Alarm load scorer
Four numbers from one operating position, graded against the published EEMUA 191 / ISA-18.2 / IEC 62682 benchmarks — with the rationalisation moves that fit your profile.
Grade one operating position
- Measure and trend the percentage of time above 10 alarms per 10 minutes — EEMUA 191 treats flood time as a primary health metric. Make it a weekly operations KPI with an owner.
- Clear the standing-alarm list: every alarm that stays active for days teaches operators to ignore the annunciator. Re-engineer, re-prioritise, or remove each one with a documented decision.
- Run a bad-actor review: the top 10-20 alarm tags typically produce most of the daily load. Rationalise cause, consequence, and response for those tags first (ISA-18.2 rationalisation stage).
The benchmarks behind the bands
The scorer compares your inputs against the published alarm-management benchmark family. These figures share one lineage — the EEMUA 191 work that followed the Bransby & Jenkinson operator-workload survey — and they are targets for screening, not contractual ceilings. ISA-18.2's 2016 edition deliberately dropped per-day KPIs to stop exactly that misuse, which is why the scorer also weighs floods and standing alarms instead of grading on the daily count alone.
Primary references: EEMUA Publication 191, ANSI/ISA-18.2, and IEC 62682. A deeper treatment of the numbers and their misuse: “Alarm Management by the Numbers,” Chemical Engineering.
Questions operations leaders ask about alarm load
How many alarms per day is acceptable for one operator?
The benchmark family converges on roughly one alarm per ten minutes in steady state — about 144 per day in IEC 62682, with ISA-18.2's 2009 edition citing around 150. Above roughly 12 alarms per hour sustained, the guidance treats the load as more than an operator can reliably manage.
Why did ISA-18.2 remove the alarms-per-day KPI in 2016?
Because a single per-day figure was being misread as a pass/fail line. Operator workload depends on alarm priority mix, task demands, and flood behaviour, not just the count. The metrics still appear in EEMUA 191 and IEC 62682 as targets — useful for screening, not as a contractual ceiling.
What is an alarm flood and why does it matter more than the average?
A flood is conventionally more than 10 alarms in 10 minutes for one operator. Averages hide floods: a plant can pass the daily target while its operators are overwhelmed during every upset — precisely when the alarm system matters most. EEMUA 191 treats time-in-flood as a primary health metric.
What is alarm rationalisation?
The ISA-18.2 lifecycle stage where each alarm is reviewed against the philosophy: does it indicate a malfunction needing operator action, what is the cause, consequence, priority, and response? In practice the top 10-20 'bad actor' tags usually produce most of the load, so rationalisation pays back from the first pass.