of iron rust. The coating of iron rust can be neglected during simulation of grounding grid corrosion.In order to obtain the objective function for four point method, the theoretical expression of objective function and its partial derivative have been obtained and computed by complex image method, which satisfies all the need of traditional optimal methods of direct search and gradient-based. The traditional optimal methods have been used as the inversion method of earth parameter on inversed error, iterative number and CPU time. Most of these methods are trapped into a local solution because the earth parameter inversion is a highly nonlinear problem. The least square method and trust region method are the better methods for earth parameter inversion for their performance on accuracy and numerical stability.To improve the consistance of the inversed results, the least square method and trust region method with constrained conditions have been proposed. The constrained inversion with weights can normalize the resistivity and reflect the upper and deeper soil parameters, so the least square method with constrains and weights is recommended as the earth parameter inversed method of grounding grid design. In order to obtain the apparent resistivity curve precisely, the general configuration of pole distance is recommended to increase by1:1.5for the adjacent pole distance.The swarm intelligent methods have been introduced to inverse the earth parameter. The performance of particle swarm optimal method, artificial firefly algorithm and artificial bee colony method has been examined. Some cases were listed to show the performance of these methods, and particle swarm optimal method and artificial bee colony method are the preferred methods. Though the initial value is not necessary, the swarm optimal methods are prone to local convergence. The traditional optimal methods trend to be convergent quickly but initial point is needed and the results sometimes are related to the initial point.The combining methods of traditional optimal and swarm optimal methods are put forward to improve the searching ability and convergence behavior. These methods include trust region particle method, least square particle method, trust region artificial bee colony method and least square artificial bee colony method. Some cases were used to show better results and higher robustness.The computation model of maximum step and touch voltage has been proposed with artificial bee colony algorithm to obtain the location and value of the maximum step and touch voltage for the equal-distance sampling detection on a certain observation surface cannot get correct result with enough accuracy. A compared case was listed to recognize the over-limit region of step/touch voltage, providing the consideration of application of high resistivity layer. Based on the artificial bee colony algorithm, the simplex method is used for optimal configuration of grounding grid to overcome the limitation of accuracy and computational speed of generic algorithm.The optimal configuration is affected by earth parameter and current injected point. After optimal configuration, the region with maximum touch voltage transfers to the center of the substation site. It just increases the risk of touch voltage shock, and certain equipment cannot be earthed locally. So the safety operation has been challenged and more rigid demand is needed for conduction, corrosion resistance, thermal stability and mechanical characteristic of the downleads.Though the touch voltage has been decreased, optimal configuration is with limitation because the over-limit touch voltage region is greatly increased, the safety of personnel is threatened and the costs for insulating surface layer increase. So the optimal configuration of ground grid is not practical.Based on constructed statistical process of corrosion of grounding conductor, the corrosion of grounding grid is modeled with the uncertainty of reduction of section area of the grounding conductors. The grounding resistance, the conducting resistance and the ground potential are used as statistical magnitudes during the Monte Carlo simulation of grounding grid corrosion. The influence of grounding grid corrosion on grounding resistance, the conducting resistance and the ground potential distribution is evaluated by Monte Carlo simulation. The validity and sensitivity of grounding grid corrosion is analyzed and some suggestions are recommended.The grounding resistance has less sensitivity for corrosion of grounding grid. The ground potential can show the totally corrosive effect on grounding grid but is affected by environment. It means ground potential is not a proper parameter to recognized grounding grid corrosion. The conducting resistance can show the serious corrosion nearby but it cannot reflect the corrosion faraway and the above ground network make conducting resistance less sensitive to corrosion. The on-line monitoring system for corrosion has been presented. By building the monitoring system of conducting resistance of the substation grounding grid, the conducting resistance data will be collected to show the slightly change of the conducting resistance and indicate the status of corrosion status of ground grid.