As discussed in What do Nuisance Alarms, the 80-20 Rule, and Mental Models Have in Common?, there are typically a handful of alarm points (10 to 20) that create the majority of notifications (50-80%) to the operator (referencing the 80-20 rule). These nuisance alarms are affectionately called “bad actors” within the alarm management community. The focus on resolving bad actors (called “Bad Actor Knockdown”) is often the introduction to alarm management for many plant personnel. They jump in with no training, no alarm philosophy, and no defined alarm rationalization process, eager to improve the performance of the alarm system. 

And why not, you might ask? For a system with 10,000 configured alarms, the potential to fix just 10 alarms and reduce the number of alarms the operator receives by 50% is compelling. Any process or  controls engineer that can improve performance this drastically is sure to be appreciated by the operations staff. 

So, what is the downside to focusing on bad actor knockdown?

In a typical Bad Actor Knockdown program, one focuses on evaluating the validity and design of a single alarm without considering how it relates to the rest of the process. It can become something like the game of Whac-A-Mole: as soon as you get rid of one, another one will raise its head. Despite your best efforts, new nuisance alarms appear randomly and ad infinitum, seemingly triggered by changes in the season, the weather, the process, as well as plant or equipment trips. 

Bad Actor Knockdown can be beneficial at eliminating or correcting obvious design mistakes, but it does not improve the overall quality of the alarm system (for example, correcting alarm priority to accurately reflect each alarm’s relative importance); it only eliminates the outliers. I have heard of cases where plant management don’t think that any other activities are required to achieve acceptable performance. This is a dangerous misconception. 

Bad Actor Knockdown can be an effective starting point that gains buy-in and momentum for pursuing a more comprehensive alarm management program—one that follows the ISA-18.2/IEC 62682 alarm management lifecycle. It might even help you meet an acceptable alarm rate (no more than 1 to 2 alarms / 10 minutes). It cannot, however, deliver an alarm system that is compliant with the ISA-18.2 and IEC 62682 standards.

An effective alarm management program would include an alarm philosophy document that lays out the criteria for an alarm, alarm prioritization, and the proper application of alarm deadband and on/off delay. It would also include alarm rationalization; a comprehensive process for determining the need for an alarm and its priority, while also documenting the likely causes, consequences of inaction, corrective actions, and time to respond in a master alarm database.  

Following are some recommendations to consider if you are going to do Bad Actor Knockdown:

• First define the criteria for what is a valid alarm and the guidelines for application of alarm deadband, on/off delay. (This is typically done in an alarm philosophy.)
• If you have an alarm philosophy in place, follow the rationalization process and rationalization rules.
• Document the changes and rationale for the changes in a master alarm database or in a form that it can be put into a master alarm database later.

Want to learn more about how to build an effective alarm management program? Watch Alarm Management and ISA-18.2 / IEC 62682: How Do I Get Started?