Fire Alarm System Failure Fire Watch in Seattle

The First Hour After an Alarm Failure

A fire alarm failure is a compliance emergency with a very short fuse. The building does not have to be evacuated in most cases, but it does have to be protected — and that protection has to be documented from the moment the system went down. Work the first hour in sequence, and do not wait for the alarm contractor to finish diagnosing before you start: confirm the scope of the outage (the whole system, or only a zone, device, or notification circuit), note the exact time the impairment began because that is the start of your log, notify the fire department using the impaired system reporting process, dispatch fire watch coverage so patrols begin immediately rather than after the repair is scheduled, notify occupants that the alarm is out of service and how they will be warned in the interim, and keep egress paths clear and unlocked while detection is offline.

Partial failures still count. Managers frequently assume a single failed zone or a panel in trouble does not require a fire watch; if the impaired portion covers occupied space, it generally does. The safer position is to cover the affected area and document it — the cost of a few hours of fire watch is trivially small next to a violation notice.

Restoring service is also a documented event. The fire watch does not end when a technician says the panel looks good — it ends when the system is restored, tested, and confirmed back in service, and that time belongs in the log alongside the start time.

What a Fire Alarm Failure Means for the Building

A fire alarm failure means the building has lost automatic detection, occupant notification, or both, and the code trigger under IFC 901.7 is the same regardless of what caused it: a panel in trouble or off-normal condition, a communications fault to the monitoring station, a batch of devices pulled offline for testing, or a wholesale system replacement mid-project. Without working detection and notification, occupants have no automated warning of a fire, so the fire watch exists to be the human substitute — eyes that would otherwise belong to smoke detectors and ears that would otherwise belong to horns and strobes.

This is different from a sprinkler or standpipe impairment because there is no water-based backup to lean on; if the alarm system is down and sprinklers are also out, the building has no automatic protection at all and most AHJs will treat that combination as grounds for a tighter interval or, in high-occupancy buildings, a conversation about whether occupancy should continue.

What the AHJ Expects on an Alarm Impairment

Fire marshals want to know the scope of the outage in device terms, not just system terms — is the entire panel down, or only a zone, a floor, or a device type such as duct detectors or pull stations? A partial outage may allow a narrower watch focused on the affected zone, while a full panel failure or monitoring communications loss typically requires watch coverage of the entire occupied building. The AHJ will also expect the monitoring company's central station to be notified of the outage so they do not treat every phone call from the building as a false alarm, and expect a manual means of notification — an air horn, a phone tree, or a staffed intercom — to be available for the Officer to use if a fire is discovered.

Where the failure results from ongoing testing or device replacement rather than an outright fault, the AHJ typically wants the contractor's scope and expected restoration date attached to the fire watch notification so they can judge whether the timeline is reasonable.

Patrol Interval and Route Without Detection

Because there is no automatic detection, the Officer's patrol route has to substitute for the devices that are down. That means walking areas a smoke detector would normally cover — mechanical rooms, electrical rooms, storage areas, and concealed spaces — not just the corridors a visual sweep would naturally follow. Intervals commonly run 15 to 30 minutes, with the tighter end used in sleeping occupancies, healthcare, and any building where the notification appliances themselves (not just detection) are down, since occupants there have no way to hear a warning even if the Officer discovers a fire immediately.

Route planning also has to account for how the Officer will notify occupants if smoke or fire is found, since the normal horns and strobes will not activate. We confirm before the first patrol whether the building has a manual fire alarm capability, a public address system, or whether the plan relies on the Officer physically alerting occupants floor by floor, and we build that into the response plan the Officer carries.

Logging and Restoring Detection Service

Alarm impairment logs record which zones or devices were out, the patrol areas covered each round, and any activity near mechanical, electrical, or storage spaces where a fire could start undetected. Because central station monitoring is often affected too, the log also notes when the monitoring company was notified of the outage and when normal monitoring resumed, since that is a separate restoration point from the physical repair.

Restoration is confirmed by a functional test of the repaired devices and a communications test to the central station, both witnessed or reported by the alarm contractor, with the AHJ's acceptance where the jurisdiction requires it for that class of repair. The most common citation on alarm-related watches is stopping the watch when repairs are physically complete but before the communications path to the monitoring company has actually been verified — we hold coverage until both are confirmed.