Fire Alarm Audibility Test
Measuring Sound Coverage
Audibility is the part of fire alarm performance that inspection reports most often mark as failed or needing investigation. Here is what NFPA 72 §18.4 actually requires, how to measure it correctly, and the common design traps — closed doors, sound-absorbing surfaces, noisy ambient conditions — that make a compliant system fail in real use.
Why Audibility Is the #1 Failed Line on FA Inspections
The Northside Duluth outpatient inspection report surfaces three sound-test failures in the Discrepancy Report: 1st Imaging Area, 1st OR Surgery Area, 1st Surgery Area / Pre-op / Endo Areas. This is not unusual. On a typical healthcare or commercial fire alarm inspection, sound-test failures and special notes account for 30–50% of all recorded discrepancies. Visual (strobe) tests rarely fail; audibility fails frequently.
Audibility is brittle because the audible performance of a NAC depends on variables that were not present when the system was commissioned: new walls, new doors, new HVAC equipment running louder than planned, new sound-absorbing surfaces like acoustic ceiling tile, and occupant density. A system that measured 75 dBA in an empty surgery suite at commissioning can measure 60 dBA after the suite is in daily use.
The NFPA 72 Audibility Rules
Public Mode (§18.4.3)
The evacuation signal must be at least one of the following, whichever is greater:
- 15 dBA above the average ambient sound level having a duration of 60 seconds or more, measured 5 ft (1.5 m) above the floor, OR
- 5 dBA above the maximum sound level having a duration of at least 60 seconds.
In a quiet office (45 dBA ambient), the 15-dBA rule drives the design: alarm must exceed 60 dBA throughout. In a mechanical room running at 85 dBA ambient, the 5-dBA-above-max rule drives: alarm must exceed 90 dBA at the ambient peak.
Private Mode (§18.4.4)
Only 10 dBA above ambient (instead of 15). Used in signaling to trained personnel only — e.g. at a constantly attended nurses station or security desk, not for general occupant notification.
Sleeping Areas (§18.4.5)
At the pillow of every occupied sleeping position, 75 dBA minimum. Since the 2013 edition, a 520 Hz low-frequency waveform is also required in rooms where hearing-impaired occupants may sleep — hotels, multifamily housing, dormitories. The low-frequency signal is more effective than the traditional high-frequency horn at waking people with age-related hearing loss or sensorineural impairment.
Temporal-3 Pattern
NFPA 72 §18.4.2 requires the evacuation signal to use the Temporal-3 pattern: 500 ms on, 500 ms off, 500 ms on, 500 ms off, 500 ms on, 1.5 s off, repeat. This is the ISO 8201 standardized evacuation tone. A system producing any other pattern (continuous, slow-whoop, etc.) for evacuation is non-compliant.
How to Measure
- Equipment. Type 2 dBA meter (Extech 407730, Quest 1600, Brüel & Kjær 2240), calibrated within the last 12 months, A-weighting, slow response. Tripod if possible for repeatability.
- Ambient baseline. With the system silent, record ambient at every measurement point. Note the highest sustained (≥60 sec) level. This sets your rule threshold.
- Measurement height. 5 ft (1.5 m) above finished floor. Not at ceiling, not at the speaker — at occupant ear level.
- Measurement points. Every room at least once. Large rooms at multiple points, including corners and behind doors. Sleeping rooms at the pillow.
- Measurement duration. Activate the system, let it run at least 30 seconds, read the average. For peak-driven rooms (mechanical, industrial), measure during ambient peak conditions — not off-hours.
- Doors closed or open. Match normal operating conditions. Hospital patient rooms with doors normally closed must be tested with doors closed.
- Record. Document at every point: ambient level, alarm level, delta, pass/fail against the greater-of rule.
Common Failure Modes
- Door absorption. Room measures 75 dBA with the door open, 55 dBA with it closed. Solution: appliance inside the room, not in the corridor.
- Ceiling tile absorption. High-NRC acoustic tile eats high-frequency horn output. Speakers at or below ceiling perform better than horns mounted on ceiling in rooms with absorptive ceilings.
- Ambient creep. New equipment added after commissioning raises ambient by 10+ dBA; system falls below the 15-dB delta without the appliances changing. Solution: add appliance capacity or up-size.
- Voltage drop. Horn at end of a long NAC run is operating at 20 V instead of 24 V; output dBA is 4–6 dB lower than rated. Solution: voltage-drop analysis and potentially a booster mid-run.
- Occupant density. A lecture hall empty vs full can swing 8–10 dBA. Solution: test with occupancy representative of worst case.
- Silenced while occupied. Staff silences alarm for nuisance trip and forgets to reactivate; next-shift inspection finds the system partially out of service.
ITM Frequency
Per NFPA 72 Table 14.4.5:
- Annual functional test — every notification appliance must sound, and technician verifies audibility subjectively or by measurement in each protected area.
- Periodic dB audit (new requirement, 2022 edition) — measured dBA survey recommended at least every 10 years or after major renovation that changes room acoustics.
- Acceptance testing — at initial commissioning and after any system extension, full measured audibility survey documented.
References
1. NFPA 72 (2022), §18.4 — Audible Characteristics.
2. NFPA 72 (2022), §18.4.5 — Sleeping-area audibility.
3. ISO 8201 — Acoustic — Audible emergency evacuation signal.
4. ANSI/ASA S12.2 — Criteria for Evaluating Room Noise.
Was this article helpful?
Rate this article to help us improve
Discussion (2)
Biggest miss I see is doors. A room with the door closed can be 20 dBA quieter than the same room with the door open. If your occupancy relies on doors being closed at night (hotels, hospitals, dorms), your audibility survey MUST be done with doors closed — not propped open for the convenience of the test.
We ran into a gym where the speakers measured fine in an empty room but failed at 20 percent occupancy because the audience absorbed the sound. Private colleges with assembly occupancies are not exempt from audibility — design for the room full, not empty. Add 3-5 dBA margin beyond the minimum code requirement.