Scalable method for optimising the energy flexibility of neighbourhoods

Gebäudetechnik

Development of a decentralised automation method to improve the flexibility options of buildings and neighbourhoods on the consumer side. The data-driven algorithms used promise high scalability and therefore low installation and operating costs. The developed method will be validated using different building types (high-tech office buildings, low-tech office buildings, residential buildings).

Acronym

scaleFLEX

Project running time

01/01/2023 - 31/12/2025

Projectbudget in EUR

> 500.000

Meeting the European climate targets requires a reorganisation of the entire energy system. In addition to measures to improve efficiency, decentralised and intelligently networked system solutions are needed to securely control flexibility options. Buildings can also be integrated into the grid via sector coupling, thereby significantly increasing their contribution to decarbonisation. Although these developments are more cost-effective than electrical storage capacities, they require the use of intelligent building automation systems to continuously adapt energy consumption to the volatile energy supply and regional energy infrastructure.

With the increasing networking of systems, existing automation concepts with centrally organised optimisation are reaching their limits. The reasons for this include the high complexity of the overall system, the limited interoperability of the subsystems, individual usage-related requirements and data protection. In addition, the automation methods currently available are dependent on information from the construction phase, which makes innovative refurbishments and extensions to existing buildings considerably more difficult. New, scalable methods for holistic load optimisation are therefore needed at building and neighbourhood level.

This project proposes a corresponding transfer of the centrally organised automation strategy into distributed regulation and control units with decentralised intelligence and decision-making powers. Grid-friendly operation is achieved through higher-level coordination based on the Stackelberg model approach. The synergetic use of data-driven, decentralised optimisation units with a higher-level reinforcement learning optimisation process facilitates scalability and interoperability. The methodology developed can therefore be applied extremely cost-effectively in new and existing buildings and neighbourhoods. This optimises the potential of existing energy infrastructures and provides sustainable support for innovative energy services and business models.


FFG - Österreichische Forschungsförderungsgesellschaft

Projectleader

Prof.(FH) DI(FH) Dr. Christian Heschl

Tel: +43 5 7705-4121
christian.heschl(at)hochschule-burgenland.at

Projectpartner/Researchpartner

client/sponsor