Digital Twins (DTs) can offer advantages for real-time monitoring, maintenance and simulation. They can provide a platform for the facility management (FM) to identify inefficiencies and deviations from optimal operation in order to derive measures to improve key performance indicators of technical building systems. A major challenge for DTs is currently still the merging of heterogenous data sources, such as sensors for relevant building and user parameters or external services, e.g., weather forecast, in a coherent DT, as well as the possibility of interaction for FM. This requires sophisticated data processing and interoperability frameworks. Building Information Modelling (BIM) provides a structured, standardized approach to creating digital representations of physical assets. Therefore, an extended BIM framework was created for modelling and commissioning building automation systems. This is capitalizing on a configurable middleware that allows the dynamic adaptation of control logic in addition to the aggregation and provision of operating data for the DT. This study extends our current solution with an open-source framework using Unreal Engine 5 (UE5) for the DT, which serves as a service level of FM, visualizes building information, allows for real-time interaction, and enables dynamic changes to the control logic from the BIM modelling framework. The results show, that the synergy between BIM and DT and the continuous analysis of real data is essential for the continuous optimisation of energy efficiency and comfort.