Fire safety for London care homes and supported living

Not at the time, and this is worth understanding properly because it is being used to sell people systems they may not need. Heat detection in sleeping rooms was widely specified for years, largely to avoid unwanted alarms. BS 5839-1:2025 changed the expectation: smoke, multi-sensor or carbon monoxide detection is now preferred in sleeping rooms unless the fire risk assessment justifies heat. That is a meaningful change, because a heat detector only responds once a fire is already producing significant heat, which in a room where somebody is asleep is late. It is not retrospective, so an existing certificated system does not become non-compliant, but it is exactly the kind of thing a review of your assessment should now be looking at.

It means moving residents sideways into an adjoining fire compartment rather than out of the building, which buys time for people who cannot use stairs quickly or at all. Whether it applies is a question for your fire strategy and your fire risk assessment rather than for us, but the thing worth knowing is what it depends on. It only works if the compartmentation actually holds, which means the fire doors, the walls above the ceiling and every service penetration through them. It also only works if there are enough staff on the floor at the time to move people. A strategy that assumes six staff and a building that runs on three at four in the morning is not a strategy.

A staff alarm arrangement, where an activation alerts staff before any general evacuation signal, is a recognised option in this sector and is used where a full evacuation signal would cause more harm than good. It is not a default and it is not something to specify from a catalogue: it has to be justified in the fire risk assessment, it depends on staff being present and trained to respond within a defined time, and it changes what the system has to be able to do. We will build it where the assessment calls for it and we will say so if the assessment does not.

In this sector they matter more than almost anything else on the system, because the evacuation strategy is usually built on the assumption that a compartment will hold for a defined period. A wedged door, a failed self-closer, a gap where a cable was pulled through a wall, and that assumption stops being true while everybody still believes it. This is also why hold-open devices linked to the alarm are worth the money here: they let a corridor stay open for care, and close on activation, without anybody having to remember.

Two service visits a year under BS 5839-1 by a competent person, plus the weekly test the responsible person carries out and records. The 2025 edition added a tolerance of plus or minus one month on the six-monthly interval, so five to seven months between visits is inside the standard rather than a lapse, which is useful when a visit falls awkwardly. Emergency lighting needs its monthly function test and annual full-duration test under BS 5266-1, and since the 2025 edition a photometric verification at least every five years.