Towards intelligent simulation platforms for scalable validation of digital energy systems

Power systems are undergoing rapid digitalization, with distributed software applications emerging to optimize and manage grid operation. These include virtual power plants, coordinated electric vehicle charging, and distributed flexibility services. Their main challenge lies in their large scale and complex interactions across software, communication networks, and physical grid models. Validating such systems requires simulation environments capable of handling thousands of interacting components in a reproducible and automated manner. To address this need, this paper presents a modular Kubernetes-based framework for automated real-time emulation of large-scale cyber-physical energy systems. A unified system description is used to automatically generate grid models, communication interfaces, and application deployments. This enables scalable and reproducible experiments using container orchestration technologies. The framework is demonstrated on benchmark distribution networks, scaling seamlessly from a 13-bus low-voltage system to a 134-bus medium-voltage system.