Evaluation and prediction of contamination level in coastal region insulators based on leakage current characteristics

The paper addresses the serious issue on insulators owing to partial discharge and flashover resulted from severe NaCl salt contaminant accumulation on transmission lines insulators located near the coastal areas. Disc type insulators such as porcelain and glass were tested at various NaCl salt pollution level in an artificial fog chamber according to IEC60507 standard. The leakage currents associated with insulator types were recorded continuously at equal operating conditions (voltage, temperature and pressure). The statistical data of the leakage current such as the mean value (I
mean
), maximum value (I
max
) and standard deviation (σ) are considered. The recorded leakage currents are employed in Neural Network model as an input and using Feed Forward Back Propagation Network, the resultant contamination severity is predicted. This model is suitable to predict the Equivalent Salt Deposit Density (ESDD). The Neural Network Model is used to determine (predict) the intensity of leakage current (in mA) during security stage (<50mA) in the course of inception voltage.