Dynamic real-time control for a multi-energy prosumer with electricity and heat

Abstract

Multi-energy prosumers have emerged as a promising avenue for curtailing energy consumption by integrating diverse energy vectors in a synchronised operation. Prevailing investigations have predominantly focused on static analyses directed at optimising cost functions for power dispatch problems, often overlooking the dynamic facets of the system. This paper introduces a dynamic real-time control scheme for a multi-energy prosumer encompassing electricity and heat as energy vectors. A novel state-based energy management system (EMS) is designed, with the goal of ensuring energy balance (electrical and thermal), while prioritising the utilization of renewable energy and diminishing reliance on local electrical distribution networks. To this end, three different operating modes are defined regarding the real-time renewable capacity. The EMS and multi-energy prosumer are subjected to evaluation across several weather conditions in a 4-h variable load profile. A sensibility analysis considering 250 simulations of 1-h duration, with a wide range of irradiance, water demand, underfloor heating load, and state-of-charge conditions were used to validate the control response, demonstrating the lack of use of the local grid. Moreover, a real-time experiment employing hardware-in-the-loop testing with an OPAL-RT4512 unit and a dSPACE MicroLabBox control prototype confirms the adequate response in a practical scenario. Comparative validation against a fuzzy-logic benchmark throughout a 24-h dynamic horizon revealed that the proposed state-based EMS reduced grid dependency by 36.61% and auxiliary gas boiler by 2.23%. Furthermore, the architecture achieved a 30-fold reduction in computational time while maintaining negligible control errors, establishing its superior suitability for real-time implementation in prosumers environments.