As refinery operations become more integrated and data-intensive, maintaining accurate digital twins has become increasingly difficult. Model drift—driven by feed variability, equipment changes, and data quality issues—can erode optimisation performance, increase energy use, and reduce confidence in simulation-driven decisions.
This article explores how a new generation of automated model lifecycle management integrates process simulation, machine learning, and cloud computing to continuously monitor and recalibrate digital twins. By shifting from manual intervention to continuous, automated maintenance, refiners can improve planning accuracy, reduce engineering workload, and enhance operational performance at scale.
Download the full technical article (PDF)
This article originally appeared in PTQ (Petroleum Technology Quarterly) via Digital Refining.