Smart electric thermal storage heating devices can be used for demand response, congestion management and incorporated within unit commitment and dispatch of generation resources for more efficient control of power systems. This paper presents an experiment-based thermal modelling approach of residential buildings which will be scaled to a national aggregate level to be used in overall power system modelling involving smart electric thermal storages. A simplified thermal network based on electrical RC-circuit analogy was developed to replicate building’s thermal dynamics and model residential heat demand at national scale. To obtain the equivalent parameters for the RC model, physical experiments were conducted during which buildings were let to cool down and then heated for several times and indoor and outdoor temperature, heat consumption and solar radiation was recorded. The identified model exhibits a good performance which improves when solar gains are considered within it. Different control strategies of the heating equipment were examined and the hourly heat demand over the year was estimated.