Adaptive Droop Gain Control for Optimal Kinetic Energy Extraction From Wind Turbines to Support System Frequency

Low-inertia power grids would benefit from frequency support from inverter-based generation, given the scarcity of synchronous generators as decarbonisation efforts progress. Wind turbines store kinetic energy in their rotors as they produce electric power, which can be partially released to the grid in the event of a generation loss, in order to contain the frequency drop. This paper introduces a novel fast droop control for wind turbines (WTs) based on an adaptive droop gain strategy, extracting the optimal amount of kinetic energy depending on the overall system conditions. This requires mapping the low-level inverter controls into a system-wide economic optimisation, which has not been successfully addressed yet. To overcome this challenge, we propose a data-driven methodology for integrating frequency stability constraints into an Optimal Power Flow (OPF) formulation, which is explicitly dependent on the low-level control parameters of the wind turbines’ converters. A modified Optimal Classification Tree (OCT) is used to encode frequency-security guarantees for the grid, due to its suitable structure for being included in optimisation while maintaining tractability. Through relevant case studies, we demonstrate the effectiveness of the proposed system-aware control for WTs, achieving over 8% system cost savings compared to a system-unaware controller.