Challenges in the detection, diagnosis and visualization controller performance data

Process control performance is a cornerstone of operational excellence in the refining, petrochemicals, pulp and paper and the mineral processing industry. Control performance assessment and monitoring applications have become mainstream in these industries and are changing the maintenance methodology surrounding control assets from predictive to condition based. The large numbers of these assets on most sites compared to the number of maintenance and control personnel has made monitoring and diagnosing control problems challenging. Identifying specific control issues on a plant-wide basis and their root-causes is analogous to looking for a needle in a haystack. Tree mapping, a data visualization technology developed in 1990 is proposed as one useful solution to this problem and is demonstrated to work very effectively within a condition based maintenance framework. This paper would explore the challenges associated with control performance assessment and monitoring and show how tree mapping technology eases the visualization challenge associated with monitoring hundreds or thousands of control assets. Both regulatory and advanced process control examples would be demonstrated.