Knowledge management and project performance: a quasi-experimental study in EPC oil and gas projects

Purpose This study addresses two interconnected gaps in project-based knowledge management (KM) research: the absence of prescriptive, stage-specific intervention frameworks for Engineering, Procurement and Construction (EPC) projects, and the lack of causal evidence linking digital KM interventions to objective operational outcomes. Design/methodology/approach An integrated practical framework for knowledge transfer in EPC projects is developed by embedding Szulanski's knowledge transfer model within the EPC disciplinary structure; its Stage 2–4 mechanisms are validated through a quasi-experimental Difference-in-Differences design using longitudinal administrative data from 12 welders over eight weeks in an international oil and gas EPC project in Iraq. Findings The framework's prescribed mechanisms produce a 4.4% increase in production efficiency, a 21% reduction in defect rates and a 17% decrease in safety incidents simultaneously (p < 0.001). These findings are consistent with infrastructural barriers playing a substantial role in fragmented EPC environments, and all mechanisms are operationalisable through an easy-to-deploy, cloud-based platform requiring no specialist development. Originality/value First, the study develops an integrated practical framework that transforms Szulanski's descriptive typology into a prescriptive design tool specifying stage-specific and discipline-specific knowledge management system mechanisms. Second, it provides causal validation of a KM framework using objective operational data, an evidentiary standard that perception-based studies cannot achieve. Third, it demonstrates that the framework's prescribed mechanisms are fully operationalisable through an easy-to-deploy, cloud-based platform requiring no specialist development, closing the gap between framework validity and framework adoptability that has historically limited the practical impact of KM research.