Fire & Security FAQ Database

Trigger the mapped detection zone or control input and confirm the vent opens fully within the expected time, then closes/resets correctly. Check the manual override controls at the fire service access point, the standby supply, and that fault monitoring reports to a supervised location. Wind and weather interlocks can legitimately block a test — understand the control logic before recording a failure. Note travel times and any struggling actuators; they fail on the day they're needed, not on the bench. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

Test end-to-end with the receiving centre on the line, not just the dialler LED: place the system on test with the ARC, trigger fire and fault signals, and have the operator confirm what was received and how it was tagged. Verify both transmission paths on dual-path units, and check the failure-reporting time matches what the site's category requires. Take the system off test before you leave and record the ARC reference for the test in your report. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

Enough to be safe and useful: how the equipment gets its address (static vs DHCP), what a subnet is, how to ping a device, and how to check a link light before blaming the equipment. Get graphics terminals, gateways and remote monitoring onto documented addresses with the site's IT involved — undocumented static IPs cause outages years later. Never change network settings on shared infrastructure without the client's IT; a fire system that takes down a corporate VLAN is a bad day. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

The mechanical side kills more doors than the electronics: check closer adjustment, alignment and lock engagement before blaming controllers. Verify the door reports its states honestly (open, forced, held), test the fire-alarm release interface where fitted, and confirm standby battery condition. Review the event log for doors generating constant forced/held events — that's usually alignment or user behaviour, and it trains security staff to ignore real alerts. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

They must reflect what was actually installed — device positions, addresses, zone boundaries, cable routes and interfaces — not the design intent from before the site changed everything. Update them at handover and after every modification. On service and fault visits they're the fastest route to a hidden junction box or an unlabelled interface; without them every callout starts with archaeology. Missing or inaccurate as-fitteds are a defect worth reporting in their own right. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

Watch the airflow readings against the commissioning baseline — rising flow faults usually mean a cracked or disconnected pipe; falling flow means blocked sampling holes or filters. Clean or replace filters per the manufacturer's schedule, and investigate any airflow deviation rather than re-normalising it: re-baselining a blocked pipe hides a real loss of coverage. Keep the pipe layout drawing with the site records; fault-finding a hidden pipe network without it is guesswork. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

Every panel, PSU, interface and isolation point should carry a durable label matching the site records — device references that agree with the zone chart, circuit references at the distribution board, and warning labels where isolations affect life-safety. Consistent labelling turns a 40-minute hunt into a 2-minute check, and it's what lets a relief engineer work the site safely. Replace missing labels as you find them and note systematic gaps in the service report. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

You need four numbers: standby (quiescent) current, full alarm current, required standby duration and required alarm duration — all from the panel documentation, device datasheets or measured at the battery terminals. Capacity = (standby current × standby hours) + (alarm current × alarm hours), then add the safety margin the design standard and panel manufacturer call for (commonly 25%) and round up to the next standard battery size. The classic mistake is using only the panel's quiescent figure and forgetting the loop devices. Confirm durations against the project specification — don't assume 24 hours is always right. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

Analytics are only as good as the scene: mounting height, angle, lighting and a clean lens decide whether detection rules work. Commission analytics against agreed test scenarios (walk tests at the required distances) and record the results; re-verify after any camera move, refocus or firmware update. Watch false-trigger rates in the VMS — an analytic that cries wolf gets ignored, which is operationally worse than no analytic at all. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

Work from the site's approved cause-and-effect matrix — not from memory. Trigger each input condition (or simulate it where live release is unsafe), and confirm every mapped output: sounders, plant shutdowns, door releases, dampers, AOVs, ARC signals. Record each input-output pair as pass/fail with times. Anything that can release suppression or move plant needs isolations agreed in advance and restored afterwards, with the restoration witnessed and recorded. An unversioned or missing matrix is itself a defect worth reporting. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

A working demonstration, the documentation pack (certificates, as-fitted drawings, zone chart, cause-and-effect, operation and maintenance instructions), the log book started with commissioning entries, and named training for the people who will actually use the panel — silencing, resetting, testing, and what to do on fire and fault. Record who was trained and when. A handover the client didn't understand generates a call-out within the month. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

Enough for the next engineer to understand the system without you: as-fitted drawings, zone chart, cause-and-effect matrix, device schedule with addresses and locations, loop test results, battery calculations, sound-level results, soak-test outcome, outstanding variations, and the commissioning certificate. Version everything and leave copies on site and in the office file. Half of the difficult service visits trace back to commissioning information that was never handed over. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

Change default passwords on anything network-connected, restrict remote access to documented, authorised routes, keep firmware current from official sources, and record who holds credentials. Segregate life-safety equipment from general networks where the design allows. Treat engineer access codes as site-specific and confidential — never publish or share defaults. Most compromises of building systems start with a default password nobody changed. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

Judge each defect by its effect on the system's ability to detect and warn: anything that stops detection or alarm in an occupied area, or takes the panel or signalling down, needs immediate action and formal notification to the responsible person. Degraded redundancy, single-device faults and non-urgent wear items follow, with agreed timescales. Put the priority and the consequence in writing — clients act on "this means the second floor has no automatic detection" far faster than on a code. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

Optical chambers accumulate dust, insects, steam residue and decorating debris; the analogue value creeps up until the panel flags contamination or pre-alarm. Check the environment first — a detector near an air-handling vent or in a workshop will foul repeatedly no matter how often it's swapped. Clean or replace per the manufacturer's guidance (many heads are exchange-only), record the analogue values before and after, and note repeat offenders: a location that fouls every few months needs a detection-type review, not another clean. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

Duplicate addresses after replacements, addresses that don't match the zone chart, and soft-addressed devices reprogrammed in the wrong order are the usual failures. After any device change, verify the address against the device schedule and run the panel's device list to check for duplicates or missing entries. Update the schedule and zone chart the same day — an accurate paper trail is the difference between a five-minute fix and an afternoon of head-scratching on the next callout. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

Call each outstation to the master and back, checking speech is intelligible in both directions — a line that rings but is unintelligible is a fail. Verify the labelling matches the location, the master handset shows the correct identity for each call, battery backup is healthy, and any ancillary indicator lamps work. Log every outstation tested; on larger sites rotate coverage so all are proven across the maintenance cycle. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

Trigger the alarm condition mapped to the doors and watch them: every held-open door on the interface should release and close fully — a door that releases but doesn't latch hasn't done its job. Confirm re-set behaviour, check local release buttons still work, and verify the interface's monitored wiring reports faults to the panel. Coordinate with the site before testing: releasing every fire door in a busy corridor unannounced doesn't make friends. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

Confirm which loop or circuit the panel is flagging, then disconnect it at the panel — if the fault clears, the problem is in the field wiring or a device, not the panel. Test the disconnected cable with an insulation-resistance tester, conductor-to-earth on each core, then split the circuit at convenient points (isolators, junction boxes) and re-test each half to close in on the fault. Damp devices, crushed cable, and screws through hidden runs are the usual culprits. Record the readings, not just "fault cleared" — the megger values are your evidence. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

Simulate mains failure for the rated duration and inspect every luminaire at the end — not the start — of the test, because failing batteries drop out late. Record failures by location, replace or repair, and re-test what was fixed. Schedule full-duration tests for times when the building can tolerate discharged batteries afterwards (they need recharge time before they can protect again). Short functional tests in between confirm operation but don't prove duration. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

Escalate when the fault affects life-safety cover you can't restore that visit, when the fix needs parts, permissions or specialist knowledge you don't have, when manufacturer support is needed for panel firmware or configuration, or when what you're finding contradicts the site records. Escalation is a competence indicator, not a failure: record what you verified, what remains unknown, and who now owns the next action, and make sure the responsible person understands any interim impairment. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

Start with the event log, not opinions: identify which devices, times and conditions generate the unwanted alarms. Then match the fix to the cause — wrong detector type for the environment (steam, dust, cooking), detector siting too close to vents or kitchens, transient activities without temporary management, or genuine faults presenting as alarms. Multi-sensor or heat detection, relocation, cause-and-effect delays where the standard permits them, and client housekeeping all have their place. Change one thing at a time and monitor the log after each change so you know what actually worked. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

Know the version on every panel you maintain and record it in the site file. Update only using the manufacturer's official releases, tools and procedure, with the configuration backed up first and the client aware the system will be impaired during the work. Read the release notes — sometimes an intermittent fault you're chasing is already fixed in a later version, and sometimes an update changes behaviour the site relies on. Never load firmware from unofficial sources. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

The graded framework matches the system's resistance to the anticipated intruder skill level — higher grades demand more supervision, tamper security and signalling resilience. The grade is set by the risk assessment and, where relevant, insurer requirements, and it drives component selection and signalling: you can't mix a low-grade signalling path with a high-grade requirement and stay compliant. Check the current standard's edition and the insurer's spec before quoting or modifying a graded system. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

Measured at the panel and at points around the loop, the loop voltage shows whether devices are actually receiving what the panel thinks it is delivering. A healthy loop shows close to the panel's output at the far end; a significant drop points to high-resistance joints, undersized or damaged cable, or an overloaded loop. Compare readings against the manufacturer's stated limits for the protocol, and record where you measured — a voltage figure without a location is useless on the next visit. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

Confirm whether the supply is actually missing (measure at the panel terminals) or whether the panel's charger/monitoring is the problem. Check the dedicated circuit at the distribution board, the isolator, and the fuse in the panel's mains terminal block. If the supply is healthy but the fault persists, look at the PSU and charger side. While the mains is off, remember the system is running on batteries — note the time and don't leave site with the standby period already eaten into without telling the client. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

Use the manufacturer's test key so the frangible element isn't broken, operate the point, and confirm the panel shows the correct address and text and that the required outputs operate. Rotate through different call points on the routine weekly test rather than using the same convenient one every week, and cover every call point across the service programme. Record which points were tested — "weekly test done" without the device reference doesn't demonstrate coverage. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

Establish which node(s) dropped and when, from the event logs of more than one panel — one side of the story misleads. Check the network topology (ring or bus), because a single break in a ring may show as degraded rather than down. Inspect the network wiring and connectors at the affected node first, then the node's network card. After repair, confirm every panel sees every other and that global cause-and-effect still operates across nodes — a network that looks up but doesn't pass alarms is worse than one that is obviously down. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

Identify the boundary: the last device that still answers and the first one that doesn't — the break sits between them. On a Class A loop the panel may keep talking to everything via the return leg, so use the panel's loop diagnostics to see which direction is lost. Check the boundary device bases and every junction box in that cable run before condemning cable. A device removed without a continuity plug, a loose terminal, or another trade's drill bit account for most of these. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

Survey what exists: loop protocols and device compatibility, cause-and-effect logic, interfaces to other systems (doors, dampers, plant, ARC), network arrangements, and the state of the as-fitted records. Decide what is being reused and verify the new panel genuinely supports it — protocol compatibility claims deserve testing, not trust. Plan the changeover window, temporary cover while the system is down, retesting scope, and handover documentation. The panel swap is the easy part; the interfaces are where jobs overrun. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

Service effort is weighted by the risk profile of the site and the system's history rather than treated as a uniform checklist: a sleeping-risk premises or a site with recurring faults gets more attention than a low-occupancy unit with a clean record. The maintenance regime still has to meet the recommendations of the applicable standard and the site's fire strategy — risk-based means prioritising within that framework, not cutting visits below it. Document the reasoning so the regime can be defended. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

What you found, what you tested, what you changed, what remains outstanding, and what the client should do next — with device references and readings, not vague phrases. "Fault cleared" is only acceptable when the report shows the checks that prove it. Record impairments and isolations explicitly, get outstanding defects in writing to the responsible person, and keep the site history so recurring problems are visible across visits rather than rediscovered each time. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

Loop isolators have sectioned off a short circuit between the loop conductors so the rest of the loop keeps working. The panel shows which section is isolated — that's your search area. Disconnect devices in that section one at a time until the short clears, and inspect for crushed cable, moisture in bases, or a trapped conductor. After repair, confirm the isolators have reset and every device in the section is back online before leaving site. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

Walk the whole building, not the drawings: record construction, ceiling heights and voids, occupancy and sleeping risk, environmental challenges (dust, steam, humidity), existing equipment and its condition, cable routes, and anything that will affect detector choice or coverage. Photograph the panel, a sample of devices, and problem areas. Talk to whoever manages the building — the operational reality they describe is the difference between a design that works and one that false-alarms for a year. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

Prove the chain: alarm input → control output → damper movement → position feedback to the panel or BMS. Witness the damper actually travel where access allows, don't just trust the status LED. Check the fail-safe state matches the design (open or closed varies by strategy), and that damper faults report somewhere someone watches. Record which dampers were exercised — on large sites rotate through them across the maintenance programme with the ventilation contractor where needed. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

Add up the alarm-state current of every sounder, beacon and VAD on the circuit from the datasheets — then compare against the circuit's rated output with the manufacturer's derating applied. Measure the actual current in full alarm where practicable; installations drift from their paperwork. An overloaded circuit may work on a bench-fresh PSU and fail on batteries at the end of the standby period, which is precisely when it matters. Fix by re-balancing circuits, not by hoping. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

Before touching anything, isolate the releasing circuit or fit the actuation blocking devices per the manufacturer's procedure, and prove the isolation — an accidental discharge can be lethal and is always expensive. Agree the isolation with the responsible person, tag it, and record it. Test detection and control logic with the release physically prevented, then restore and verify the system is back to automatic with the isolation removed. Never leave site with a release circuit isolated and nobody formally told. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

Keep the readings and the references: which devices were tested, measured values (insulation resistance, battery, sound levels, analogue values), times, and results. Photos of meter readings and panel screens cost nothing and settle arguments later. Evidence is what separates a defensible service history from a signature on a sheet — if an incident is ever investigated, "tested — OK" with no data will not stand up. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

Radio conditions decide whether a wireless system is viable in that building, on that day, with those contents — and steelwork, stock levels and refurbishments change them. Survey with the manufacturer's tool before design, record signal margins per device location, and re-check when the building use changes or when devices start reporting communication faults. Fit devices at surveyed positions, not convenient ones; a device moved half a metre behind a steel beam can drop off the network. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.

It must let a stranger — a responding fire officer or a relief engineer — find any device quickly: clear building orientation, zone boundaries that match reality, device locations and addresses, and a revision date. Update it whenever devices are added, moved or re-addressed. A zone chart that no longer matches the installation is worse than none, because people trust it. Flag an out-of-date chart as a defect in your service report. Standards are referenced for orientation only — always confirm requirements against the current edition of the applicable standard, the site's fire strategy and manufacturer documentation.