An agent-based framework for smart grid balancing exploiting thermal flexibility of residential buildings

The Smart Grid is a complex system that encompasses many different fields of expertise. As a consequence, co-simulation tools are emerging as a possible solution to test future scenarios and strategies thanks to their ability to reuse domain-specific simulators in a broader context. Therefore, we propose an agent-based co-simulation framework able to act as a test-bed for multiple Smart Grid strategies. In particular, we tested demand response programmes that exploit the thermal behaviour of residential buildings at the district level. The proposed framework is modular, thus it eases further integration of new modules. Moreover, it is highly flexible thanks to the numerous configuration parameters that allow creating a realistic scenario. The system has been tested over 1000 buildings in a district and an analysis of the effects on the balancing at primary substation, due to micro-deviation from scheduled temperature set-points in the building’s premises, is proposed. Results demonstrate that power imbalances can be mitigated already with minor set-points deviations.