Every year, poor alarm management results in the loss of billions of dollars due to accidents, equipment damage, unplanned plant or unit outages and costs related to environmental and safety infractions.
Clearly, despite the fact that it is one of the most important factors in maintaining a safe and stable plant, alarm management continues to be one of the most overlooked and undervalued parts of the manufacturing process.
It has certainly come a long way over the years with the advent of the Distributed Control System (DCS), but the issue is that operators are becoming overwhelmed with unnecessary alarms, causing them to start suppressing the biggest offenders.
This may seem harmless during regular day-to-day operations, but it could be disastrous in the event of actual plant accidents.
This is one of the reasons a proper alarm management system is so important.
The problem
Up until recently, hard-wired panel board alarms were the main mechanism for alerting control room operating staff to potential problems.
These alarms were limited by both the amount of available board space and the cost of running wiring, as well as the need to hook up an annunciator, indicator, and switches.
Therefore, the addition of alarms was done sparingly and operators were rarely flooded with alarms during abnormal situations.
With the advent of the DCS, operators gained far more control over alarm systems and had the ability to add more alarms when needed. Unfortunately, when the new alarms were developed, many of them were poorly configured.
To make matters worse, today we expect these static alarm settings to adapt to dynamic plant conditions, which just isn’t possible.
These two factors combined are the culprit for alarm floods, which prevent operators from properly accessing the root cause of problems.
The only solution to this issue is a total overhaul of the alarm management system.
Revamping the system
Many plants are still using the alarm management philosophy that was developed when the plant was first built, failing to realise that just as the needs of the plant change, so must the alarm management system.
It can certainly be overwhelming to overhaul an entire alarm management system at once, which is why a simple, five-fold procedure is recommended.
Define
The alarm philosophy document is the most important part of the system. This document will form the basis of the overall design guidelines and will record all the KPIs that will be used to measure the success of the alarm management system.
Measure
Benchmarking is an effective way to determine whether the system is improving.
Record things such as the number of alarm counts per day and the amount of alarms operators handle on an hourly basis.
Analyse
Checking and resolving bad actors (alarm rationalisation) is crucial to improving alarm management. This process can be assisted with the use of software programs.
Improve
Once the alarm rationalisation process is completed, it is essential to keep management controls in place, or else the system will revert back to its old ways. There are also software programs to ensure all alarm changes are tracked, as well as identifying changes to alarm settings.
Control
Continued compliance to the alarm philosophy is important. It is worth considering the nomination of an “alarm champion” that will monitor the alarm KPIs and make any required changes.
Alarm rationalisation
This is one of the most important parts of the alarm management system. Simply eliminating the top ten most problematic alarms by resolving bad actors and chattering alarms will quickly make a significant improvement in the overall alarm count.
There are a number of software programs that can be used to obtain bad actors, such as Yokogawa’s alarm/event analysis software tool, Exaplog, or its alarm reporting and analysis software, Exaquantum ARA.
The difference between the two is that Exaplog enables the user to run manual reports when needed, whereas in ARA a report can be generated automatically and sent through email.
These software programs can also evaluate “chattering alarms”, which (according to the EEMUA#191 alarm standard) are alarms that go from alarm mode to normal again more than five times in 60 seconds.
Typically, this is the result of incorrect alarm limits, and can be easily resolved. EEMUA#191 recommends no more than one alarm per operator every 10 minutes.
Considering the current state of most alarm systems, this goal would be difficult to achieve in many plants. However, significant progress can be made just by employing alarm rationalisation.
Of course, it is of equal importance to continue monitoring and adapting the alarm management system.
Poor alarm management can be extremely costly (in more ways than one), so it is of the utmost importance that the system is regularly monitored and given the priority it deserves.
Yokogawa Australia
(02) 8870 1100
www.yokogawa.com/au/