The shift to decentralized energy systems complicates energy scheduling optimization and interoperability. It requires dynamically accounting for diverse power plants and exogenous constraints while ensuring seamless integration. KEMASS (Knowledge-Enabled Multi-Agent System for Energy Scheduling Support) addresses these challenges by using knowledge graphs and ontologies to dynamically configure multi-agent systems, enabling seamless data integration, automated system setup, and cost-effective optimization. Applied within EDF in a testing environment, it shows flexible management, scalability, and interoperability, reducing human efforts in adaptation to traditional systems. The multilevel modular ontology supports interoperability among multiple systems that enables the potential of ontology-driven system configuration for adaptive integrated energy management.