Abstract

The paper investigates the method of detecting fault location in 35–10–6 kV electrical distribution networks based on empirical criteria in the coordinates of three symmetrical components, specific for each type of emer-gency mode. Voltages and currents of the direct, reverse and zero sequence are used in the formulas of the criteria. The formulas themselves were developed according to the following principle: if voltages or currents increase when the fault point moves along the length of the line, they are used as numerators, and if vice versa, the voltages or currents values are used as denominators. For each criterion, a graph of the interpolating function was obtained which is stored in the computer's memory. When a fault occurs, the value of the criterion is calculated and, according to the graph, the place of the emergency mode is found. The calculated mathematical error of the developed method is revealed with a small measurement error of ±0.2 %. It is shown that all errors are less than 1 %, except for single-phase earth fault modes, when the error reaches 2.6 %. As calculations show, the error of the fault location is directly proportional to the resulting measurement error and the accuracy of setting the initial data. So, with a resulting accuracy of ±2.5 %, the maximum error of the fault location is close to 3%, and with an accuracy of ±5 %, it is close to 6 %. Existing fault de-tection devices operate with an error of 3–20 %. Moreover, they are unable to detect damage with ground faults and breaks. The effectiveness of the considered method is confirmed by technical and economic calculations. So, based on one 35, 10 and 6 kV feeder, the economic effect is 200, 150 and 110 thousand rubles per year, respectively, with a payback period of about a year. When applying the method for 1000 feeders, the effect will be about 200, 150 and 110 million ru-bles per annum. Thus, the criteria obtained for fault location are effective from both technical and economic points of view.

Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call