EUI3: Application of self-learning techniques for energy systems integration in future residential heating systems

Decarbonisation of the residential thermal sector is a key pathway to meeting EU targets on carbon emissions. Future large-scale electrification of residential heating, beyond its present level, in tandem with the expansion of renewables on the electricity grid, provides one possible pathway to decarbonisation. Switching from fossil fuels to electricity is technically very challenging without a significant investment in the electricity generation and distribution network. In order to facilitate the switch, the residential sector will need to reduce its thermal demand by minimising losses, increasing system efficiencies, in tandem with increased demand flexibility.

 

​Energy integration of building energy conversion systems is one promising way to address these challenges. Buildings contain many energy systems (such as heating, ventilation, etc.) that provide a necessary service to meet occupant comfort expectations and health requirements. However, the different systems are often operated independently and in a sub-optimal way. Machine learning algorithms integrated into building controls systems, coupled with relevant environmental sensing, can be used to unify the monitoring and control of the separate systems. Moreover, actions identifying wasteful energy behaviour can be identified and with targeted efforts, be corrected. This holistic approach will allow the operation of the systems to be optimised with respect to reducing energy consumption and maximising occupant comfort and health. The use of advanced control systems will also facilitate the integration and operation of new technologies that are expected to provide increased levels of service to the power system. For example, electric vehicles, thermal energy storage and renewable energy systems (e.g., photovoltaics), more efficient heating systems (e.g., heat pump systems) combined with ICT and cloud technology, will enable demand response management measures to shift peak electricity loads to times of high renewable energy production or low system demand.

 

Members


Dr Donal Finn
Funded Investigator, Associate Professor, School of Mechanical & Materials Engineering, University College Dublin
donal.finn@ucd.ie
01 716 1947
Dr Eleni Mangina
Funded Investigator, Associate Professor, School of Computer Science, UCD
eleni.mangina@ucd.ie
+353 1 7162858
Dr Mohammad Saffari
Postdoctoral Researcher
mohammad.saffari@ucd.ie
Adamantios Bampoulas
PhD Researcher
adamantios.bampoulas@ucdconnect.ie