This paper demonstrates an application case for the concept of Data Service design and composition techniques established by the Big-Data Data Service (BD-DS) modelling method realized using the ADOxx meta-modelling platform. In the domain of energy-efficiency assessments of buildings and their operations, the collection of energy related data does not pose a problem anymore as the necessary infrastructure is available in a non-intrusive way, at low installation and operation costs. An issue identified relates to the realization of value-adding services based on these continuous data streams in such distributed, heterogeneous environments. In the context of the ORBEET project, a dynamic, close-to-real-time data access/composition design and exploration framework has been realized. This framework builds upon the concept of Energy Data as a Service combining different sources for a dynamic and enhanced operational rating deployed in 4 different pilot sites in Europe.