Partial discharge analysis in distribution networks

Partial discharge (PD) is a localized electrical discharge that does not completely bridge the electrodes of an insulation system, yet it can significantly impact the longevity and reliability of electrical equipment. This study investigates PD phenomena using a combination of acoustic and electrical measurement techniques to comprehensively evaluate insulation health and identify potential fault sources. Measurements were conducted at eight different locations, encompassing varied environmental and operational conditions, to provide a diverse dataset for analysis. Electrical signals were recorded in COMTRADE format using a power quality device with a high sampling rate, while ultrasonic sensors were employed to capture acoustic emissions generated by PD events. The collected data were processed in MATLAB and analyzed through a multi-step approach, including waveform analysis, time-domain analysis, frequency-domain analysis, and phase-resolved partial discharge (PRPD) analysis. Waveform analysis provided insights into the temporal evolution of PD events, identifying critical anomalies and discharge patterns that could indicate insulation degradation. Time-domain analysis focused on peak detection, repetition rates, and the timing characteristics of PD events, offering valuable information about the periodicity and potential fault mechanisms. By applying Fast Fourier Transform (FFT), frequency-domain analysis revealed dominant frequency components associated with different types of PD, such as corona and surface discharges, enabling their classification based on frequency signatures. PRPD analysis was used to correlate PD activity with specific phases of the AC cycle, allowing the identification of phase-dependent discharge patterns. This approach was particularly useful in distinguishing between surface, internal, and corona discharges. Additionally, statistical parameters such as mean discharge magnitude, standard deviation, skewness, and kurtosis were computed to characterize the overall health of the insulation system and assess its degradation over time. A comparative evaluation across the 7 locations provided further insights into the influence of environmental factors and equipment aging on PD characteristics. The findings of this study highlight the critical role of integrated acoustic and electrical measurements in detecting and classifying PD events. By leveraging complementary measurement techniques, this research offers a robust framework for diagnosing insulation faults and developing predictive maintenance strategies. Furthermore, the identified patterns and dominant frequency components provide valuable benchmarks for future studies and practical applications in energy transmission and distribution systems. This work underscores the importance of PD analysis in ensuring the reliability and safety of electrical infrastructure. It also serves as a guide for implementing multi-method approaches to improve insulation monitoring and mitigate the risks of catastrophic failures in high-voltage equipment.