Effective assessment of power transformer insulation using time-varying model without temperature constraints

This paper proposes a method for insulation diagnosis using the time-varying model. The parameters of the stated model are unique and can be recognized by the polarization current. Several methodologies have been reported for insulation diagnosis using various insulation models. However, moisture information without considering the effect of measurement temperature is bound to provide inaccurate result. Temperature significantly affects the dielectric response of materials. As the temperature rises, various important changes take place. Increased temperatures can boost molecular mobility, resulting in higher polarization, which in turn affects the material's dielectric constant and loss. Additionally, higher temperatures typically raise conductivity due to improved charge carrier mobility, further influencing the dielectric response. Hence, the effect of measurement temperature on insulation diagnosis is discussed in this paper so that responses recorded at different temperatures can be effectively compared. The proposed methodology for determining insulation state is tested using data from real-life power transformers.