What is alarm fatigue?
Today’s complex automated building management systems (BMS) generate hundreds and sometimes thousands of alarms. BMS are constantly receiving data from all kinds of sensors within a building (chillers, lighting systems, access control systems and so on) yet often they display interrelated alerts without any indication of how they are connected.
Whether the issue is ‘nuisance’ repetitive alarms,
redundant alerts, incorrect settings, or faulty hardware, there is one
unfortunate consequence for the operator to all these irritations – alarm
fatigue. This term describes the desensitization that occurs when an
administrator is exposed to so many alarms, and so frequently, that the alerts
simply become “background noise”. This fatigue can cause delays in
response times or lead to the missing of genuine alarms altogether which can
have critical ramifications.
Drowning out the noise
Typically, the programming and commissioning of
alarm levels within systems is done in bulk, with limited information on how
the operational model of the building in question has changed over time, or how
it is likely to operate in the future. Because systems are interconnected
throughout the building, it means that one single issue can trigger multiple
alarms. For example, a chiller provides
cold water to multiple air-handling units (AHUs), in order to cool multiple
areas. If the chiller then fails, it might trigger alarms in all the connected
systems’ AHUs and room temperatures. If those alarms aren’t displayed in a
logical way to the operator, it can become difficult to make the right
operational decisions. Often, without knowing how to process them, operators
will simply turn off or ‘kill’ these alarms and, inevitably, the notifications
will eventually reappear on their screen.
A better focus
Good alarm organization can reduce the noise of non-relevant alarms. After the noise is reduced, operators can better focus on the relevant alarms, reducing downtime and risk. So, what are the optimum strategies we can use to manage alarms and thereby combat alarm fatigue?
- Prioritize via analytics – Using Fault Detection Diagnostics (FDD) – an analytical tool that identifies faults or equipment not running as specified – gives operators a clearer insight into the current status of all facilities. By providing meaningful information on the operation of assets, complex analytics help operators prioritize the critical alarms, understand their causes, and ultimately reduce their number.
- Solving root causes – It may seem obvious but solving the root cause is crucial in reducing fatigue. For example, maybe a door alarm is repeatedly being activated because it has been mounted incorrectly. After the installation issue has been resolved, the frequency of alarms from that site will reduce. Addressing the cause is the initial step in lowering the number of alarms and in turn, administrator apathy towards them.
- Clustering alarms – Identifying the most frequent alarms is another crucial step. Typically, 20% of individual alarm points are causing 80% of alarms – it is the same alarms creating the majority of notifications. We call this the 80/20 rule. Pin-pointing the most problematic alarms using an easy-to-understand portal, and resolving them, can drastically reduce noise – and ultimately administrator desensitization.
Poor alarm management can ultimately lead to unplanned downtime, increased costs and crucially, increased risk of danger to the occupants of the building. On the other hand, effective alarm management improves both productivity and safety while increasing the speed of root cause identification and correction. Good alarm management also considerably reduces instances of alarms being triggered at night or weekends, which would otherwise result in an operator needing to visit the site and associated call-out charges from the FM company. If used to its full potential, a building’s BMS should provide the operator with the relevant information to benefit from the alarm functionality. When operators do not have to spend their time sifting through alarms trying to assess which are the most critical, considerable time savings can be made as well as reducing the “noise” that operators are exposed to.
A recent customer example
In one recent example, the building owner of a shopping center wanted to streamline their BMS and only see alarms that required action. By being able to report on alarm history, we could easily visualize the frequency and severity of alarms. Criticality was displayed in a way that made it easy to digest, with the alarm categories labelled as A (critical), B (high risk) and C (medium risk). They were also color-coded accordingly. Categorization of the types of alarms also helped give a comprehensive view of what was causing the alarms. ‘Hard’ alarms are those referring to fire and safety, whereas ‘soft’ refers to alarms that are programmed to go off at certain points. In this instance, it was possible to reduce the 15,500 alarms per year to 4,671, a reduction of 70%. This meant a reduction of 42 alarms on average per day to just 13, granting the operators a considerable reduction in alarm ‘noise’ and ultimately, alarm fatigue.
Panu Kärävä is a global portfolio manager for energy and asset performance services at Siemens Smart Infrastructure.