Alarm management is a hot topic of late, perhaps because incorrect presentation of alarms often confuses operators. This confusion can lead to operator error, a leading cause of many industrial incidents. Transmission & Distribution Worldreports that operator error was responsible for the November 4, 2006, blackout in Europe that left millions in the dark.
It’s easy to blame operators, but better for engineers and engineering managers to improve alarm management and other systems to make operators’ jobs easier. The benefits of improved alarm management include reduced incidents and shutdowns, improved process control, improved safety, faster incident reviews and more efficient maintenance.
Matrikon is a leading firm in the areas of alarm management, OPC connectivity and control asset monitoring. Its web site identifies a six-step process for life cycle management of alarm systems.
Step 1 is benchmarking and assessment. Alarm system performance can be benchmarked against industry guidelines such as EEMUA’s document 191. Some of the metrics used for benchmarking include the average alarm rate, the peak alarm rate and the percentage of hours in which there are more than 30 alarms.
These metrics are used because they effectively characterize alarm system performance, are simple to calculate and can be generated automatically to track the ongoing performance of the alarm system. In addition to this quantitative analysis, work practices are also scrutinized during benchmark and assessment to ensure sufficient due diligence exists for adequate maintenance of the alarm system.
Step 2 is to create an alarm philosophy document that defines governing rules for alarm settings. This document helps ensure the rest of the alarm management lifecycle is executed in a consistent manner throughout the plant and across the organization. It helps establish how alarms are prioritized, what criteria must be met before an alarm is configured and who is accountable for alarm management.
Step 3 is alarm rationalization. This step consists of review and rationalization of each configured alarm in the control system based on the rules defined in the alarm philosophy document. The primary goals are to ensure that every alarm requires unique intervention and that each alarm is prioritized in a consistent manner. Alarm rationalization is done from a process-centric, rather than an equipment-centric perspective, as is the case during hazard and operability studies. Representatives from both operations and engineering should be on the alarm rationalization team.
Step 4 is implementation and execution. This involves the implementation of the rationalized alarm settings on the automation systems. During this stage, transitional training is required to ensure past bad habits, such as alarm complacency, are discontinued.
An example of a detailed implementation and execution issue would be an existing system where three temperature transmitters operate in parallel for redundancy on a critical process. The existing system might have three separate alarms. The new system should have a single alarm that would only be activated when one of the three temperature transmitters violated its limit. An alternate approach would be to have the alarm activate when two out of the three transmitters violated their limits.
Step 5 is maintenance. This step involves long-term alarm system performance monitoring, whereby a plant integrates alarm management into existing workflows and establishes change procedures for alarm management. If left to its own devices, alarm system performance will degrade over time due to process and equipment changes. Maintenance ensures that when inevitable changes occur, alarm system implications are taken into account. This ensures continued system reliability and high performance of the alarm management system.
Step 6 is continuous improvement, which is the process of identifying new opportunities for advanced alarm management strategies, such as dynamic alarming and early event detection. Dynamic alarming is the programmatic suppression of alarms during abnormal situations to prevent alarm floods. Early event detection involves the creation of pre-alarms using multivariable statistical analysis. When correctly configured, pre-alarms can help plant personnel prevent abnormal situations before they occur.
In addition to the Matrikon web site, another good source of information on alarm management is the Alarm Management Handbook, recently published and coauthored by Eddie Habibi and Bill Hollifield of PAS. PAS is involved in advanced process control and alarm management. A recent issue of our sister publication Control Design provides a review of the handbook.
An online resource focused on alarm management can be found at the aptly named www.AlarmManagement.com. This web site is hosted by the alarm management firm TiPS, and it gives end users an opportunity for online discussions about the latest developments in alarm management.