One of the key objectives of a process control system is to maintain product quality parameters within specification. On-line analyzers that automatically provide measurements of material properties to the control system may be too expensive, may have proven to be unreliable, or simply may not be available because of the complexity of the sampling system or the complexity of the analysis. As a result, most process plants include one or more labs that are centrally located in the plant or distributed throughout the process to provide an analysis of the material properties that must be controlled. Thus, a gap often exists between the time a grab sample is taken for analysis and when the lab results are available to the plant operator.
As is addressed in Advanced Control Foundation – Tools, Techniques and Applications, through the use of on-line data analytics on-line decision support for operations personnel can provide:
• Product quality predictions: Identify quality problems while there is time to make on-line corrections to prevent them.
• Early process fault detection: Detect abnormal process operation and/or equipment problems before they affect production. Provide root cause analytics to direct operations or maintenance personnel to quickly correct the cause of problems.
However, applying on-line data analytics to a continuous process is complicated by the fact that changes in throughput or the product grade may occur at any time. When the plant’s normal production schedule requires frequent changes in throughput or product grade, then to utilize data analytics for on-line quality parameter prediction and fault detection some means must be provided to account for these changes. The approach we have taken in the implementation of on-line continuous data analytics is to apply the concept of operational states.