Management: Prof. Rahimpour
Research partner: Hitachi Energy Germany AG
Duration: 2026-2029
Large High Voltage (HV), Extra High Voltage (EHV), and Ultra High Voltage (UHV) power transformers are critical assets in modern power systems. Their insulation systems are exposed to severe electrical stress caused by transient overvoltages (TO), particularly during switching operations, lightning events, and internal fault conditions. Among these phenomena, Very Fast Transient Overvoltages (VFTO), characterized by steep-front waveforms and high-frequency components extending into the MHz range, have become increasingly significant with the widespread adoption of Gas-Insulated Substations (GIS). These fast transients can excite internal resonances within transformer windings, leading to amplified local overvoltages and accelerated insulation degradation.
Accurate high-frequency transformer models are therefore essential for insulation design optimization, resonance analysis, and advanced diagnostic applications. Lumped Parameter Models (LPMs), typically implemented as RLC ladder networks, are widely used for applications such as Frequency Response Analysis (FRA), Partial Discharge (PD) localization, and transient overvoltage distribution studies. However, most existing LPM-based studies rely on the Disk Pair Model (DPM) as the fundamental modeling unit. While suitable for lower-frequency transient analysis, DPM-based approaches may lack sufficient spatial resolution for VFTO studies.
To capture the steep voltage gradients and high-frequency behavior associated with VFTO, finer segmentation of transformer windings is required. This project therefore focuses on the Single Turn Model (STM) within the LPM framework (STM-LPM), enabling turn-by-turn representation of winding parameters and improved accuracy in internal voltage distribution analysis.
Model validation will be conducted through experimental measurements on prototype windings, ensuring consistency between simulated and measured responses in both time and frequency domains. As the project progresses, findings will be disseminated through international conferences and peer-reviewed journal publications.