A Multi-Level Operation Method for Improving the Resilience of Power Systems under Extreme Weather through Preventive Control and a Virtual Oscillator

Chenghao Li, Di Zhang, Ji Han, Chunsun Tian, Longjie Xie, Chenxia Wang, Zhou Fang, Guanyu Zhang2024 · DOI

TL;DR

This paper introduces a novel multi-level operation method to boost power system resilience against extreme weather. It combines a predictive preventive control strategy to anticipate and mitigate cascading failures with an adaptive virtual oscillator control for faster and more stable frequency regulation, significantly outperforming traditional methods.

What Problem Does This Paper Address?

Extreme weather events are becoming more frequent and severe, posing a significant threat to the stability and reliability of power systems. These events can lead to cascading failures, where the loss of one component triggers a chain reaction of failures, potentially causing widespread blackouts. Existing resilience strategies often focus on responding to failures after they occur. However, this paper highlights a critical gap: the need for proactive measures that anticipate and prepare for successive failures, especially in isolated or weakened grid sections. Furthermore, in such scenarios, maintaining frequency stability becomes challenging due to reduced inertia and limited resources. Traditional frequency regulation methods are often too slow to effectively counter these rapid disturbances.

Methods

The proposed solution is a two-tiered approach:

  • Upper-Level Preventive Control: This novel method focuses on predicting the time intervals between successive line failures using a Poisson model. Based on these predictions, it implements a preventive control strategy that considers both the ramping capabilities of local generators and the transmission capacity from the external grid. The goal is to ensure that the local power supply can meet demand even if multiple lines fail, by proactively adjusting generator outputs and power flows before failures occur. This is achieved by optimizing a cost function subject to power flow, reserve, generator operation, and crucially, the power constraint between successive failures.
  • Lower-Level Adaptive Virtual Oscillator (VO) Control: This strategy addresses frequency stability in the event of failures. It employs a Virtual Oscillator (VO) control method, which mimics the behavior of synchronous generators but with faster response times and without requiring a Phase-Locked Loop (PLL). The key innovation here is an adaptive VO control strategy where the VO loop coefficients (specifically virtual capacitance) are dynamically adjusted based on the system's operating conditions. This allows for a larger virtual inertia during frequency drops and a smaller one during recovery, optimizing the frequency response.

Datasets

The study utilizes the IEEE-39 power system for simulation experiments. This benchmark system includes 38 transmission lines, 10 generators, 3 wind farms, and 6 energy storage units. Specific parameters for generators, wind farms, and energy storage units are provided in Tables 1, 2, and 3, respectively. The parameters for the VO controllers of the energy storage units are detailed in Table 4.

Key Insights and Ideas

  • Predictive Resilience is Crucial: The paper strongly emphasizes that anticipating and preparing for successive failures, rather than just reacting to single events, is key to true power system resilience.
  • Integrated Constraints for Prevention: Effectively preventing cascading failures requires considering the interplay of generator ramping, transmission limits, and the timing of potential future faults.
  • Adaptive VO for Faster Frequency Response: Dynamically adjusting virtual inertia through adaptive VO control offers a significant advantage over static methods, leading to quicker frequency stabilization.
  • VO Outperforms Droop Control: The simulations clearly demonstrate that the proposed VO strategy achieves faster frequency recovery and better stability compared to traditional droop control, especially in scenarios with reduced inertia.

Key Results

  • Accurate Fault Prediction: The preventive control method accurately predicted the timing of line faults, with a maximum error not exceeding 3 minutes compared to actual occurrences.
  • Faster Frequency Stabilization: The virtual oscillator control (VOC) strategy stabilized frequency within 2 seconds, whereas traditional droop control showed continued fluctuations beyond 20 seconds.
  • Improved Stability: The adaptive VOC strategy demonstrated superior effectiveness in frequency stability and control, effectively avoiding overshooting and achieving a more stable steady-state frequency.
  • Reduced Power Flow Oscillations: The preventive control method successfully reduced power flow in transmission lines connecting separated regions, mitigating severe oscillations during faults.

Limitations and Future Work

  • Challenges with Major Generator Tripping: The authors acknowledge that the proposed method might struggle in extreme scenarios like the tripping of a major generating station.
  • Future Work on Adaptive VOC: Future research could focus on developing more advanced predictive models and control strategies to enhance the adaptability of VOC, especially for real-time monitoring and automatic adjustment in response to potential instability signals.
  • Machine Learning Integration: Integrating machine learning is suggested as a way to design control systems that can learn from past events and adapt to new scenarios more effectively.

Summary updated

BibTeX
@misc{li2024multi,
  author = {Chenghao Li and Di Zhang and Ji Han and Chunsun Tian and Longjie Xie and Chenxia Wang and Zhou Fang and Li Li and Guanyu Zhang},
  title = {{A Multi-Level Operation Method for Improving the Resilience of Power Systems under Extreme Weather through Preventive Control and a Virtual Oscillator}},
  year = {2024},
  doi = {10.3390/s24061812},
}
Figure