In the last decade, Knowledge Graphs (KGs) have increased their use as a tool to represent knowledge and query data. Some industrial solutions use KGs to allow the integration between the systems, to use artificial intelligence, or to facilitate communication between humans and the systems. DANKE is a platform developed by Tecgraf Institute at PUC-Rio University to support key searches over preprocessed and integrated data. It was configured and used for searching over data related to real-world operation and maintenance activities of an Oil and Gas production scenario. The Knowledge Graph of this system, called DOnG-KG (DANKE Oil and Gas - Knowledge Graph), is its core and allows searching and linking information from concepts collected from diverse data sources. However, it was developed under agile methodology, sacrificing some modeling decisions to empower performance. Moreover, it does not consider existing data models and core ontologies like ISO15926-part 14 for life-cycle data for process plants to guide the graph construction. This work proposed a refactoring of the DOnG-KG using well-known ontologies like SKOS, DC, OntoCape, ISO15926 part 2 (Reference Data Library), and part 14 (Upper Ontology) to define some concepts. We also extend the graph, including more concepts previously treated as data properties. In that way, the graph was adequate to an international semantics standard for life-cycle data for process plants, including Oil and Gas production facilities, allowing the inclusion and linking of new terms in the graph and different standards-subgraph for specific subdomains without inconsistency and ambiguity. Finally, the KG resulting from this work could be extended by plugging new sources and even used for other applications in the Oil and Gas context.