The rapid increase of inverter-based resources (IBRs) in modern power systems is fundamentally changing system dynamics and introducing new stability challenges. Unlike synchronous generators, IBRs are governed by fast control dynamics, leading to complex interactions both among converters and with the network, which can result in poorly damped oscillations and emerging instability mechanisms. These challenges are especially important in hybrid power plants (HPPs), where grid-following (GFL) and grid-forming (GFM) inverters coexist and interact through different synchronization and control principles.

Existing literature has proposed multiple modeling and analysis methods to investigate converter-driven stability phenomena. Electromagnetic transient (EMT) models provide a detailed representation of converter dynamics. Moreover, they are often used as a high-fidelity reference, but are computationally demanding and less practical for studying many operating conditions. Linearized state-space models are more convenient for stability analysis because they reveal eigenvalues and oscillatory modes, but they are only valid near a chosen operating point. Nonlinear continuous-time models provide an intermediate option, retaining more of the original system dynamics than linearized models while being less expensive than full EMT simulation. These models are obtained from the EMT model by averaging the switching behavior over a suitable time period.

This paper examines how these modelling choices affect the stability conclusions drawn for a hybrid inverter-based power plant. The study considers three representations of the same system: a PSCAD EMT model used as a high-fidelity benchmark, a nonlinear continuous-time ordinary-differential-equation model derived from the EMT model, and a family of linearized models obtained around multiple operating points. To this end, we will begin by considering both GFL and GFM inverters connected to a power network, so that the influence of operating point and disturbance type on the consistency of the models can be assessed. This study forms the basis for HPP level modeling and analyses.

The objective is to systematically compare the dynamic responses and stability predictions obtained from these models under varying operating conditions and disturbances. Attention is given to identifying operating regions where the models provide consistent results, as well as conditions under which significant mismatches arise. The relevance of these mismatches is further assessed in terms of their potential impact on stability conclusions, including the risk of mischaracterizing system behavior.

The results provide insight into the applicability and limitations of commonly used modeling approaches for HPPs connected to the power network. This work contributes to improving the reliability of stability studies by clarifying when simplified representations are sufficient and when high-fidelity modeling is required.