The Study Of Fault Location Of DC1500V Traction Supply

It is a must to find and eliminate the faults as soon as possible after the short-circuit occurs. The fault location of DC traction supply for metro was studied to shorten the troubleshooting time after the faults in order to improve the reliability of the traction supply and ensure the safety of the metro.The DC traction supply is featured by low-voltage supply, short distance between adjacent substations, large traction currents and changes substantial frequently. The structures and the characters of the DC traction supply and the power system are quite different, so the fault location algorithm developed on the impedance based on travelling waves and the phasors could not be applied to the DC traction supply directly. The three-winding transformers was decoupled into two two-winding transformers, based on the two second windings use a first winding at the same time. The winding impedance of the transformer is transfer to the DC side, then the equivalent model of24-pulse rectifier is established. All the conductors of the traction supply are seen as cylinder. The sum of internal and external impedance of the conductor is seen as the impedance of the conductor. The DC traction supply is seen as a linear system and decomposed to a pre-fault network and a pure fault network, in order to avoid the effect by the load current. The pure fault network is used for fault location.The single-ended currents based, the two-ended voltage and currents based and two-ended currents based fault location algorithm are presented. To obtain the optimal solution, least square method is used to solve the redundant matrix equation. The simulation is taken under the PSCAD, at the same time, a simulation test is completed. The data of the simulation and the test are used to verified and improved the algorithms. The characters of the algorithms is evaluated after the test. The effects of the selection of the data window, the load currents, fault impedance, impedance of the conductor and the data unsynchronized are analyzed. The results show that the fault impedance has no effect on the error of fault location, the impedance of the internal impedance of the rectifier need to be calculated precisely for the single-ended algorithm. The two-ended algorithm has a smaller error than the single-ended. The parameter of the line has little effect on the two-ended algorithm. The result of fault location will be affected by the unsynchronized data. The two-ended currents algorithm has a better performance when it is applied to the fault between the catenary and the overhead ground wire.