A specialty chemical plant located in the Southeastern US realized they had issues with the way they were doing alarm management. exida provided the tools, training, and DeltaV alarm management expertise to enable the team to be successful.
A specialty chemical plant located in the Southeastern US realized they had issues with the way they were doing alarm management. Like many process plants, they had too many alarms - and many of the alarms were nuisance alarms that were not meaningful to the operator. Alarm priority was not useful; fully 60% of the alarms seen by the operator were High (Critical) priority, which made it hard for them to differentiate the relative importance of one alarm versus another. Moreover, they had reached the point where focusing on eliminating the “bad actors” was no longer providing benefit.
The plant reached out to exida, an Emerson strategic partner for alarm management of DeltaV™ systems. As the plant set out to improve their alarm management, they started by identifying their specific needs. They wanted to create a continuous improvement program that would be sustainable over time and would integrate with their existing OSHA Process Safety Management (PSM) program. Since they ran a batch process, addressing operator action requests, prompts, and batch-based alarming were important. They didn’t want to break the bank and they wanted to keep the knowledge in-house. Enter exida to provide the tools, training, and DeltaV alarm management expertise to enable the team to be successful.
Successful alarm management programs are developed by adopting the ISA 18.2 standard, which is considered recommended and generally accepted good engineering practice (RAGAGEP) by OSHA. Along with its international counterpart, IEC 62682, ISA 18.2 not only defines the alarm management lifecycle, but it gives a clear definition of what an alarm should be: “An audible and/or visual means of indicating to the operator an equipment malfunction, process deviation, or other abnormal condition requiring a timely response.” Alarm management issues are created when alarms are defined which do not meet this definition. One of the main goals of an alarm management program is to ensure that alarms are actionable by the operator and represent an unexpected situation.
To be thorough as well as efficient, the team turned to exida’s alarm management experts to guide them through the steps to implement an ISA-18.2 compliant alarm system and create a robust program tailored to their needs. The exida alarm management program, which can be applied to any control system in both greenfield and brownfield applications, includes the following steps:4
Figure – exida Alarm Management Program
The alarm management program starts with benchmarking to identify systematic and systemic issues, as well as to document the starting point for performance improvement. DeltaV Analyze was used to benchmark alarm system performance using 30 days of data for each of two different operator positions (with similar spans of control). The team discovered that a high percentage of the alarms presented to each operator were High (Critical) priority. Although the situation was different between the two operator positions—90% for Operator 1 and 48% for Operator 2—it was a significant gap compared to the ISA-18.2 recommendation (~5% High priority).
The benchmarking also analyzed the annunciated alarm rate. “Acceptable” performance was defined as ≤ 1 per 10 minutes. Using this criteria performance was deemed “Acceptable” 67.6% of the time for Operator 1 and 52.5% for Operator 2.
Figure – Initial Alarm System Performance Benchmark (DeltaV Analyze)
With exida’s help, the team developed an Alarm Philosophy document, which is the foundation of a successful alarm program and is a pre-requisite to alarm rationalization. A three-day training / workshop was held, bringing together key stakeholders. For the plant, this document helped them affirm the following key topics:
Next, the exida team helped the plant tackle alarm rationalization. Rationalization entails reviewing each alarm against the criteria documented in the alarm philosophy to justify that it meets the requirements. It includes defining and documenting the design attributes (such as priority, setpoint, type, and classification) as well as the cause, consequence, time to respond, and recommended operator response in a master alarm database (MADB). It is a team activity, similar to a HAZOP, involving production / process engineers, process control engineers, experienced operators, and other personnel as needed.
The process began with two days of rationalization and facilitator training, followed by five days of team workshops, all using exida’s SILAlarm tool (and master alarm database). After this, the plant was on their own to plow through the rest of the rationalization.
The Rationalization Process (ISA-18.2 TR2):
Alarm rationalization can be overwhelming without an astute plan of attack—one of the keys to success is to assign a core set of senior operators to be part of the rationalization team. Operations input and buy-in is a critical success factor.
It is also important to have knowledgeable persons take part who can quickly access process historian trends, process hazard analysis (PHA), and operating procedures. Process Safety Information (PSI) from their OSHA PSM program such as critical variables, safe operating limits, consequences of deviation, and corrective actions, was used to jumpstart the process.
The plant found that four hours maximum for each session was optimal. Overall, 60 sessions were conducted to complete rationalization of approximately 27,000 alarms. The process engineer who took on the role of facilitator / scribe ensured that causes, consequences, and corrective action were documented consistently so the results could be copied to similar equipment in the SILAlarm database. This minimized the number of sessions and the time commitment of the team, allowing them to complete rationalization for upwards of 900 alarms per day.
Alarm rationalization resulted in a significant change to the design of the alarm system, as shown below. More than 10,000 alarms were eliminated. Non-alarm notifications, such as alerts and prompts, were identified separately from alarms and moved to display in their own summary list.
The configured alarm priority distribution is significantly different and is now consistent with the plant’s alarm philosophy.
Criteria |
Before Rationalization |
After Rationalization |
Enabled Alarms |
17,188 |
4,773 |
Enabled Non-Alarms (Alerts, Prompts) |
------ |
3,141 |
High (Critical) Priority Alarms |
85.5% |
5.6% |
Medium (Warning) Priority Alarms |
12.3% |
21.2% |
Low (Advisory) Priority Alarms |
2.2% |
73.2% |
One of the benefits of using exida’s SILAlarm tool was that DeltaV’s Alarm Help faceplates could be configured automatically using the information documented during rationalization. Now an alarm’s cause, consequence, corrective action, time to respond, etc. are shown instantly to the operator in context when the alarm occurs. If done correctly, rationalization can be an effective knowledge capture process. Display of alarm help, particularly for junior operators, is a form of knowledge transfer, allowing them to see how the most experienced operators would respond.
During rationalization the plant identified opportunities for implementing conditional alarming to eliminate nuisance alarms. Their standard class-based control modules included the ability to dynamically disable / enable a low flow alarm (for example) based on the run status of an upstream pump.
Rationalization results, including changes to alarm parameters, were implemented in the control system in November 2017 by using DeltaV’s bulk edit capability to import a configuration file produced by SILAlarm.
An additional benefit was delivered by updating the documented operator actions associated with Safe Operating Limits and Consequences of Deviation for their PSM program. Updated procedures were created automatically from a SILAlarm output file (in Excel format).
After implementing the results of rationalization, alarm system performance was measured again using DeltaV Analyze. A major shift was seen in the annunciated alarm priority distribution. The percentage of High (Critical) priority alarms dropped from 90% to 0.6% for Operator 1, and from 48% to 3% for Operator 2—both consistent with the ISA-18.2 recommendation (~5% High priority). Now the high priority alarms really stand out and get the operator attention they deserve.
Figure – Alarm Priority Distribution Before and After Rationalization
The average alarm rate was also impacted. The percentage of time the average alarm rate was “Acceptable” went from 67.6% to 77.5% for Operator 1 and from 52.5% to 82.1% for Operator 2.
These days, on a monthly basis, the team looks at performance metrics monthly for Operator 1 and 2 via DeltaV Analyze. From this report, they create their action items and share it with the group. They can also track how performance changes from one month to the next.
As of 2019, the team is now focusing on stale alarms.
The plant has seen significant benefits from their successful alarm management program.