How the rate is calculated
Annex F of the standard gives the method: the number of false alarms in the last 12 months, multiplied by 100, divided by the number of automatic fire detectors on the system. The result is false alarms per 100 detectors per year. Two details from the worked examples matter:
- Manual call point false alarms are left out of the rate. They still count towards the other triggers and still need reviewing.
- An aspirating system or an optical beam counts as one detector, however many sampling holes or metres it covers.
The rate is worked out per system, not per estate: a large site with two panels has two rates.
When a preliminary investigation is needed (31.2)
| Trigger | Threshold |
|---|---|
| a) Rate | More than 4 false alarms per 100 automatic detectors over the previous 12 months |
| b) Volume | More than 10 false alarms since the previous service visit |
| c) Repeats | More than 1 false alarm (not counting good intent) from the same call point, detector or detector location since the previous service visit |
| d) Persistent cause | A persistent cause has been identified |
| e) Fire service calls | More than 2 unwanted fire alarm signals in the previous 12 months |
An "unwanted fire alarm signal" is a false alarm that was passed on to the fire and rescue service, for example by an alarm receiving centre, whether or not an appliance actually turned up.
When an in-depth investigation is advised (31.4 and 31.5)
| System size | In-depth investigation if, in a rolling 12 months |
|---|---|
| More than 40 automatic detectors | The rate is above 5 per 100 detectors, or more than 2 false alarms came from a single call point, detector or detector location |
| Fewer than 41 automatic detectors | More than 2 false alarms in total |
An in-depth investigation is a job for suitable specialists with knowledge of false alarm causes. Clause 44.3 then covers the action taken to bring an unacceptable rate down.
The four kinds of false alarm
Clause 30 sorts false alarms into four categories, which the logbook should record:
- Unwanted alarms: the system responded as designed, but to something that was not a fire, such as cooking fumes, steam or dust.
- Equipment false alarms: caused by a fault in the system.
- Malicious false alarms: someone set it off knowing there was no fire.
- False alarms with good intent: someone genuinely believed there was a fire.
Good-intent alarms are excluded from the repeated-device trigger in 31.2 c), because a person acting in good faith is not a sign the equipment is wrong.
A worked example
A hotel system has 120 automatic detectors. In the last 12 months it has had 6 false alarms from detectors and 2 from call points. The rate is 6 × 100 ÷ 120 = 5.0 per 100 detectors. That is above 4, so a preliminary investigation is needed under 31.2 a). It is not above 5, so the in-depth trigger in 31.4 a) is not met on the rate alone. If three of those alarms came from the same kitchen-corridor detector, 31.4 b) would apply and an in-depth investigation would be advised.
Who checks this, and when
The servicing engineer should record the 12-month false alarm rate at every six-monthly visit (clause 43.2) and start an investigation where a trigger is met. Between visits, the responsible person's logbook is the source: every false alarm, its cause, the device and whether it was passed to the fire service.
Common questions
What is an acceptable false alarm rate for a fire alarm?
BS 5839-1:2025 does not call any rate acceptable, since every false alarm should be reviewed. A rate above 4 per 100 automatic detectors per year needs at least a preliminary investigation, and above 5 per 100 (on systems with more than 40 detectors) an in-depth one.
How do you calculate a fire alarm false alarm rate?
Multiply the false alarms from automatic detectors in the last 12 months by 100 and divide by the number of automatic detectors on the system (Annex F). Call point false alarms are left out of the rate, and an aspirating system or beam counts as one detector.
Do manual call point false alarms count?
They are left out of the per-100-detector rate, but they still count towards the other clause 31 triggers, such as more than 10 false alarms since the last service or more than 2 in a year on a small system.
What is an unwanted fire alarm signal?
A false alarm that is passed on to the fire and rescue service, for example by an alarm receiving centre. More than 2 in 12 months triggers at least a preliminary investigation (clause 31.2 e).
Is there a free false alarm rate calculator?
Yes. The FireCompliance Pro calculators page works out the Annex F rate and shows which clause 31 triggers apply, at firecompliancepro.com/tools/.
FireCompliance Pro schedules every service inside the 5 to 7 month window, issues the certificate the moment the engineer signs off, and gives your client a live portal showing exactly where each building stands.
Start your free trial