Search

The dc breakdown strengths of 4-15 [micro]m thick capacitor grade polypropylene (PP) films were determined. The measurements were performed with an automatic measurement system using five electrodes with areas between 0.045 and 9.3 [cm.sup.2]. Some samples were studied with larger electrodes. Samples were not aged but were measured as received. The measurements were performed in air at room temperature and typically 40-80 breakdowns per sample per electrode area were collected. The Weibull distribution parameters a and p were fitted to the data. Depending on the electrode area and the film grade, the a-parameter values obtained were between 450 and 850 V/[micro]m. The breakdown strength values showed decreasing area dependence with decreasing electrode area and the Weibull scaling law was not valid for the smallest electrode areas. The [alpha]-values for the large sample areas were extrapolated from the small area data by area and Weibull extrapolation. The area extrapolation overestimated the breakdown strength at 4 [m.sup.2] by 40-50% whereas the Weibull extrapolation gave an accuracy of [+ or -]15%. The extrapolated breakdown strengths for the full-scale power capacitors deviated from the capacitor manufacturers test data. This indicated that different defect distributions were responsible for the breakdown strengths in the oil impregnated power capacitors than in the small dry film samples. Index Terms--Polypropylene (PP), biaxially oriented PP (BOPP), breakdown strength, Weibull distribution, area dependence, extrapolation, capacitor, capacitor film, capacitor dielectric.

An automatic method for determining the breakdown strength of thin polymer films was evaluated. The equipment consists of a scanning electrode arm with which the breakdown voltage over the film sample is measured. The measurement electrodes were of solid brass and the ground electrode was of thin aluminum foil. Five different electrode areas between 0.045 and 9.3 [cm.sup.2] were used and 40-100 individual measurements per electrode area per sample were performed. All measurements were performed in air. The Weibull function parameters a and p were fitted to the obtained data. Different features concerning the measurement system and conditions, e.g. the criteria for the automatic detection of the breakdowns, the effect of the electrode edge design, the DC ramp speed, partial discharges and humidity were analyzed. In order to control the reproducibility of the results, also another type of measurement setup was tested. It was concluded that the measured [alpha]-parameter values were stable and repeatable over a period of several years. The average [beta]-parameter values varied [+ or -]1030%, more for the large than the small electrodes, and were also sensitive to changes in the sample, in the measurement conditions and the choice of the electrode system. Index Terms--Breakdown strength, polymer film, polypropylene film, dielectric film, capacitor insulation, measurement method, area dependence, Weibull distribution function.

A technique is developed for extraction of the wave propagation properties of power cables from S-parameter measurements. The method extracts the complex propagation constant and the characteristic impedance, as well as the LCRG Telegrapher's Equation parameters. The extraction process is developed after clarifying the effect of the connection between the measurement port and the power cable. It is concluded that treating the connection solely as a characteristic impedance change could lead to considerable errors in the parameter extraction. Furthermore, the method corrects for electrical lengths, which are not accounted for by the standard Network Analyzer calibration. The extraction is demonstrated for a medium voltage crosslinked polyethylene (XLPE) cable over the frequency range 300 kHz to 300 MHz. The results are compared to a time domain short pulse propagation method for cable characterization. Both measurement methods are evaluated against a cable model. Index Terms--Attenuation measurement, cross linked polyethylene insulation, parameter estimation, power cables, scattering parameters.

Thermally stimulated depolarization is modeled from dielectric response functions deduced from isothermal measurements. The key assumption made is that the same state of polarization can be achieved both via the real non-isothermal thermally stimulated depolarization (TSD) process and via an equivalent isothermal process. The change of polarization with temperature and time is traced through a continuous adjustment of the starting and stopping times and polarizing field of the fictitious isothermal process. The theory is applicable to general cases of non-Debye relaxation with variable activation energy. A comparison of the modeled and measured TSD thermograms is presented for films.

A HV dielectric spectroscopy system has been developed for diagnostics of water tree deteriorated extruded medium voltage cables. The technique is based on the measurement of nonlinear dielectric response in the frequency domain. Today's commercially available systems are capable of resolving low loss and small variations of permittivity as a function of frequency and voltage. Experience from more than 200 field measurements was combined with laboratory investigations. Small samples were used in an accelerated aging test to elucidate the correlation between water tree growth and dielectric response. Furthermore, field aged cables were investigated in the laboratory. It has been shown that the dielectric response of water tree deteriorated crosslinked polyethylene (XLPE) cables can be recognized and classified into different types of responses related to the aging status and breakdown strength. The influence of termination and artifacts such as surface currents was investigated. The measurement method enables us to separate the response of the cable from the influence of accessories. Finally, two different field studies of the implementation of the diagnostic method are presented. The field studies show that the fault rate decreased significantly when replacement strategy was based on the diagnostic criteria formulated.