A Smart Grid Overvoltage Identification System Associated with Partial Discharge Signal and Dielectric Loss Detection

Guojin Chen, Yucheng Zhu, Zihao Meng, Weixing Fang, Wei Xie, Ming Xu, Wenxin Li2023 · DOI

TL;DR

This paper introduces a novel integrated system for smart grid overvoltage identification by simultaneously monitoring partial discharge (PD) and dielectric loss. By synchronizing these measurements with four current signals from the power grid, the system enhances hardware integration and provides a more comprehensive approach to diagnosing capacitive equipment health and identifying overvoltage events.

What Problem Does This Paper Address?

The increasing complexity and voltage levels in smart grids necessitate robust monitoring of capacitive equipment (like transformers and bushings) to ensure operational stability and prevent severe failures. Current methods for monitoring overvoltage, insulation degradation, and partial discharge often rely on multiple, unintegrated detection apparatuses, leading to cumbersome operations and a lack of holistic data. This paper addresses the need for a more integrated and efficient system that can simultaneously capture and correlate diverse electrical signals to accurately identify overvoltage events and assess equipment health.

Methods

The core of this work is a smart grid overvoltage identification system that correlates partial discharge (PD) signals with dielectric loss detection. The system utilizes a novel 4-way wide-frequency-range synchronous acquisition device. This device integrates the acquisition of:

  • Reference voltage: For phase comparison.
  • End-screen current (power frequency): Used for dielectric loss calculation.
  • Reference system interval end-screen current (power frequency): Also used for dielectric loss calculation.
  • High-frequency partial discharge (PD) signal: Captured in the 30 MHz range.

The system employs techniques like busbar voltage inversion to derive voltage from current signals and calculates the relative dielectric loss factor by analyzing phase differences between voltage and current waveforms. For PD detection, it focuses on high-frequency signals due to their distinct detecting features and reduced susceptibility to electromagnetic interference in outdoor substation environments. The hardware design emphasizes integration, including a power surge protector and 485 communication protection module to enhance anti-interference capabilities in harsh substation environments.

Datasets

Not discussed in the paper. The paper focuses on system design and presents results from field tests conducted on a 550 kV substation.

Key Insights and Ideas

  • Integrated Monitoring is Key: The paper strongly advocates for the synergistic monitoring of dielectric loss and partial discharge, arguing that combining low- and high-frequency signals provides a more comprehensive understanding of equipment health than individual methods.
  • Hardware Integration is Crucial: The authors highlight the limitations of existing multi-dimensional detection technologies due to poor integration and cumbersome operation. Their proposed system aims to overcome this by integrating multiple acquisition channels into a single device.
  • Wide Frequency Range is Essential: The system's ability to capture signals across a wide frequency range (50 Hz to 30 MHz) is critical for simultaneously detecting dielectric loss (lower frequencies) and PD (higher frequencies).
  • Substation Environments Demand Robustness: The design incorporates specific measures (surge protectors, shielding) to combat the high electromagnetic interference present in substation environments, ensuring reliable data acquisition.
  • Correlation Unlocks Deeper Insights: By correlating PD and dielectric loss data with overvoltage events, the system can more accurately diagnose specific faults, such as poor wiring contact or capacitive screen wetting.

Key Results

The paper presents results from field tests on a 550 kV substation, demonstrating the system's ability to:

  • Monitor Overvoltage: The system successfully captured and visualized bus voltage fluctuations, identifying intervals with sustained high-amplitude overvoltage (e.g., up to 1005 kV in Figure 27, and even higher in Figure 24 and 30).
  • Detect Dielectric Loss Anomalies: It identified instances where the relative dielectric loss factor deviated significantly from normal ranges, indicating potential insulation degradation (e.g., exceeding 1% in Figure 19 and Figure 22).
  • Identify Partial Discharge Events: The system detected high-frequency PD signals, including instances of significantly high instantaneous discharge magnitudes (e.g., 3200 pC in Figure 20, 11,928 pC in Figure 26).
  • Diagnose Specific Faults: By correlating these measurements, the system successfully identified various fault conditions, including:
  • Poor primary-side wiring contact (Figure 18, 20).
  • Poor secondary-side wiring contact (Figure 21, 23).
  • Capacitive screen wetting (Figure 24, 25, 26).
  • Insulated cable paper wrinkles (Figure 27, 28, 29).
  • High-amplitude overvoltage events without immediate damage (Figure 30, 31, 32).

Limitations and Future Work

  • Limitations:
    • The paper focuses on the design and validation of the system through field tests, but a comprehensive quantitative performance evaluation (e.g., accuracy metrics for fault classification) is not detailed.
  • While the system is designed for 550 kV substations, its scalability to other voltage levels or different types of equipment is not explicitly discussed.
  • The paper mentions the use of an "optimized FFT algorithm logic" but does not provide extensive details on its performance gains compared to conventional methods.
  • Future Work:
    • The authors imply that the system's comprehensive data acquisition and analysis capabilities pave the way for more advanced fault prediction and diagnosis.
  • Further research could explore the integration of machine learning algorithms with the collected data for automated and more sophisticated fault identification and prognostics.
  • Expanding the system's applicability to a wider range of grid equipment and voltage levels would be a logical next step.

Summary updated

BibTeX
@misc{chen2023smart,
  author = {Guojin Chen and Yucheng Zhu and Zihao Meng and Weixing Fang and Wei Xie and Ming Xu and Wenxin Li},
  title = {{A Smart Grid Overvoltage Identification System Associated with Partial Discharge Signal and Dielectric Loss Detection}},
  year = {2023},
  doi = {10.3390/s23187727},
}
Figure