Application of system cultivation and the philosophy of statistical process control to a PI flow control loop

In recent years, interest in quality, productivity, and competitiveness has led to extensive use of the techniques of statistical process control (SPG). While the philosophy of SPG is very sound, the applications of its traditional techniques are limited in the multivariable environment of continuous chemical processing. Furthermore, traditional SPG is not useful in the analysis of the performance of control systems. However, a new technique, called system cultivation, has emerged to provide a means of systematically sorting through process data to analyze multivariable relationships for continual process improvement. It also provides a means of monitoring the quality of control systems. This study shows how system cultivation may be used to study the steady state behavior and the dynamic behavior of a PI flow control loop. It shows how the cultivation techniques of time-shifted distribution, hypothesis feedback modeling, and controlled resolution may be used to monitor controller performance, changes in a flow system, and valve hysteresis. An important conclusion is that the philosophies of statistical and conventional process control are compatible and mutually beneficial.