Pre-Alarm Signals on Fire Alarm Systems
Pre-alarm is one of the more useful things an analogue addressable system does and one of the most casually ignored. It sits in an awkward place: not urgent enough to empty the building, not broken enough to raise a job, and so on plenty of sites it becomes background noise that someone silences on the way past.
The essential idea worth holding onto is that a pre-alarm is not a malfunction. It is the system telling you, accurately, that a detector is reading higher than it should be and lower than an alarm. Something is going on at that address. Troubleshooting it properly is usually quick, and the answer is nearly always worth having.
Who this is for
This is for fire alarm engineers and service technicians, and for duty-holders trying to work out what the panel is telling them. Experience level: competent engineer for the diagnostic sections; the procedural sections are written to be usable by a responsible person without an engineering background. Pre-alarm terminology, thresholds and configuration are manufacturer-specific, so the equipment documentation for the system on site is the authority on how it behaves.
What a pre-alarm actually is
On a conventional system a detector is essentially a switch: it is either in alarm or it is not. On an analogue addressable system the panel is reading a value from each device and deciding what that value means, which allows it to recognise a state between "normal" and "fire".
That intermediate state is what most manufacturers call pre-alarm, and it is generated when a device's reading passes a lower threshold configured in the panel without reaching the alarm threshold. Because the panel is holding the decision logic, the threshold, the indication and what else happens as a result are all configurable, and panel behaviour differs considerably between platforms. Some systems will annunciate it at the panel only; some will drive an output through the cause and effect programming; some will log it silently and show it in the analogue values. What it will not normally do is operate the general alarm.
Two things follow from that. First, the specific numbers involved are properties of the equipment and its configuration, not of any standard, and they should be read from the manufacturer's documentation for the device and panel rather than assumed. Second, pre-alarm is a facility, not a requirement — plenty of perfectly compliant systems do not have it, and its presence or absence is a design decision.
Pre-alarm, alarm, fault and staff alarm
These four get confused constantly, and the distinctions matter when you are writing procedures for a site.
A fire alarm is the system's determination that there is a fire condition, with the evacuation response the building has planned. A fault is the system reporting that it is impaired — something is disconnected, missing, out of range or unpowered. A pre-alarm is a healthy system reporting an elevated reading below alarm level. A staff alarm with an investigation period, by contrast, is a designed delay in the alarm response itself, used where false alarms would be disruptive and there are staff available to check; BS 5839-1 gives recommendations on arrangements of that kind, and it is a design feature agreed at the outset rather than something that happens by accident.
The practical difference: a fault generates a repair, a pre-alarm generates a look, and a staff alarm generates a timed, procedural response. Treating a pre-alarm as a fault sends the wrong person with the wrong expectations. Treating it as an alarm empties the building unnecessarily and teaches everyone to distrust the system. Related reading on the designed-delay side is in staff alarm and investigation delay.
The likely causes
The common causes are few and they repeat. Roughly in order of probability as they turn up on service:
Contamination and baseline drift. Dust, grease, insects and general age raise a detector's resting reading. Many panels compensate for slow drift within limits, and a pre-alarm is frequently the first visible sign that a head has drifted past what compensation can absorb. That relationship is worth understanding properly — see analogue sensitivity and drift compensation.
The environment at that location. Steam from a shower room or kitchen, exhaust in a loading area, dust from a workshop, aerosols in a cleaning cupboard, candles and cooking in residential settings. The detector is doing its job; it has simply been asked to do it somewhere unhelpful.
Airflow and siting. Detectors near diffusers, doorways, high-level voids and stairwells see different air from the rest of the room, and can sit persistently elevated or swing about. Siting is a design question, and a repeat pre-alarm is sometimes the building telling you the original design assumption no longer holds because the space has been repartitioned or re-used.
Sensitivity settings and mode changes. Where day and night sensitivity modes are configured, a device that is comfortable during the day can pre-alarm overnight on the more sensitive setting. If the pre-alarms cluster at a time of day, check the day and night operating modes before anything else.
A genuine, slow-developing condition. Rare, and the reason none of the above should be assumed. Overheating electrical equipment, smouldering material in a bin store, a fault in plant. This is exactly the scenario the facility exists to catch, which is why "it's probably just a dirty head" is a thought to have after you have looked, not before.
Investigating a pre-alarm
On arrival, the initial observations are quick and they matter.
Note the address and location, and read the event log for the pattern: is this the first occurrence, or the fortieth this month; does it happen at a particular time; does it clear on its own; is it one device or several in the same area. The event log will answer most of that in a couple of minutes, and it is the difference between diagnosing a device and diagnosing a location.
Then go and stand where the detector is. Site observations are decisive here in a way they are not for wiring faults — what is in the room, what has changed, what is being stored, what is being done nearby, whether the space is being used as designed. Evidence gathering at this stage is mostly a matter of looking and asking.
If the panel exposes analogue values, read the device's current value and, if the platform stores it, the trend. A device sitting consistently elevated points at contamination or environment; one that spikes and recovers points at an activity happening in the room.
Safety warning. Approach a pre-alarm as though it might be real until you have established otherwise, particularly in plant rooms, electrical intake rooms and storage areas. Do not open panels or enclosures you are not authorised to open in order to satisfy curiosity about a heat source, and if you find anything you are not comfortable with, treat it as a fire condition and follow the site's emergency procedure.
Investigation flowchart
As an investigation flowchart:
- Identify the address and its actual physical location, not just the zone text.
- Read the event log for frequency, timing and clustering.
- Attend the device and assess the environment around it.
- Read the analogue value and trend if the panel provides them.
- Decide between contamination, environment, siting, configuration and genuine condition.
- Clean or replace the head if contamination is indicated.
- Review type and siting if the environment is the cause.
- Review sensitivity and mode configuration if the pattern is time-related.
- Treat as a fire condition if a genuine source is found, and follow site procedure.
- Record what you found and what you changed, and tell the site what to expect next.
Fixing the cause
Repair depends on which of those five it turned out to be.
Contamination is the straightforward one: clean or replace to the manufacturer's instruction, following contamination and cleaning practice, and note that on many devices replacement is the manufacturer's preferred route rather than field cleaning.
Environment and siting need a design conversation rather than a part. That might mean a different detector type — a multi-sensor or a heat detector where a plain optical is fighting the room — or a re-site, or a change to what the space is used for. Choosing between types is covered in choosing fire detector types, and any change of this sort is a modification to the installed system that belongs in change control rather than being done quietly.
Configuration changes — thresholds, modes, whether pre-alarm is annunciated at all — should be deliberate, agreed and recorded. Turning the indication off to stop it being annoying is the one option to be genuinely wary of.
If the work needs a device or a section out of service while you deal with it, agree the isolation strategy with the site first and keep it as small and as short as it can reasonably be.
A short repair checklist: cause identified rather than assumed; device cleaned, replaced or re-sited as appropriate; any configuration change recorded with the reason; testing after repair completed on the affected device and its neighbours; the site told what was found; the record updated so a recurrence is recognisable next time.
Verification and testing after repair
Verification is simple but easy to skip. Prove the device functionally after any intervention, confirm it reports a normal value at the panel, and where the platform allows it, look again a few days later to see whether the reading has stayed where you left it. A head that pre-alarms again a fortnight after cleaning is telling you the cause was the room, not the device.
Common engineer mistakes
Silencing pre-alarms visit after visit without recording that they happened. Cleaning a head repeatedly when the environment is the actual cause. Assuming contamination without ever reading the analogue value. Disabling the indication as a fix. Treating a cluster of pre-alarms in one area as five separate device problems rather than one location problem. And, occasionally and seriously, dismissing a genuine early warning because the site has a history of nuisance ones.
Estimated repair time
Estimated repair time is not worth generalising: a head swap is minutes, a detection strategy review for a repartitioned floor is a design exercise. What helps the client is an early, honest split between "sorted today" and "needs a proper look", together with your confidence in the diagnosis.
Procedure, training and records
A pre-alarm facility is only worth having if somebody responds to it, which makes this as much a procedural question as a technical one.
The premises needs a written answer to: who gets told, what they do, how quickly, and what happens if it recurs. That belongs with the responsible person, who in England and Wales carries duties under the Regulatory Reform (Fire Safety) Order 2005 — including a maintenance duty in respect of fire safety equipment provided in the premises (article 17) — with GOV.UK setting out those responsibilities in plain terms. BS 5839-1 is a code of practice offering recommendations on system design, installation, commissioning and maintenance; it is guidance on how to do it well, not legislation. Users also need to be told what the indication means, which is part of user training and familiarisation.
Recording matters because pre-alarms are a trend, not an event. A single occurrence is noise; the same address eleven times in a quarter is a finding, and it only becomes visible if somebody wrote the first ten down. Report writing should therefore give the count and the pattern, not just the last occurrence. This is also good false alarm practice, since pre-alarms frequently precede the unwanted alarms covered in false alarm management.
Report example
A useful report example: "Panel logged pre-alarm at zone 2 address 07 (first floor kitchenette, optical) on eleven occasions since the last service, clustered between 07:30 and 09:00. Device attended; analogue value elevated and consistent with contamination from adjacent toaster use. Head replaced like for like, tested and proved, value returned to normal range. Detector type and position discussed with site: a multi-sensor or a re-site away from the counter would reduce recurrence. Recommend this is considered as a small remedial item; recurrence to be reviewed at the next service."
Related faults
Related faults and topics worth reading alongside this: false alarm management, detector contamination and cleaning, and when a panel will not reset where a device that has passed pre-alarm into alarm will not clear.
When not to rely on this alone
When not to use this article: do not use it to set or change pre-alarm thresholds, to interpret a specific manufacturer's terminology, or to decide that a pre-alarm can safely be ignored. Threshold values and configuration behaviour are manufacturer-specific and belong to the equipment documentation. Whether any particular indication warrants a particular response at a particular premises is a matter for the responsible person and the fire risk assessment.
Relevant standards
Recommendations on the design, installation, commissioning and maintenance of fire detection and fire alarm systems in UK non-domestic buildings are given in BS 5839-1, current edition, including on limiting false alarms and on staff alarm arrangements. Requirements for the equipment itself sit in the BS EN 54 series. Both are standards rather than law; the statutory duty in England and Wales rests with the responsible person under the Regulatory Reform (Fire Safety) Order 2005. Always work to the current edition and to the manufacturer's data for the devices installed.
Professional disclaimer
This is an educational resource for competent engineers and for duty-holders trying to understand their system. It does not replace the current British Standards, manufacturer documentation, a competent fire risk assessment or professional judgement. Confirm any configuration change against the manual for the equipment and agree it with the responsible person before making it.
Related documentation
Read this with analogue sensitivity and drift compensation, choosing fire detector types and false alarm management. Tracking recurrence across visits is easier with the digital logbook and the fault database.
References
- The Regulatory Reform (Fire Safety) Order 2005 — legislation.gov.uk
- Fire safety in the workplace — GOV.UK
- BS 5839-1 (current edition), BSI
- BS EN 54 series (current parts), BSI
- Detector data sheets and panel programming and service manuals for the equipment on site