Study on Impurity Particle Detection in Transformer Oil Based on Microflow Imaging Method

ABSTRACT The concentration of particles in transformer oil has a significant influence on its electrical performance, making oil cleanliness a critical parameter in the operation and maintenance of power transformers. However, the light obstruction (LO) method recommended by current standards has inherent limitations, including low sensitivity to irregular particles and the inability to characterise particle morphology. To address these deficiencies, this study introduces microflow imaging (MFI) as an alternative technique for insulating oil analysis. MFI is a high‐precision method based on dynamic microscopic imaging that enables simultaneous measurement of particle concentration, size distribution and morphological characteristics. In this work, both LO and MFI methods were applied to oils collected from in‐service transformers and accelerated thermal ageing of oil‐paper insulation. For relatively clean oil, the ratio of particle counts measured by MFI to those obtained by LO ranged from 0.48 to 1.26, demonstrating good consistency. However, as contamination levels increased, the particle count detected by MFI reached up to 5 times that measured by LO. Notably, accelerated thermal ageing of oil–paper insulation at 130°C for 60 days resulted in an ∼28.94% increase in the proportion of elongated particles, with the total particle count measured by MFI being ∼4.82 times that obtained using LO. These results demonstrate that the LO method tends to underestimate particle quantities in oils with high concentrations of particles, particularly when irregular particles are present. By contrast, MFI exhibits significantly higher detection sensitivity. This study is the first to systematically demonstrate the technical advantages of MFI in evaluating particle contamination in transformer oil, providing a novel detection method for precise diagnostics of insulation conditions in power equipment.