An important tool for improving the reliability of HV insulation systems are partial discharge (PD) measurements. The interpretation of such measurements aims at extracting from the measured data information about insulation defects which then are used for estimating the risk of insulation failure of the equipment. Because the physical understanding of PD has made substantial progress in the last decade, it can now be exploited to support interpretation. In this paper a concept is presented which merges the available physical knowledge about various PD types into a generalized model which can be applied to arbitrary insulation defects. This approach will be restricted to PD of the streamer type in gases and at gas-insulator interfaces which cover a large fraction of the cases encountered in technical insulation systems. The generalized model allows us to derive approximate relations between defect characteristics, insulation design parameters and test conditions on one side, and measurable PD characteristics on the other. The inversion of these relations yields rules for extracting defect information from the PD data. The application of the generalized model is illustrated by two simple examples, namely, spherical voids in an insulator and electrode protrusions in SF/sub 6/. >

A generalized approach to partial discharge modeling

Experimental data are presented of the development of PD activity in artificial air-filled spherical voids in epoxy resin over a period of /spl sim/6 months. The data are presented in the form of phase averaged plots of the charge distribution over the time (aging plots) and phase resolved histograms (PRPD patterns) at various aging states. They are interpreted by comparison to numerical simulations which are based on a physical discharge model. This provides insight into the physical changes in the void during aging. The simulation is the basis for a physically meaningful interpretation of the PD data. It provides rules for the identification of the defect type, the defect quantification, and an assessment of the aging stage. >

Measurement and simulation of PD in epoxy voids

The decay of surface charges on polymeric insulators in air and SF6 is studied. Surface charges decay by three mechanisms, by volume and surface resistivity of the insulator as well as by neutralization from the gas. These mechanisms are investigated. In particular the mechanism of gas neutralization is explained in detail. It is shown that within a certain volume ions generated by natural radiation lead to a neutralization of surface charges on specimens and can contribute significantly to surface charge decay. Relevant gas parameters like ion pair generation, ion mobility and ion-ion recombination coefficient are given for air and SF6. The decay of surface charges generated by corona discharge is simulated by taking the three mechanisms into consideration. The intention of the simulation is to quantify the effect of the individual mechanisms of the decay of surface charges and to provide the patterns needed to interpret the experimental result given in part II of this study.

Surface charge decay on insulators in air and sulfurhexafluorid - part I: simulation

A survey is given of the present state of understanding of leader breakdown of electronegative gases in nonuniform field gaps and under fast-rising voltage waveforms. The basic physical processes involved are the formation of a pulsed streamer corona, the transformation of one of the streamers into a leader step, the temporal development of the leader channel, and the stepped propagation of the leader through the gap. The growth and structural characteristics of the corona are modeled and the results obtained are used to derive simple approximations for the spatial corona extension and the corona charge. Two leader inception mechanisms are discussed, namely, the stem and the precursor mechanisms. Inception criteria and characteristic time scales are derived for both of them in terms of the experimental parameters. A simplified model, which also predicts the random aspects of leader propagation through the gap, is presented. Some typical gas-insulation design problems are treated as examples. >

The mechanism of leader breakdown in electronegative gases

For pt.II see ibid., vol.6, no.5, p.35 et seq. (1990). The question of how to relate the magnitude of the measured, effective, partial discharge (PD) to the size, shape, and position of the cavity within a solid dielectric system is addressed. The causes of cavities are examined, and a brief review of gas breakdown is given. It is then shown that for a range of geometries, such as parallel plate and coaxial, one can develop analytic solutions (i.e. formulas), and that for any system the solution can be found using computer-based methods for calculating the electric field throughout a sample. Scaling rules are derived from the physical model. >

Partial discharge. III. Cavity-induced PD in solid dielectrics

Inhomogeneous field breakdown in SF/sub 6/ has been investigated, and a theory has been developed which predicts the breakdown voltage for fast-rising (rectangular) waveforms applied to arbitrary electrode geometries and gas pressures. This theory facilitates the analytical design aspects of many GIS components where previously, engineering judgement and development testing were necessary. It is suggested that extensions of the theory, will permit breakdown voltage prediction for an arbitrary surge waveform, which will facilitate the assessment of test waveform efficacy and the development of a more reliable field, factory, and type tests. >

Inhomogeneous field breakdown in GIS-the prediction of breakdown probabilities and voltages. I. Overview of a theory for inhomogeneous field breakdown in SF/sub 6/

For pt.I see ibid., vol.3, no.3, p.923-30 (1988). An improved model for the calculation of statistical time lags in SF/sub 6/ insulated electrode arrangements for positive impulse voltage is presented. On the basis of physical considerations, the production rate of initial electrons is ascribed to the local density of negative SF/sub 6/ ions and to field dependent electron detachment. Values of negative ion densities are measured directly, while the detachment coefficient is determined by a fitting procedure to measurements of statistical time lags. >

Inhomogeneous field breakdown in GIS-the prediction of breakdown probabilities and voltages. II. Ion density and statistical time lag

For pt.II see ibid., vol.3, no.3, p.931-8 (1988). Extensive investigations into various phases of discharge development in SF/sub 6/ (corona formation, streamer to leader transmission, leader, and propagation) are summarized, with the goal of familiarizing the engineer with the theoretical and empirical basis of a quantitative model SF/sub 6/ breakdown, including breakdown in highly inhomogeneous fields. The quantitative predictive capability of this model is tested for a number of examples, which demonstrate the correct prediction of geometry and pressure dependencies of the breakdown voltage. Combined with the modified volume-time 'law', the breakdown model becomes a valuable tool for the power engineer, capable of predicting statistical breakdown characteristics for a wide range of gap geometries and positive voltage waveforms as a function of pressure. The model does not cover negative polarity waveforms and becomes increasingly inaccurate as the waveform risetime increases over about 10 mu s. Future research planned to address both of these limitations and practical applications are outline. >

Inhomogeneous field breakdown in GIS-the prediction of breakdown probabilities and voltages. III. Discharge development in SF/sub 6/ and computer model of breakdown

F. Pinnekamp

Papers

Inhomogeneous field breakdown in GIS-the prediction of breakdown probabilities and voltages. I. Overview of a theory for inhomogeneous field breakdown in SF/sub 6/

Inhomogeneous field breakdown in GIS-the prediction of breakdown probabilities and voltages. II. Ion density and statistical time lag

Inhomogeneous field breakdown in GIS-the prediction of breakdown probabilities and voltages. III. Discharge development in SF/sub 6/ and computer model of breakdown