Dynamic management of schedulable household assets for solar self-consumption maximization with demand side management

Sammanfattning: A crucial challenge introduced by the decentralized installations of photovoltaic (PV) systems in the residential sector, is the mismatch between PV electricity generation and the load curve for energy consumption. To overcome this incompatibility between production and consumption, energy storage and demand response are seen as effective solutions. Smart meters and the installation of intelligent smart appliances in homes have paved the way for efficient energy consumption monitoring and active household load control in the residential sector.  The aim of the thesis work is to develop a dynamic energy management algorithm tailored to optimize the energy consumption pattern of controllable household assets to maximize PV selfconsumption. A rolling horizon algorithm based dynamic model was designed using mixedinteger linear programming (MILP) and later compared with the baseline model to understand the real-time operational benefits of the rolling horizon approach.  Analyzing device scheduling patterns based on the feed-in-tariff showed considerable differences in the scheduling approach for both optimization models. A comparative analysis was conducted to understand the system benefits offered by both optimization models under different feed-in-tariff structures. Higher self-consumption rates were achieved through annual scheduling approach, but it does not reflect the real-time operation of the systems in the household. A rolling horizon optimization reflects the real-time operation of the energy system and has a lower self-consumption rate due to a limited optimization horizon. The method indicates the significant potential of self-consumption specially in lieu of decreasing feed-in tariffs.