Analysis of probabilistic properties of harmonic currents of loads connected to the high voltage networks

Abstract

The paper presents the results of a statistical analysis of the probabilistic properties of harmonic currents of the loads connected to the high voltage networks. The analysis was based on the measurements of current of three facilities: an aluminum smelter shop, a paper mill and a railway traction substation. These loads have a large capacity and receive electric power from the 220 kV network. The main technological equipment of the facilities is a source of harmonics. The facilities have different load curves. The load of the aluminum smelter shop during a day time remains almost constant. The load of the paper mill varies during a day time due to the shift operation. The load of the traction substation is abruptly variable. The analysis shows that the arrays of measured harmonic currents are non-stationary and contain anomalous elements. In the currents of each load there are prevalent harmonic components. It is established that in the process of measurement the direction of harmonic current flows changed. The harmonic currents sometimes flow from the network to the load and sometimes vice versa. The probability density functions for the values of harmonic currents in most of the cases are a composition of several known distributions.