Converter-driven stability has generally been accepted as a new category of power system stability. This paradigm shift has occurred due to observations of real-world power system oscillations linked to areas of large penetration of inverter-based resources (IBRs). It is anticipated that methods for analyzing converter-driven stability will see broader adoption by system operators, plant developers, and device manufacturers, driven by the growing share of IBRs in the power system.

Among the available methods for analyzing converter-driven stability, impedance- or admittance-based approaches using frequency-response data have proven particularly useful in practice, owing to several advantages:

1. Protection of OEM intellectual property, as frequency-response data conceals control structures and parameter tuning details.
2. Compatibility with system-wide stability assessments, particularly through integration with traditional Y-bus methods.
3. The ability to be empirically extracted from both electromagnetic transient (EMT) simulations and physical measurements.

The validation of empirically extracted converter admittance matrix data from real measurements against model-based data is the central focus of this paper. While numerous academic and industry sources present comparisons between analytically derived and empirically obtained converter impedance/admittance matrices—using data from EMT simulations or real measurements—such comparisons are typically qualitative in nature. Descriptors like “good accuracy,” “good alignment,” or “reasonable match” are often used, without a clearly defined quantitative threshold. However, the lack of a standardized, objective error metric limits the ability to rigorously assess model fidelity. This becomes especially critical in stability studies, where reliance on an inaccurate model may lead to misleading conclusions regarding system robustness or the risk of instability. Therefore, a quantitative validation criterion is necessary to support confident decision-making by system operators, plant developers, and equipment manufacturers.

This paper proposes a relative error metric based on the norm of the residual matrix to quantify the accuracy of converter admittance validation—whether between models or between model and measurement. This metric enables the definition of acceptance criteria, such as requiring a relative error below 10%. Its suitability is supported through analytical derivation, demonstrating invariance to reference frame choice and linking it to the stability characteristics of the converter–grid system. The procedure for extracting converter admittance from real-world measurements is also presented in detail.

The proposed relative error metric and admittance validation process are applied to measurements from a Siemens Gamesa Renewable Energy (SGRE) wind turbine converter connected to SGRE’s Grid Converter Test Rig (GCTR). The converter admittance extracted from measurement is validated against the corresponding data from electromagnetic transient (EMT) simulations and an analytical model over the 0.2–1000 Hz frequency range. Results demonstrate that relative errors below 10% are achievable across multiple operating points. The validation methodology and experimental findings offer practical guidance for system operators, plant developers, and device manufacturers in evaluating the accuracy of converter stability models.