ABSTRACT Increasing inverter penetration shifts frequency stability in low‐inertia power systems from an inertia‐dominated to a transient‐dominated regime governed by converter control dynamics. While grid‐forming (GFM) inverters are widely regarded as essential for fast frequency support, the dynamic limitations and shaping potential of grid‐following (GFL) control architectures remain insufficiently understood. This paper presents a systematic experimental comparison of inverter‐based frequency‐support concepts based on their transient frequency–power dynamics. GFL droop control, GFM droop control, and virtual synchronous machine (VSM) control are implemented on identical inverter hardware using a common inner control structure. The dynamic mapping from frequency deviation to active power injection is characterized through laboratory experiments and low‐order system identification. An ‐based matching controller is further introduced to shape the closed‐loop frequency–power dynamics of a GFL inverter towards GFM behaviour. The results demonstrate that classical GFL droop control exhibits delayed power responses due to measurement‐based synchronization and cascaded tracking loops, leading to deeper frequency nadirs in weak grids. GFM strategies enable immediate power injection and explicit damping, with inertia and damping acting as complementary tuning parameters. GFL inverters equipped with matching control are shown to reproduce key GFM frequency–power characteristics, including rapid initial power injection and a substantially improved frequency nadir. These findings indicate that fast frequency support is governed by the closed‐loop frequency–power dynamics of the control structure rather than by GFM or GFL classification. A strict GFL/GFM paradigm dichotomy is therefore insufficient to assess frequency‐support capability.
Building similarity graph...
Analyzing shared references across papers
Loading...
Alexander Och
Irina Zettl
Alexander Winkens
IET Generation Transmission & Distribution
RWTH Aachen University
Building similarity graph...
Analyzing shared references across papers
Loading...
Och et al. (Thu,) studied this question.
www.synapsesocial.com/papers/69cd7b065652765b073a8b36 — DOI: https://doi.org/10.1049/gtd2.70285