Master's Thesis

Larmhantering i processindustrin

Björn Stenquist


The objective of this thesis is to reduce the alarms in the control rooms of the processing industry and to investigate a strategy for working with the alarms. The problem with the alarms in the automation system is a result of the highly automated production and failure to design the alarms for the operator. The situation analysis showed problems with badly tuned alarms and cascade of alarms. The badly tuned alarms result in alarms triggers without any danger for the process or personal, it also results in silent alarms never signaling although the risk for failure is high. The cascades of alarms occur when one fault triggers many alarms due to the correlation of process parameters. From the situation analysis was concluded that a structured work with alarms and forming groups addressing the problems would be the best way to reduce troubles with alarms. To structure the work with alarms a document with guidelines were created. This document includes the personnel of the alarm group and their tasks. It also describes the definition of an alarm which is a signal sent to the operator showing that the process doesn´t behave as it should and the system therefore call for the operators’ attention to act. A part of the alarm document also describes how to organize the information about the alarm points. The alarm points were documented by extracting data from the automation system to Excel and formatted by VB-script. To make it easy to update the information the script can automatically compare the data in the automation system with the data stored in documentation. The personnel will add information about the intention of the alarm and the measure to take when the alarm is signaling. The user added information is to justify the alarm. A pilot project with forming an alarm group documenting and structuring the alarms was conducted. Results are promising and show the working with alarms really can cut the alarm frequency and improve the working environment for the operato

Master's Thesis ISRN LUTFD2/TFRT--5855--SE, Department of Automatic Control, Lund University, Sweden, June 2010.

Download full document

Download Popular Science Summary