Where very early smoke detection makes sense

Very early smoke detection makes sense where a small fire would cause a large loss, such as data rooms, telecom rooms, archives and critical electrical rooms, and where ordinary detectors cannot do their job, such as high-airflow rooms and very tall spaces. It only pays off if someone is ready to act on the early warnings it gives. Without a response plan, it becomes an expensive way to raise the same alarm a little sooner.

What "very early" means

Fires in buildings rarely start as flames. Many electrical fires begin with a component overheating: insulation breaking down, a connection getting hot, a power supply failing. In this incipient stage, tiny particles are released long before there is visible smoke or flame. The stage can last a long time or a very short time, depending on what is burning.

Conventional spot smoke detectors respond once there is enough visible smoke at the ceiling. Heat detectors and sprinklers respond later still, once the fire is producing significant heat. Very early warning systems, most often aspirating smoke detectors, are set sensitive enough to pick up the incipient stage, which gives staff the chance to find and isolate the failing equipment before there is a fire at all.

Spaces where it usually fits

• Data centres, server rooms and telecom rooms. High-value equipment, high airflow that dilutes smoke, and a large cost of downtime. Early warning lets staff find a failing power supply or cable before a fire forces a shutdown or a suppression discharge.
• Critical electrical rooms. Switchgear, UPS and battery rooms, where overheating connections are a known source of fire.
• Archives, museums, galleries and heritage interiors. Contents that cannot be replaced, and often ceilings where visible detectors are unwelcome.
• Clean rooms and laboratories. Very high airflow and processes that cannot tolerate smoke or water.
• Tall and open spaces. Atriums, warehouses and arenas, where the issue is less about sensitivity and more about getting detection where smoke actually goes. Here imaging beam detection is also an option.
• Places that are hard to maintain. Cold storage, secure areas and high ceilings, where a detector at a reachable location sampling through pipes makes testing safer and simpler.

Spaces where it usually does not

Sensitivity is a disadvantage where the normal background contains smoke, dust or fumes. Commercial kitchens, parking garages with vehicle exhaust, woodworking shops, loading docks with diesel trucks and spaces with frequent construction dust tend to produce constant alerts. In ordinary offices, classrooms and apartments, conventional detection designed to code usually does the job well, and very early warning adds cost without a clear benefit.

A simple way to decide

The decision rests on two questions: how much does a small fire or an early shutdown cost in this space, and how well can ordinary detectors work here? The answers place most spaces in one of four groups.

The response plan decides the value

A very early warning is only useful if it starts a response. Before choosing the system, decide what happens at each alarm level, who is notified, and what they do. In practice that means a written plan, agreed with the people who will follow it.

Questions the plan should answer:

• Who receives early warnings during business hours, nights, weekends and holidays?
• How do they receive them: a building automation alarm, a text message, a monitoring company call?
• Are they trained to enter the space safely and allowed to shut down equipment?
• What is the expected time to get someone to the space?
• How are alerts recorded, and who reviews recurring alerts?

A worked example

Take a typical case: a regional office building with a small server room and a separate telecom room on each floor. The server room has a clean agent suppression system and two spot detectors. IT staff are on site during business hours and on call after hours.

The server room is a good candidate. Its cooling units move a lot of air, a fire would cost far more than the equipment, and IT staff can respond. An aspirating detector sampling the room and the cooling unit returns, with early levels going to IT by text and the building automation system, gives them time to find a failing component. The telecom closets are small, lightly loaded and not staffed; conventional detection designed to code is likely enough there, perhaps with temperature monitoring added.

Costs to weigh

The purchase price of the detector is only part of the cost. When comparing options, include:

• Design and commissioning. Pipe modelling, threshold setting and smoke testing take more effort than placing spot detectors.
• Integration. Connections to the fire alarm control unit, plus a path for early warnings to reach staff, whether through the building automation system, a monitoring company or a notification service.
• Maintenance. Filter replacement, airflow checks and transport time testing, on top of the fire alarm system's regular inspection under CAN/ULC-S536.
• Staff time. Someone has to respond to alerts, including the ones that turn out to be harmless.

Against those costs, set what a single incident would cost: equipment, data, downtime, a suppression system refill, cleanup and lost business. In the right spaces, the comparison is not close.

Common mistakes

• Buying sensitivity without a plan. Early levels that go nowhere add nothing.
• Setting thresholds once and forgetting them. A room's background changes when equipment, cooling or use changes, and thresholds should be reviewed with it.
• Using it in dirty spaces. Constant alerts teach people to ignore the system.
• Treating it as a replacement for suppression or good housekeeping. It detects early; it does not prevent or put out fires.

How ehZee can help

Our VESDA and OSID smoke detection service starts by asking why you need early detection and who will respond, then recommends the right approach for each space. Read more in how aspirating smoke detection works, open-area beam smoke detection and data centre cooling.

Common questions