Assessment of Macro Fiber Composite Sensors for Measurement of Acoustic Partial Discharge Signals in Power Transformers

Abstract

This paper presents a performance assessment of macro fiber composite (MFC) sensors for measuring acoustic emission (AE) signals from partial discharges (PD) in power transformers filled with mineral oil. MFC sensors are low-profile and flexible, allowing them to be attached to uneven surfaces, such as a transformer wall. Two types of MFC sensors were assessed: P1 (d33 effect) and P2 (d31 effect), which are optimized for different deformations in the structure, such as elongation and contraction, respectively. In addition, a conventional AE sensor, R15I-AST model from Physical Acoustics South America, was also used as a reference for comparative analysis. Four metrics were applied to the signals: root mean square, energy criterion, Akaike criterion, power spectral density, and correlation. The experimental results indicate a high similarity between the MFC sensors and the conventional AE sensor, which expands the research field in acoustic PD measurement in power transformers by using low-cost and flexible sensors.