Research On Fault Location For HV Transmission Lines

Fault location algorithms for HV and EHV transmission lines can promptly and accurately detect the fault point. It not only can greatly decrease the restoring work and outage time, but also can detect the defective insulation. Fault location plays an important role in the security, stabilization and economic operation of power system. The micro-computer based fault location equipments are now widely used in HV and EHV transmission lines, the effects are good as a whole, but some limitations are also found through long-term filed operations. In order to solve these problems, systematic researches of principle problems of fault location for HV and EHV transmission line are discussed in this thesis.The basic principles of fault location using one-end or two-end data are derived based on theoretical analysis. From the derivation, it can be obtained that fault location algorithm using one-end data can't fully eliminate the influence of fault resistance and the source impedance at the same time. The two-end fault location methods totally solved this problem and are more accurate. In order to further improve the accuracy of two-end fault location, the influence factors which are ignored are discussed in this thesis.Short-circuit calculation and its inverse operation and the fault characteristics analysis are the basics of fault location algorithm. In this thesis, the equation of uniform transmission line is derived, which is based on the distributed parameter transmission line module. Also the decoupling of three-phase system is achieved. The formulas to calculate the voltage and current of each point along the line are obtained. The characteristics which are useful to locate the fault are discussed, which lays the basis for researching new fault location algorithm.For the situation that phasors obtained from both end of transmission line are asynchronous, the current two-end and three-end methods based on distributed parameters all required search, iteration or fitness, thus much more computation is required. And the accuracy depends on the length of search step. Taking account of those above, the fault distance and the asynchronous time are solved as the unknown variables. Positive component and pure-fault positive component are used to obtain simultaneous equations. From the equations, the asynchronous time can be eliminated. Thus the fault distance is presented by a simple equation. And the method to identify the pseudo-root is presented. Additionally the presented method is extended for three-terminal lines. Comparing to the existing similar methods, this fault location algorithm not only reduces the amount of computation, but also improves the accuracy of fault location.A bolted fault or a fault through a low resistance, especially near the source or substation, can easily cause saturation of the current transformer (CT) core, greatly distorting the secondary current and lowering the accuracy of the fault location based on fundamental frequency phasors. At present, this problem has not been well solved. In this thesis, based on the distributed parameter, the positive, negative and zero sequence voltage and current of each point along the line are calculated based on the short-circuit sequence network analysis without currents from the saturated terminal. And phase voltages and currents of each point can be obtained using the symmetrical component transformation. Because the fault impedance can be seen as a pure resistance, it can be derived that if and only if at the fault point, the voltage drop over the fault resistance and the fault current have the same phase angle, so the fault can be located immune to the saturation of current transformer. Taking account of that this proposed algorithm based on distributed parameter requires to search fault point leading to a large amount of computation, a more practical method based on the doubleπtransmission line module is proposed for the application for the transmission lines which are not very long. The Monotonicity of fault location function is demonstrated, so the fault point can be easily detected by the dichotomy search method. Thus the amount of computation greatly decreases.As the development of power system, more and more teed lines are used in the grid. At present, the existing fault location algorithms for teed lines need to identify the fault section first, and then convert the three-terminal line to a two-terminal line to locate the fault. While the fault occurs near the teed node, especially through a large resistance, the fault location methods currently used can't identify the fault section correctly, thus fail to locate the fault. So there is a dead zone of fault location near the teed node. For this problem, two fault location algorithms with different principle are presented in this thesis. In algorithm one, the voltage and current of each point along the line are obtained using the equation of uniform transmission line. Because of the fault impedance can be seen as a pure resistance, the fault can be located at the point where the voltage and current have the same phase angle value. Algorithm two assumes the fault occurring at one branch, respectively. Then calculate the teed node voltage and current flowing into the assumed fault branch from teed node. Finally three fault distances will be solved, respectively. It is proved in this thesis that there is only one fault distance meets the relationship that it is positive and less than the length of the assumed fault branch, which indicates the value is the true distance, and the fault happened at the assumed fault branch. These two methods both break the mold of traditional methods that it must identify the fault section before locating the fault. The proposed two methods both can locate the fault without identifying the fault section first. Compared to traditional methods, the highlighted advantage of these two methods is that fault occurring near the teed node through large resistance can be accurately located, and dead zone of fault location doesn't exist.