A Validation Method for UHF Partial Discharge Localization in Power Transformers Based on Pulse Injection

View at publisher
Authors: Zirui Wu, Cheng Pan, Zijun Pan, Qingdong Zhu, Hui Wang, Ju Tang
Cited by: 1

Abstract

Accurate localization of partial discharge (PD) in power transformers is critical for condition monitoring and fault prevention. However, localization algorithms are primarily validated through laboratory experiments or offline tests, making it difficult to assess their performance accurately in operational transformers. In this paper, a method is proposed to evaluate the positioning accuracy of PD localization algorithms by injecting pulse current into an ultra-high frequency (UHF) sensor to simulate PD source. The proposed method is first validated on an in-service 220 kV transformer equipped with six built-in UHF sensors. The results show that this method is safe, controllable, and capable of flexibly changing the location of PD. Furthermore, extensive experiments are conducted on another 220 kV transformer equipped with 18 built-in UHF sensors to systematically analyze the impact of sensor number and placement on the localization error of time difference of arrival (TDOA). It is found that increasing the number of sensors can reduce the likelihood of large positioning errors to some extent. But for highly accurate positioning, the best result is achieved when using five or six sensors. In terms of sensor placement, TDOA-based localization is less robust when sensors are placed on the same side or in adjacent areas. Uniform placement on all sides of the transformer can improve robustness, but it is difficult to achieve a high accuracy of the positioning error within 0.5m.

Fields of Study & Subjects

High voltage insulation and dielectric phenomena Power Transformer Diagnostics and Insulation Magneto-Optical Properties and Applications

Publication details

Journal / source
IEEE Transactions on Dielectrics and Electrical Insulation
Publisher
Institute of Electrical and Electronics Engineers (IEEE)
Pages
1–1
Publication date
1 January 2026