Skip to content
CodeGreen ITM Blog
Inspection and testingFire alarmNFPA 72 Ch. 14

Smoke detector sensitivity testing: measured, not sprayed

Canned smoke proves a detector responds. It does not prove it is in calibration. Here is the schedule, and how the longer interval is earned and lost.

Two technicians reviewing test readings on a rugged laptop at a fire alarm panel
Reviewed by Joe Montuori, NICET Level IV October 4, 2026 · Updated October 23, 2026 · 4 min read

Two different tests get called the same thing. A functional test proves a smoke detector responds to smoke. A sensitivity test proves it responds at the right level. NFPA 72 Chapter 14 requires both, and they are not interchangeable.

Sensitivity has to be measured

Blowing an unmeasured aerosol at a detector proves it responds to smoke, not that it is in calibration. Sensitivity is checked against the detector's listed and marked sensitivity range. Where a detector carries no marked range, the reference is 4 percent obscuration light grey smoke.

From the field

Canned smoke is a functional test, not a sensitivity test. The two get confused constantly.

The schedule

StageWhen sensitivity is checked
New installationWithin the first year after installation.
After thatEvery alternate year.
After two consecutive passing calibration testsThe interval may be stretched to a maximum of 5 years.
Any zone where nuisance alarms increased over the previous yearCalibration tests are performed regardless.

This schedule does not apply to one and two family dwellings.

The 5 year interval is earned, and it can be lost

The longer interval is not a choice. It is earned by two consecutive calibration tests showing the device inside its listed and marked range. Once it is in use, records of nuisance alarms and the trends in them have to be maintained. In any zone or area where nuisance alarms show any increase over the previous year, calibration testing comes straight back.

That means the nuisance alarm log is part of the testing program. A site on the extended interval with no nuisance alarm records is not meeting the condition that comes with it.

Detectors you cannot safely reach

Devices that are inaccessible for safety reasons, such as continuous process operations, energized electrical equipment, radiation or excessive height, may be tested during scheduled shutdowns if the AHJ approves. The extended interval cannot exceed 18 months.

The test is safety, not convenience. A detector that is merely awkward to reach does not qualify, and the AHJ approves the arrangement, not the contractor.

Duct detectors

Under NFPA 72 Chapters 14 and 17, a duct detector has to be accessible for cleaning, with access doors or panels provided for the purpose, and the location of every duct detector is permanently identified and recorded.

From the field

A duct detector 30 feet up over a mechanical room with no access is a finding waiting to happen.

Who does the work

Under NFPA 72 Chapters 10 and 14, service personnel are expected to understand the alarm code and the electrical code, the equipment, and the test methods, and to be able to read the system documentation. This is the clause behind certification requirements in most jurisdictions.

Get new articles by email

One email when a new article goes up. Confirm once, unsubscribe any time.

Keep reading

Drawing: corridor plan with a pull station within 5 feet of each exit door and the 200 foot travel distance to the nearest one
Fire alarmInspection and testingNFPA 72 Ch. 17

Manual pull stations: placement, height and travel distance

Measure to the handle, put one by every exit, then walk the floor to check nobody is more than 200 feet from one.

October 7, 2026 · 4 min read
A fire department connection marked AUTO SPKR on the outside wall of a building
Fire department accessSprinklerNFPA 1 Ch. 18

Fire department access: the address, the road, the hydrant, the key box and the FDC

Everything inside the building depends on the fire department getting to it. Five things to check from the street before you walk in.

October 10, 2026 · 4 min read
Two technicians checking horn strobes and notification appliances on a test wall
Fire alarmNFPA 72 Ch. 18

Fire alarm audibility: sound levels, ambient measurement and the evacuation signal

A horn passes or fails against the room it is in, not its data sheet. Measure the ambient first, then check the margin, the ceiling and the signal itself.

October 13, 2026 · 5 min read
Drawing: sprinkler riser with a waterflow switch and a valve supervisory switch wired to the fire alarm control unit
SprinklerFire alarmNFPA 72 Ch. 17

Waterflow and supervisory signals from the sprinkler system

The 90 seconds is the device, not the horns. What the waterflow and supervisory devices have to do, and how many go on one circuit.

October 1, 2026 · 5 min read
A pair of white doors with closers and panic hardware in a dark hallway
Fire doorsInspection and testingNFPA 80 Ch. 5

Fire doors: self closing, latching and what actually voids the label

A door that closes but does not latch has lost its rating. Not every hole drilled on site has, and the listing laboratory is who decides.

September 28, 2026 · 4 min read
Horn strobes, a speaker strobe and pull stations mounted on a test wall
Fire alarmNFPA 72 Ch. 18

Strobe mounting height and candela: how NFPA 72 sizes visible notification

Measure the lens, size the room correctly, and never drop a strobe below 80 inches without redoing the coverage.

September 25, 2026 · 5 min read