Frequency Domain Spectroscopy Characteristics Dielectric Loss Normalization Method of Transformer Oil-Paper Insulation and Under Wide Temperature Range

As core components in oil-immersed power transformers, oil-paper insulation require precise dielectric property evaluation to enhance equipment condition monitoring. This study systematically investigates dielectric loss factor (tand) behavior across a broad temperature range <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\left(-20^{\circ} \mathrm{C}-70^{\circ} \mathrm{C}\right)$</tex> using frequency domain spectroscopy frequency domain spectroscopy(FDS) on oil-paper insulation samples in varying states. Analysis reveals an exponential relationship between <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\tan \delta$</tex> and temperature, driven by dielectric loss growth under thermal effects. Leveraging this trend, two temperature imputation models anchored to specific frequency points were developed, mitigating temperature fluctuation impacts on test results. Building on this discovery, the study proposes an innovative temperature normalization method for tan <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\delta$</tex>, standardizing measurements to reference temperatures regardless of testing conditions. This approach advances dielectric loss factor correction techniques for power transformers, offering theoretical and practical tools to improve equipment condition assessment accuracy. Its engineering applications hold significant value for real-world diagnostics.