An Improved Non-Unit Traveling Wave Protection Method with Adaptive Threshold Value and Its Application in HVDC Grids

Research output: Contribution to journalArticlepeer-review

63 Scopus citations

Abstract

The performance of the traditional derivative-based non-unit traveling wave protection (TWP) in high voltage dc (HVDC) grids shows the problem of low-sensitivity when high-impedance faults occur. In this article, the characteristics of the fault initial traveling wave are analyzed first, two conclusions can be drawn: i) the distortion degree of the fault initial traveling wave is related to the fault location, ii) the amplitude of the fault initial traveling wave is related to the fault impedance. Then a zero-sequence current fitting method is proposed to obtain parameters respectively reflecting fault impedance and fault distance. When a fault occurs, a fault information extraction method based on the fitting parameters will give the estimation interval of the fault impedance, which provide reference for the adaptive threshold value selection of TWP. In this way, the influence of the fault impedance on TWP can be eliminated. A ±400 kV modular multilevel converter (MMC) HVDC grid is built in PSCAD/EMTDC. The verification of the proposed method considers different installation positions of the current limit inductor. The simulation results show that the proposed method can fast and correctly identify the dc line faults and effectively improves the protection sensitivity in detecting high-impedance faults.

Original languageEnglish
Article number8907453
Pages (from-to)1800-1811
Number of pages12
JournalIEEE Transactions on Power Delivery
Volume35
Issue number4
DOIs
StatePublished - Aug 2020

Keywords

  • Adaptive threshold value
  • current fitting
  • fault information extraction
  • HVDC grid
  • non-unit traveling wave protection

Fingerprint

Dive into the research topics of 'An Improved Non-Unit Traveling Wave Protection Method with Adaptive Threshold Value and Its Application in HVDC Grids'. Together they form a unique fingerprint.

Cite this