The decarbonization of the buildings’ sector implies advanced control strategies for the heating, ventilation, and air-conditioning systems, as well as combining conventional and renewable energy technologies. This paper presents the conceptual foundations for implementing a digital twin-enabled control architecture for a hybrid HVAC system integrating a methane-fueled boiler and an air-source heat pump. The approach integrates physics-based models, an experimental-scale infrastructure with real-time data acquisition with predictive optimization techniques to enable energy-efficient system operation. A modular digital twin architecture is currently under development, providing capabilities for real-time state estimation, predictive simulation, and informed decision-making under dynamic weather conditions, tariff structures, and energy demand profiles. The proposed framework is scalable and applicable to smart buildings and different energy communities, bridging the transition toward intelligent, reliable and sustainable energy systems.