Summary
Composite insulators operating in polluted environments are exposed to complex and highly variable combinations of climatic and contamination-related stresses that govern surface electrical activity. Leakage current is a sensitive indicator of such stress; however, conventional assessment methods based on single-parameter thresholds provide limited insight into real service behaviour. This paper presents a field-based investigation of leakage current activity at
Read more Read lessMartigues outdoor test station, one of the most severe natural exposure sites in Europe.
A composite insulator was continuously energized for more than two years under marine and industrial pollution. The representativeness of the monitoring period was validated through comparison with long-term climatic and pollution records. Unsupervised machine learning techniques were then applied to jointly analyse meteorological conditions and leakage current measurements, enabling the identification of recurrent climatic–electrical patterns.
The results show that leakage current activity is primarily associated with specific environmental regimes combining transient wetting and pollution transport mechanisms, rather than with extreme values of individual parameters. The proposed approach provides a physically interpretable, data-driven framework for improved condition monitoring of composite insulators under natural exposure.
Additional informations
| Publication type | Session Materials |
|---|---|
| Reference | B2_11440_2026 |
| Publication year | |
| Publisher | CIGRE |
| Country | Spain |
| Study committees | |
| File size | 1 MB |
| Price for non member | 30 € |
| Price for member | 30 € |
Authors
DE SANTOS Héctor - SHEMAR, Spain; MOAL Eric - SHEMAR, France; PONS Christian - EDF, France
Keywords
