Use of Modeling as an Enabler for Cross-Topic Knowledge Management and Ontologies to Support Return of Experience and Replicability of Large Nuclear Projects

Abstract

Nuclear projects often produce and consume a large amount of knowledge. Capitalization on this knowledge constitutes a significant way to increase efficiency on subsequent projects for any stakeholder. In this study, modeling is used as a main approach to support this capitalization. It constitutes, through graphical layout, a more reliable and robust way to transfer information. Moreover, the use of an interconnected set of models enables organizations to break the silos between the disciplines. The approach proposed is based on the operational implementation of existing “on-the-shelf” elements to benefit from previous implementations. The presented example illustrates how, on a nuclear project, engineering processes have been modeled from knowledge of previous projects. These components are all interconnected to constitute a self-supporting set of models as a body of knowledge. The use of models as a main base for knowledge management and transfer has been initiated due to (i) increased traceability of design decisions and impact analysis required by ever-increasing safety requirements from 1980s, 1990s, and Fukushima accident for new nuclear and (ii) gradual departures of design teams and lifetime extension of existing power plants requiring efficient ways to pass down knowledge to newcomers. Additionally, transfer of knowledge from document-centric to model-centric has enabled securitization of interfaces (between organizations, between systems) due to more numerous stakeholders and reduction of rework and site works.