Abstract
This paper describes a solution based on data analysis and innovative visualization methods that aim to aid the control room operator detect faults developing in the process at an early stage. The solution also helps the operators keep an overview in disturbance situations by indicating how the disturbance propagates through the plant. By combining the result from analysis of historical process data with information on alarm and events, the operator is made aware of important changes, relationships and patterns in the process that otherwise would have been very difficult to detect. Novel visualization methods are used to highlight critical factors while at the same time allowing the user to explore and understand large amounts of information at once. A fully functional prototype has been developed and tested on various use cases, both on offline data from a set of different oil & gas facilities as well as on a training simulator at a petrochemical plant.
