Nuclear-Powered Hybrid Energy System for Clean Hydrogen Production: Time-Step-Optimized Real-Time Multi-Domain Hardware Emulation

Increasing global emphasis on decarbonization and the proliferation of renewable energy, energy storage, and nuclear power is driving a surge of research interest into integrated sustainable energy modeling, simulation, operation and control. Traditional electromagnetic transient (EMT) methods typically discretize electrical networks using the trapezoidal rule. When coupled with the ordinary differential equations (ODEs) of other physical domains, however, the absolute stability region of the numerical integration scheme can shift, and discrepancies in time constants across subsystems further complicate integration of disparate models. While the demand for integrating EMT with multi-domain co-simulations is increasing, existing commercial EMT simulation tools either lack support for multi-domain physical coupling or are not specifically optimized for such hybrid simulations. To address this gap, this paper proposes a robust multiscale time-step estimation (RMTE) framework that enables real-time co-simulation of EMT networks and multi-domain subsystems. The framework includes a fast, efficient, and adaptive approach for selecting the optimal maximum time-step across heterogeneous physical domains. The proposed method is validated through a case study involving small modular reactors (SMRs), wind farm, photovoltaics (PV) and low-temperature proton exchange membrane (PEM) electrolysis for clean hydrogen production. Real-time hardware co-emulation is achieved on a field-programmable gate array (FPGA)-based platform. The results demonstrate significant improvements in simulation efficiency and execution time.