Development of process safety knowledge graph: A Case study on delayed coking process

Abstract

Process safety is one of the essential preconditions for the achievement of green manufacturing. The improvement of process safety management requires a comprehensive risk analysis based on the collection of almost all safety related information, which are usually unstructured knowledge and experience. To handle the information and support the risk analysis, a process safety knowledge graph is prompted and the development of domain ontology on delayed coking process is elaborated. The combined top-down and bottom-up approaches are used in defining the process safety schema on ontology level. Several multi-structured data sources are introduced in establishing the process safety knowledge, in which the hazard and operability analysis (HAZOP) reports and process diagrams are most important. The ontology design and data extraction are demonstrated in the manuscript while various related applications are discussed. This process safety knowledge graph might empower the knowledge-based analysis abilities in discovering the hidden relationships between possible risk causes and consequences in an emergency situation, and could provide a foundation for more application related to process safety.