Researcher
Research institution
Champion
Focus team
Topic
Project status
Year ended
2009
Project ID
200801
Why should I care about this project?

How many alarms are too many alarms?  Alarm standards provide some guidance, but this study put those standards to the test using students.

Abstract

Dr. Craig Harvey tested when increasing alarm rates degrade performance. LSU students responded to five different alarm rates using a pipeline simulator with two methods of presenting the alarm information. Performance was constant on lower alarm rates up to and including 10 alarms per 10 minutes. Performance degraded significantly (p<.05) at twenty alarms in 10 minutes, although less when the alarms were presented grouped by priority than when presented by time of actuation. Follow-on project was identified to determine if the same effect would be seen (1) for actual operators (instead of students) and (2) if the period of alarm activity was 60 rather than 10 minutes.

Objective

Quantify the performance impact of alarm rate targets utilized in EEMUA and ISA alarm management guidelines. A secondary objective was to demonstrate that multiple variables can be evaluated within the same experiment (multi-factor design), so different formats for presentation of alarms were utilized.

Driving questions
  • At what alarm actuation rate (alarms/minute) is the operator’s ability to detect critical alarms compromised?
  • Does that vary with presentation formats?
Background

Both EEUMA 191 and ISA 18.2 provided targets/limits for alarm actuation rates of one alarm in ten minutes for steady state operation and ten alarms in ten minutes for upset operation. As these targets were selected based upon “judgement of experts”, at issue is whether they can be supported by objective, empirical data.

Deliverables
  1. Chart showing performance breakdown occurred between 10 and 20 alarm per ten minutes, with the breakdown more pronounced with the presentation by time of actuation.
  2. Submitted journal paper(s) outlining ways to improve alarms and alarm handling for SCADA operators. Conference paper accepted for presentation at 3rd Applied Human Factors and Ergonomics (AHFE) International Conference 2010 http://www.ahfe2010.org/
  3. Presentation to API Cybernetics conference in April 2010. Slides can be found at http://www.api.org/meetings/topics/pipeline/upload/Use_Simulation_for_SCADA_Alarm_Study.pdf
  4. Empirically Evaluating and Developing Alarm Rate Standards for Liquid Pipeline Control Room Operators, Graduate Thesis of Glen Uhack, May 2010.