A Study on a Knowledge Graph Construction Method of Safety Reports for Process Industries

This paper presents a framework for constructing industrial safety knowledge graphs from HAZOP and PHA reports. It combines asset management shell concepts with NLP-based entity extraction to create a standardized knowledge structure applicable across different process industry domains. For MOC, knowledge graphs are transformative because they capture the interconnected relationships between equipment, hazards, safeguards, and operating conditions. When a change is proposed, a knowledge graph can automatically trace impact propagation—identifying which downstream equipment, safety systems, operating procedures, and regulatory requirements may be affected. This enables more comprehensive and systematic MOC reviews than traditional manual checklist approaches.