The energy transition is driving research on the topic of large-scale hydrogen electrolysis to support hard-to-abate industrial and transportation sectors such as steelmaking, chemical production (ammonia and derivatives), and e-fuels.

In recent years, an increasing number of hydrogen developers have required detailed power system studies to support electrical design optimization, assess technical feasibility, and demonstrate compliance with applicable grid codes. These studies typically start with load flow, short-circuit, and harmonic analyses. To validate the resulting passive and active filtering solutions, dynamic studies are performed at a final stage to confirm equipment ratings, ensure system stability, and verify compliance with dynamic grid code requirements.

In order to perform these studies, accurate dynamic models of hydrogen rectifier systems are needed. These models become even more critical as rising grid connection costs (i.e., grid tariffs) encourage partial-grid or fully off-grid hydrogen configurations coupled with Renewable Energy Sources (RES), which often result in weak grid conditions and increased stability risks.

Despite its relevance, the dynamic modelling of hydrogen rectifiers is not comprehensively addressed in the literature. This paper presents a detailed electromagnetic transient (EMT) model of a 20 MW alkaline electrolyzer supplied through a multi-winding transformer and a thyristor-based rectifier. The EMT model, developed in a MATLAB-based simulation environment, includes a detailed representation of the transformer topology, thyristor bridge, electrolyzer electrical characteristics, control systems—such as firing angle and on-load tap changer control—and relevant protection functions.

Based on the EMT reference model, a set of equations suitable for Root Mean Square (RMS) dynamic simulation tools is proposed, with a specific implementation in PSSE. Particular attention is given to the limitations of conventional thyristor bridge models commonly applied in LCC-HVDC studies when used for electrolyzer-fed rectifiers. It is demonstrated that standard assumptions related to commutation processes and DC-side behavior lead to significant inaccuracies in representing the dynamic response of electrochemical loads.

A systematic comparison between the EMT model and the proposed RMS model is presented, considering steady-state operation, and dynamic response. Results show that the proposed RMS modelling approach significantly improves accuracy compared to traditional thyristor bridge representations, while remaining computationally efficient for large-scale power system studies.

This work results from a collaboration between Hitachi Energy Power Consulting, a hydrogen project developer, and a rectifier manufacturer, and supports more reliable grid integration studies for large-scale electrolyzer installations connected to renewable-dominated power systems.