Key points are not available for this paper at this time.
Renewable energy sources such as wind power have grown exponentially in recent years. Conventional grid-following wind turbines (GFL-WTs) run the risk of oscillations under weak grids. To operate stably over a wide range of grid strength, the hybrid wind farm with both grid-forming wind turbines (GFM-WTs) and GFL-WTs is an effective form. Due to the declining frequency stability of the renewable power system, it is important for wind farms to provide frequency support. However, effective frequency support control of such hybrid wind farm is a crucial issue. Hence, this paper proposes a hierarchical optimal frequency support (HOFS) scheme for the hybrid wind farm to optimize frequency support. Firstly, the system frequency response models of GFM-WTs and GFL-WTs are established for analysis and comparison. A duality consistency is revealed. In conclusion, the frequency support effects of GFM-WTs and GFL-WTs are consistent once both structures and parameters satisfy the revealed duality consistency formula. Secondly, a two-level HOFS scheme benefiting from the above analysis is designed. Level I is optimal frequency control of WT. Level II is coordination control of multiple WTs. The former level according to the duality consistency enables the GFM-WTs or GFL-WTs to optimally support the frequency, respectively. The latter level mitigates the additional frequency drop due to power limitation, particularly in high wind speed scenarios. Finally, case studies are undertaken on a two-area integration system with a wind farm. Both simulation and real-time experimental results verify the effectiveness of the HOFS scheme. • A hierarchical optimal frequency support scheme is proposed for the hybrid wind farm. • Frequency control is designed separately for frequency decline and recovery stages. • Coordination of multiple WTs enhances the frequency in high wind speed scenarios. • A duality consistency of GFM-WTs and GFL-WTs in frequency support is revealed. • The frequency response models GFM-SFR and GFL-SFR are developed for comparison.
Zong et al. (Tue,) studied this question.