Large-Signal Stability of Power Systems with Mixtures of GFL, GFM and GSP Inverters

Yifan Zhang, Yaoxin Wang, Yunjie Gu, Yitong Li, Sijia Geng, Yue Zhu · 2025

Read Large-Signal Stability of Power Systems with Mixtures of GFL, GFM and GSP Inverters by Yifan Zhang, Yaoxin Wang, Yunjie Gu, Yitong Li, Sijia Geng, Yue Zhu (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-05994.txt.

Introduction

Yifan Zhang and co-authors study stability in power grids crowded with inverters. Their 2025 paper mixes three inverter kinds: grid-following, grid-forming, and grid-supporting. These devices act nothing like spinning generators under big disturbances. Classic tools like the equal-area rule do not carry over.

The team derives a shared model for two-inverter systems in all mixes. They prove that systems with grid-following units admit no global energy function, blocking classic direct methods. A manifold method instead maps the true safe region. Simpler reduced models, drawn from scale separation, keep the math affordable. A stability radius then measures how much shock the grid can take.

Catalog record

CREATORYifan Zhang, Yaoxin Wang, Yunjie Gu, Yitong Li, Sijia Geng, Yue Zhu
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE53 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2501-05994

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Large-Signal Stability of Power Systems with Mixtures of GFL, GFM and GSP Inverters Thanks: This work was supported by the Royal Society under the award ICA\R2\242103 and the Ralph O’Connor Sustainable Energy Institute at Johns Hopkins University. Yifan Zhang, Yaoxin Wang, Yunjie Gu, and Timothy C. Green are with Imperial College, London, UK, Yitong Li is with Xi’an Jiaotong University, Xi’an, China, Sijia Geng is with Johns Hopkins University, Baltimore, MD, USA, Yue Zhu is with City University of Hong Kong, Hong Kong, China, and Hsiao-Dong Chiang is with Cornell University, Ithaca, NY, USA. (e-mails: [email protected] ; [email protected] ; [email protected] ; [email protected] ; [email protected] ; [email protected] ; [email protected] ; [email protected] . Corresponding author: Yunjie Gu.) Yifan Zhang, Yaoxin Wang, Yunjie Gu, Yitong Li, Sijia Geng, Yue Zhu Abstract Grid-following (GFL) inverters have very different large-signal stability characteristics to synchronous generators, and convenient concepts such as the equal-area criterion and global energy function do not apply in the same way. Existing studies mainly focus on the synchronization stability of an individual GFL inverter, and interactions between multiple inverters are less often addressed. This paper elucidates the interaction mechanisms of inverters with heterogeneous control, covering GFL, grid-forming (GFM), and grid-supporting (GSP) inverters, to determine the stability boundaries of a system with mixtures of inverters. The generalized model for large-signal dynamics of two-inverter systems with various inverter combinations are derived. This establishes that no global energy function exists for systems containing heterogeneous inverters, implying that traditional direct methods of stability assessment cannot be applied to such systems. To overcome this barrier, a manifold method is employed to accurately determine the region of attraction (ROA) of such systems. To address the computational complexity of the manifold method, reduced-order models of inverter are used based on multiscale analysis. The large-signal stability margin is assessed by the shortest distance from a stable equilibrium point (SEP) to the boundary of the ROA, which is called the stability radius (SR). The manifold analysis, verified by EMT simulation, demonstrates that GFM and GSP inverters significantly enhance the large-signal stability of a two-inverter system where the other inverter is GFL, with GFM being slightly better than GSP. Such stability enhancement is attributed to the voltage support effects of GSP and GFM inverters. The improvement is maximized if the GFM or GSP inverter is placed at the mid-point of the transmission line, where the voltage is lowest. All findings in this paper are also validated through power hardware-in-the-loop (PHIL). † † aftertitle: I Introduction Grid-following (GFL) inverters, which rely on a phase-locked loop (PLL) for synchronization with the grid, are widely used to interface renewable energy generation with the power grid. The incidents of PLL losing synchronization when encountering a grid fault have been frequently reported in recent years [ 1 , 2 , 3 , 4 ] . While the synchronization challenges of an individual GFL inverter are well-documented, the dynamics become more complex when multiple inverters are interconnected via transmission lines. This complexity is further compounded by the presence of other types of inverters, such as grid-forming (GFM) [ 2 ] and grid-supporting (GSP) inverters [ 5 ] , whose interactions can significantly influence synchronization neighboring GFL inverters. The discussion here will be of threats to synchronization induced by large perturbations such as line faults, so the issue is termed large-signal stability and the analysis must capture the nonlinear behaviors present. The large-signal stability of traditional power systems dominated by synchronous generators (SGs) has been extensively studied. The direct method, which avoids computationally expensive time-domain integration of post-fault system dynamics, is well-established and has become one of the most widely used approaches for traditional systems [ 6 ] . The method estimates the region of attraction (ROA) of the operating point to a stable equilibrium point (SEP) through the level set of the global energy function through an unstable equilibrium point (UEP). The construction of global energy functions is a vital part of the direct method. However, [ 6 , 7 ] prove that global energy functions do not exist for lossy power systems which have a transfer conductance introduced by the losses. Because the transfer conductance is typically small, there are various approaches that can mitigate this issue, such as numerical approximations [ 6 ] , extended invariant principle method [ 8 ] , and dissipative systems theory [ 9 ] . However, in power systems dominated by inverter-based resources (IBRs), the construction of global energy functions presents a great challenge because of the differences in dynamics between IBR and SG. Although a global energy function can be constructed for a single-inverter-infinite-bus system [ 3 , 10 ] , for multi-inverter systems, GFL inverters will introduce cosine interaction terms in differential equations, as will be demonstrated in this paper. This effect is similar to the characteristics of transfer conductance in conventional power systems [ 7 ] . Unfortunately, the cosine interaction terms induced by GFL inverters are much larger than those introduced by transfer conductance in traditional SG systems, exacerbating the challenge. As a result, global energy functions generally do not exist in power systems containing GFL inverters. There are no effective means to date to mitigate this issue, which prevents the application of the direct method to IBR-dominated systems. (a) modelling of GFM inverters. (b) modelling of GFL inverters. (c) modelling of GSP inverters. Fig. 1: Large-signal modeling of different types of inverters. The famous results in [ 11 ] show that multi-machine systems can often be characterized by the interaction between two clusters. This implies that an investigation into two-inverter systems is a reasonable first step towards studying multi-inverter systems. The two-inverter large-signal stability has been studied in [ 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 ] . Among these studies, [ 12 , 19 , 18 ] perform quasi-static analysis, focusing on the existence of a SEP, but the ROA around the SEP is not assessed. Other studies transform two-inverter systems into single-inverter systems, by neglecting the dynamics of the second inverter [ 15 , 17 , 16 ] , or treating the second inverter a source disturbances [ 14 , 13 ] . These methods are based on very significant assumptions, that is, the two inverters are not tightly coupled, or the second inverter is less significant than the first inverter. To address the shortcomings above, this paper presents the manifold method to identify the ROA of two-inverter systems, based on the theory proposed by Chiang et al. [ 6 , 20 ] . Chiang’s theory indicates that the ROA of a wide class of nonlinear systems (including power systems) is the region encircled by the stable manifolds emanating from all UEPs of the system. The manifold method is not dependent on a global energy function and therefore can be applied to IBR-dominated systems. The manifold method has been applied to single-inverter systems in [ 10 ] , but obstacles are encountered in multi-inverter systems due to the difficulties in computing and visualizing the stable manifolds of high-order systems. In this paper, we perform model order reduction on inverters based on multiscale analysis and through that extend the manifold method to two-inverter systems. This extension enables

Read the complete text