Universal Input-Output Model of GFM Functions and Data-Driven Verification Methods: UNIFI Consortium

An overarching goal of the UNiversal Interoperability for grid-Forming Inverters (UNIFI) Consortium is to develop vendor agnostic specifications and guidelines that ensure interoperability of gridforming (GFM) inverter-based resources (IBRs) without requiring vendors or system operators to reveal proprietary information. These UNIFI principles and specifications are envisioned to apply to a wide range of technologies and systems. The initial work on the UNIFI principles and specifications has focused on performance requirements and high-level aspirational principles outlined in [1]. Going forward, a key question is how to translate such high-level requirements and principles into rigorous specifications that can be enforced and validated for a wide range of IBRs. A wide range of different GFM controls are available in the literature that use vastly different internal dynamics and control parameters (e.g., droop coefficients, virtual inertia constants, virtual oscillator gains). However, from a system-level viewpoint, these GFM controls implement a few universal GFM functions to various degrees of fidelity. This report summarizes work on universal GFM input-output dynamics (i.e., the dynamic response observable on the IBR terminals) through a universal reduced-order input-output model that is parameterized in generalized system-level control parameters (i.e., parameters that translate to different gains for specific GFM controls). Specifically, the input-output models presented in this work are suitable to capture the small-signal response under nominal operating conditions. After presenting the modeling approach and several examples of how standard GFM controls map to the universal input-output model, we present a first attempt to map UNIFI performance requirements for operation within normal grid conditions [1] to specifications on the input-output dynamics. We emphasize that the main purpose of the input-output models is not to develop models for simulation but rather to obtain low-dimensional models of the IBR terminal behavior that can be used to formulate and validate vendor and technology-agnostic unit-level specifications that ensure stability, interoperability, and performance. To this end, a crucial question is how to verify if a GFM IBR conforms to the universal input-output dynamics with given generalized system-level control parameters without detailed information of the internal hardware and controls. To this end, we discuss two methods for identifying the input-output (i.e., terminal) dynamic behavior of GFM converters from input-output data that can be obtained using hardware experiments or blackbox models. The results on input-output modeling form the basis for developing stability certificates that can be directly verified through (experimental) input-output data. Moreover, we envision using the input-output modeling framework to develop and validate specifications that narrow down the class of interoperable GFM dynamics to obtain bounds on the unit-level dynamics within which vendors can innovate while interoperability and stability are ensured.