LCL Resonance Analysis and Damping in Single-Loop Grid-Forming Wind Turbines
Meng Chen, Yufei Xi, Frede Blaabjerg, Lin Cheng, Ioannis Lestas
TL;DR
This work shows that LCL resonance can dominate the stability of single-loop grid-forming (SL-GFM) wind turbines, even when the resonance frequency is high, by revealing strong coupling between reactive-power control and high-frequency modes. A detailed small-signal model with high-frequency dynamics demonstrates that reactive-power (RAP) control significantly shapes the LCL-resonant modes and can render the RAP loop nonminimum phase, especially in weak grids. The authors compare multiple $q-V$ RAP strategies, identify limitations of RAP-only or pure-droop approaches, and design a capacitor-voltage feedback active-damping (AD) technique to suppress resonance without adding sensors. Validation on a 14-bus, 5-machine IEEE system shows instability can originate from LCL resonance and that the proposed AD robustly mitigates high-frequency oscillations, highlighting key differences between SL-GFM and grid-following (GFL) control in resonance behavior and damping needs.
Abstract
A dynamic phenomenon known as LCL resonance is often neglected when stability analysis is carried out for grid-forming (GFM) control schemes by wind turbine systems, due to its high frequency. This paper shows that this simplification is not always valid for single-loop (SL) control schemes. A detailed small-signal analysis reveals that reactive power (RAP) control significantly influences the resonant modes, which may be dominant in determining overall system stability, even if the resonant frequency is high. The underlying mechanism via which the LCL resonance may dominate the overall system stability is systematically analyzed. Furthermore, various RAP control strategies are compared to assess their different effects on resonant modes. An active damping (AD) strategy favorable for SL-GFM control is then designed. We also provide a comparison between SL-GFM and well-studied grid-following control schemes, highlighting quite different resonance features between them. Finally, case studies associated with a 14-bus, 5-machine IEEE test system are presented. These show that instability originates from the LCL resonance rather than low-frequency interactions among multiple machines, validating the theoretical analysis and the proposed AD strategy.
