This paper presents an investigation into the nonlinear dynamics characteristics of the partial discharge (PD) sequence measured during the propagation of typical electrical trees in XLPE cable insulation. Calculation of the correlation dimension and the largest Lyapunov exponent from the PD sequence is used to illustrate the nonlinear characteristics of the PD behavior in the electrical tree growth. The results provide strong evidence for the existence of deterministic chaos in the PD phenomenon of electrical trees, where the correlation dimension and the largest Lyapunov exponent are related with the PD activity of the tree growth, rather than a presence of random stochastic process. The fluctuating correlation dimension of the PD sequence during the branch-pine tree growth is associated with the transition of PD patterns and the change of the channel conductivity of the tree structure. It is also found that the nonlinear dynamics characteristics of the PD sequence are influenced by the applied voltage, the channel conductivity and the electrical tree shape. A possible mechanism is presented for the formation of different chaotic extents of typical electrical trees in XLPE cable insulation.

Nonlinear time series analysis of partial discharges in electrical trees of XLPE cable insulation samples

The present disclosure is directed to cable diagnostic test methods, systems and apparatus that advantageously utilize “standing wave”principles to facilitate the identification and location of defect(s) along a power cable. The disclosed methods/systems are effective in measuring dissipation factors and dielectric constants associated with shielded power cable insulation at any number of points or sections along the axial length of the cable. In essence, the disclosed methods/systems perform what may be termed axial tomography, allowing the dielectric loss or dissipation factor and the dielectric constant of the insulation as well as the resistance and inductance of the cable conductor system to be determined at one or more pre-determined points/sections of the cable along its axis.

Diagnostic methods for electrical cables utilizing axial tomography

Classification of partial discharge events in GILBS using probabilistic neural networks and the fuzzy c-means clustering approach

Partial discharge (PD) field tests are widely used for cable assessment or cable commissioning. Usually, these tests are carried out by means of off-line tests where PD pulses are measured with capacitively coupled PD sensors. In off-line single-sided cable measurements, time-domain reflectometry (TDR) is used to locate the source of the PD along the cable. However, when the PD pulses are originated at the cable ends, the TDR technique cannot, by itself, properly distinguish between PD pulses produced at the cable near-end or at the far-end. This paper describes three different methods that produce an unambiguous localization of PDs produced at the cable ends, complementing the TDR.

Localization techniques of partial discharges at cable ends in off-line single-sided partial discharge cable measurements

The conducting and non-conducting properties of electrical trees in cross-linked polyethylene (XLPE) cable insulation are studied by means of partial discharge (PD) measurement, optical microscope, confocal Raman microprobe spectroscopy and scanning electron microscopy (SEM). Specimens equipped with an embedded electrode system and constituted of a needle separated by approximately 2 mm from the counter electrode were utilized. Various types of electrical trees were produced at a range of voltage levels from 9 to 15 kV. It was found that the electrical trees grown at low voltage levels exhibited very different electrical properties from those grown at higher voltage levels. Branch-pine trees were formed at the low voltage conditions, displaying buildup of conducting main channels composed of disordered graphitic carbon deposited on the tree sidewalls. The average domain of the graphitic residues was estimated at the range of 8-8.4 nm in size, which resulted in a sufficient channel conductivity to suppress the PD activity within the main tree channels and caused growth of the pine structure. The branch trees, grown at higher voltage levels, showed typical characteristics of non-conducting trees, in which continuous discharges eroded the tree sidewalls and the observed intensity of fluorescence decreased evidently with the distance from the needle electrode to the tree tip. The bush trees at high voltage levels also showed the non-conducting characteristics. However, some carbonized residues were formed locally in them near the needle electrode, which can be connected to the long and intensive discharge activity during the tree growth.

On the conducting and non-conducting electrical trees in XLPE cable insulation specimens

This paper describes typical cable defects uncovered while testing over 9,000 km of medium voltage XLPE insulated cables. After a brief review of the testing method, the procedure that led to the identification, localization, and characterization of typical defects found in operating cables is described. The paper also discusses the "conversion" of water trees (WT) to electrical trees (ET) which are associated with partial discharge (PD) activity in the context of laboratory research. Based on actual service performance of cables, several cases in which ETs associated with WTs did not lead to cable failure even after several years of service in very harsh operating environments are cited

M.S. Mashikian

Papers

Medium voltage cable defects revealed by off-line partial discharge testing at power frequency