Building ontologies for 4-5GDHC: A critical evaluation and modeling experiments on building-side components

This research addresses the critical challenge of digital integration and exchange of data and information from the building side toward 4-5th generation district heating and cooling (4-5GDHC) systems, where heterogeneous data and information from distributed components hinders integration and deployment of data-driven services at scale. The study conducts a critical evaluation of six major ontologies (Brick Schema, RealEstateCore, Project Haystack, SAREF, Flow Systems Ontology, and ASHRAE Standard 223P) and performs semantic modeling experiments on key building-side components including buildings in thermal networks, thermal energy storages, heat pumps, photovoltaic-thermal systems, and waste heat recovery systems. The analysis reveals significant gaps in current ontologies for representing district-level interactions, bidirectional energy flows, and thermal storage dynamics. While existing frameworks effectively model basic building components and sensors, they lack DHC-specific terminology and cannot adequately represent prosumer relationships or complex system topologies. The paper positions ontology-based semantic models as one layer of a broader digital information infrastructure and explores how they can interface with large language models (LLMs) to streamline information interaction across building and district energy systems. This work contributes to three key advances: a comprehensive critical evaluation of existing ontologies for DHC applications, practical semantic modeling experiments demonstrating real-world applicability and limitations, and forward-looking integration frameworks combining knowledge graphs with LLMs and design metadata. The findings highlight the need for DHC-specific ontology extensions and multi-ontology integration to address the unique challenges of 4-5GDHC systems. By bridging semantic technologies and AI. • Critical evaluation of six ontologies applied to building-side components in district heating and cooling systems. • Semantic modeling experiments confirm feasibility but reveal coverage gaps across multiple district heating and cooling systems. • Introduces a modular knowledge graph framework and integration with large language models for enhanced query and reasoning capabilities.