Monitoring of an Industrial Dearomatisation Process

Abstract

Process monitoring methods have been studied widely in recent years, and several industrial applications have been published. Early detection and identification of abnormal and undesired process states and equipment failures are essential requirements for safe and reliable processes. This helps to reduce the amount of production losses during abnormal events. In this paper, statistical multivariate methods and neural networks applied in monitoring of an industrial dearomatisation process are compared. No appriori process knowledge for the methods were assumed. The data for the comparison were generated with a dynamic simulator model of the process. Special emphasis was put on a case of internal leak in a heat exchanger.

Fault detection and isolation of an on-line analyzer for an ethylene cracking process

“…Therefore, the direct use of the available process measurements often poorly describes the process behavior, and the fundamental process variables or their approximations have to be derived from the measurements by calculations. This obviously improves the identification capabilities of most monitoring methods, as demonstrated in the previous studies of a similar nature by Jämsä-Jounela, Laine, and Ruokonen (1998), Yoon and MacGregor (2001), Bergman et al (2002), and Komulainen, Sourander, and Jämsä-Jounela (2004). A fundamental understanding of the relationships between process variables is essential in creating the right computed process variables.…”

“…Comparative studies have been made between process history-based methods, e.g. by Bergman, Sourander, and Jämsä-Jounela (2002), Jämsä-Jounela, Vermasvouri, Enden, and Haavisto (2003) and Chiang, Russel, and Braatz (2000).…”

A Toolbox for on-Line Process Monitoring With Some Industrial Applications

“…Another result was that principal component pre-processing clearly improved the SOMs' performance. The results are presented in more detail by Bergman et al (2002).…”