Prior to the era of distributed control systems (DCS), alarms were hard wired into fixed annunciator panels mounted on the walls of a control room. The panels consisted of a discrete input and a red light on a window that was connected by a wire.
Creating an alarm was a manual and permanent process because the installer had to drill a hole in the panel and run copper wire to activate the signaling device. This method required that significant thought be put into what events should be alarmed and why.
The introduction of the DCS changed traditional alarming methods. Alarms are now built into the distributed control system application software, making it a relatively simple software change to create new ones.
This has led to a proliferation of alarms that may not be meaningful or relevant to the operator. Saturating the operator with nuisance alarms can cause them to ignore new alarms, while overloading them with too many alarms can cause them to miss critical alarms.
Because it is easy to add new alarms, it is important now more than ever to implement an effective alarm management program.
While it can take months to design and implement a new plant, the plant may be in operation for more than 20 years. This means operators will potentially live with the decisions others make during the design and implementation phases of the project for a long time!
Most additions, enhancements, and migrations of distributed control systems are treated and executed as a project.
However, alarm management is an ongoing process that is never complete. Therefore one of the keys to creating a successful alarm management program is to realize that it is NOT just a project, but an ongoing process. The ISA-18.2 standard suggests that alarm management is not just about hardware or software, but it is about the work process, or alarm management lifecycle.
Rockwell Automation and exida, a recognized Encompass? Partner, have condensed the alarm management lifecycle into seven major steps to consider for executing a successful alarm management project with the PlantPAx? DCS:
- Benchmark alarm system performance. Quantify the average number of alarms/operators, identify nuisance alarms and “bad actors.”
- Develop an alarm philosophy. What constitutes the need for an alarm? How are alarms prioritized?
- Perform alarm rationalization. The goal is to create the optimum set of alarms that will maintain plant safety and keep the operation within normal operating limits. It is also important to document the rationale for why each alarm is needed.
- Advanced alarm design. Add logic to suppress alarms when equipment is not in use or after plant trips to help prevent alarm floods.
- Implementation of rationalization results. Load alarm configuration changes into the PlantPAx system and create alarm response procedures for presentation to the operator in the HMI.
- Performance monitoring and assessment. Regularly review alarm system performance reports to drive continuous improvement.
- Audit. Compare the PlantPAx alarm settings versus those from rationalization and verify that processes and procedures are being followed.
Proper alarm management is an ongoing commitment. In a PlantPAx system, there is no reason to live with a poorly performing alarm system. The system software offers easy configurability and the Rockwell Automation Library of Process Objects integrates alarm configuration and common suppression techniques to reduce engineering time and deployment.
Simple steps can be taken to execute and sustain successful alarm management with your distributed control system. To get started, access the Rockwell Automation white paper series on alarm management:
You can also learn more about exida and their tools, as well as process solutions from Rockwell Automation.