More and more grid codes require grid-forming (GFM) capabilities from grid-side converters which are needed for the integration of wind and PV power plants as well as battery storage systems. When focusing on the core functionalities—excluding hardware-related extensions of overcurrent capability or additional DC-side energy storage—these requirements can largely be fulfilled through control software alone. From an external perspective, it is therefore not directly observable whether a converter operates in grid-following (GFL) or grid-forming (GFM) mode.

The objective is to develop testing procedures that can be performed in laboratory environments and that verify the converter’s performance using assessment methods based solely on externally measurable quantities. A suitable approach is to apply defined disturbances to the grid voltage and evaluate the resulting time-domain responses of the three-phase voltages and currents at the point of connection. These measurements are readily accessible and do not require knowledge of the internal structure of the power electronics or the implemented control strategy.

The key distinction between GFL and GFM operation becomes evident within the first milliseconds following a disturbance. During this transient period, GFM converters actively contribute to stabilizing the voltage at the point of connection, whereas GFL converters primarily follow the grid conditions. To capture this performance, a space-vector-based assessment method is introduced, comparing the measured response to that of an ideal voltage source. In addition, the limitations of conventional RMS-based evaluation methods are analysed and discussed.

This work presents measurement results obtained from Power Hardware-in-the-Loop (PHIL) laboratory tests and compares them with previously conducted electromagnetic transient (EMT) simulations. Alongside a discussion of the most relevant influencing factors, the paper also proposes an analytical evaluation approach that requires only minimal information from the converter manufacturer.

Furthermore, not only self-developed white-box control models are investigated, but also commercially available household PV inverters are tested under laboratory conditions. In addition to established 10 kVA grid-following converters, a prototype GFM converter—implemented solely through a software modification—is evaluated.

The results demonstrate that a clear distinction between GFL and GFM performance can be achieved within the first milliseconds of a disturbance. The proposed space-vector-based assessment method proves to be robust and enables reliable classification with minimal modelling effort, making it a promising candidate for testing procedures in future grid code compliance assessments.