Abstract

The ambition to increase the production of paper has made paper machine runnability widely studied in recent decades. Paper machine runnability is often measured by the number of web breaks in comparison with paper machine speed. When runnability is good, a machine can be run at the desired speed with the least possible number of breaks. Web break sensitivity means the number of breaks in a day, which can also be understood as a measurement of paper machine runnability.

This study presents an application for the evaluation of web break sensitivity in a paper machine. A web break sensitivity indicator was built using the basic principles of case-based reasoning with a linguistic equations approach and basic fuzzy logic. The indicator combines on-line measurement data with expert knowledge and provides a continuous indication of break sensitivity. Web break sensitivity defines the current operating situation at the paper mill and provides new information to operators. Web break sensitivity is presented as a continuous signal with information about actual web breaks depicted as an eight-hour trend. The trend shows how the situation has developed and the current value gives a prediction for the next 24 hours if the situation stays as it is now. Together with information about the most important variables, this prediction gives operators enough time to react to the changing operating situation.

From the methodological point of view, a new tool for building case-based reasoning applications for other purposes was also tested and found to be suitable for diagnostic applications.