Process and controller performance monitoring: overview with industrial applications

Abstract

Complex processes depend on regulatory control systems to operate. Changes made to the sensor system and the process itself as well as the natural degradation over time of the equipment may cause some controllers' and process performance to suffer. Since there are hundreds of controllers in most complex processes, it is virtually impossible to monitor their performance manually. This work reviews the state-ofthe-art in controller performance monitoring including both feedforward and feedback control. The experiences of the DuPont Company in the deployment of a system for performance monitoring are presented. Information as to how such a monitoring system is implemented in the operation of a chemical process and examples of its operation are included.

Dynamic Data Reconciliation to Decrease the Effect of Measurement Noise on Controller Performance Assessment

“…As pointed out by Hoo et al . , about 60% of the industrial control loops are not performing as well as they are expected. It is necessary to design a real‐time system that could monitor and evaluate the controller performance with many control loops.…”

Automated Evaluation of Biochemical Plant KPIs based on DEXPI Information

“…In process control, there are many conventional methods to evaluate the performance of controllers, such as using dimensionless numbers [15], and fast Fourier transformation [16]. Many applications focus on the monitoring of the controllers [17], but little work focuses on the evaluation of the overall process. This is because the evaluation of the process highly depends on the process itself and cannot be generalized easily and is also dependent on the system where the process takes place.…”

Study on outlier robustness of minimum variance control performance assessment

“…The research describes industrial applications of the CPA method as well. The first reported implementation in pulp and paper industry 40 is followed by chemical engineering 2 , other process industries 41 , mechanical engineering 42 , Heating Ventilation and Air Conditioning (HVAC) 43 and others. Industry introduces complexities, non‐linearities, non‐stationarity and limited information.…”