Smoke Detectors
Photoelectric, Ionization & Beyond
The two sensing technologies, where each one works best, spacing rules from NFPA 72, and why getting this wrong is the number-one cause of nuisance alarms.
Two Technologies, Very Different Strengths
Photoelectric
A light source (usually an LED) shines across a chamber. A photodiode sits at 90° to the beam. In clean air, no light reaches the photodiode. When smoke particles enter, they scatter light into the sensor. The controller reads the scatter signal and triggers alarm when it crosses threshold.
Best at: smoldering fires, soft furnishings, slow-burning fabrics, electrical insulation. Common in bedrooms, hallways, healthcare.
Ionization
A tiny radioactive source (americium-241) ionizes air molecules inside a sealed chamber, creating a small measurable current between two plates. Smoke particles attach to the ions and reduce the current. The controller reads the current drop and triggers alarm.
Best at: fast-flaming fires (paper, wood, gasoline). Historically common in residential; less common in new commercial installations due to nuisance-alarm rates.
▶ Watch: How Do Smoke Detectors Work?
Courtesy of AMRE Supply · Open on YouTube ↑
Spacing Rules
NFPA 72 §17.7 sets spot-type smoke detector spacing for smooth, level ceilings up to 10 feet:
- Maximum 30 feet between detectors (listed spacing for most photoelectric detectors).
- Maximum 15 feet from a wall (half of listed spacing).
- No more than 900 sq ft of coverage per detector on a smooth flat ceiling.
- Beams, waffle ceilings, joists, and high ceilings trigger specific adjustments in §17.7.3.
- For corridors less than 15 feet wide: 41 feet maximum between detectors, 20.5 feet from ends.
Where airflow exceeds one air change per minute, the listed spacing must be reduced. High-airflow environments — data centers, air-handling units, laboratory fume hoods — often require either a tighter spot-type spacing or a dedicated air-sampling (VESDA-style) detector.
Placement Don'ts
- Within 3 feet of an HVAC supply register (wind blows smoke away from the detector).
- Within 3 feet of a bathroom door (steam causes nuisance alarms).
- In the apex of a peaked ceiling deeper than 4 inches without special listing (dead air pocket).
- In kitchens within 20 feet of cooking equipment (a heat detector is typically used instead).
- In unconditioned attics or outdoor locations (not listed for environment).
Sensitivity Testing
NFPA 72 §14.4.5 requires the sensitivity of every smoke detector to be tested after the first year in service and every two years thereafter. If sensitivity has remained within the listed range on two successive tests, the interval can be extended to every five years until sensitivity begins to drift. Sensitivity is measured by aerosol test method or by the panel reading for addressable systems with on-board self-diagnostics.
The listed sensitivity window for most spot-type photoelectric detectors is 0.5% to 4.0% obscuration per foot — a detector reading outside that window (either too sensitive or too insensitive) must be cleaned, recalibrated, or replaced. Addressable systems report the raw analog value continuously, so a sensitivity audit from the panel event log takes minutes instead of the hours required for aerosol testing on a conventional system.
Aspirating (Air-Sampling) Detectors
In high-value, high-airflow, or very-early-warning applications, a spot-type detector is inadequate. Aspirating smoke detection (ASD) — best known by the VESDA (Xtralis / Honeywell) brand name — uses a small network of sampling pipes with precisely drilled holes to continuously pull air from the protected space back to a centralized laser detection chamber. The chamber can detect smoke at concentrations 0.001% obscuration per foot and below — 1,000× more sensitive than a spot detector.
Typical ASD applications: data centers (where a single smoke event can destroy millions of dollars of equipment), museums and archives, clean rooms, high-ceiling warehouses, telecommunications facilities. NFPA 72 §17.7.3.4 sets the listing and installation rules. Each sampling hole represents the equivalent of one spot detector for spacing purposes, and the system must be commissioned with a smoke-transport-time measurement (typically <90 seconds) to verify proper airflow.
Beam Detectors (Projected Beam)
For very large open spaces — atria, gymnasiums, warehouses, arenas, sanctuary ceilings — spot-type detectors at 30-foot spacing are impractical. A projected-beam smoke detector uses an infrared transmitter and receiver at opposite ends of the space, up to 330 feet apart, monitoring for obscuration of the beam. Smoke anywhere along the beam path causes signal loss; the detector alarms on sustained obscuration.
Installation rules in NFPA 72 §17.7.3.5: beam spacing maximum 60 ft (perpendicular), within 0.5–10% of ceiling height. The transmitter and receiver are factory-paired; replacement requires recalibrating the pair. Structural movement (seasonal beam deflection, building sway) can misalign a beam detector — an annual alignment check is part of the ITM.
Drift Compensation & Pre-Alarm
A dusty photoelectric chamber reports higher obscuration even with no smoke present — this is drift, and it's a leading cause of nuisance alarms on older systems. Analog-addressable panels continuously monitor each detector's baseline signal. When the baseline creeps upward (dirt accumulating in the chamber), the panel automatically adjusts the alarm threshold upward to match. This is drift compensation, introduced in the 1990s and now standard on every major brand.
The catch: drift compensation only works until the detector approaches the upper limit of its listed sensitivity range. Once it hits that limit, the panel generates a maintenance trouble or pre-alarm signal — essentially telling the facility “this detector will nuisance-alarm soon, clean or replace it.” A facility that ignores pre-alarm signals for months is one cooking event away from a false evacuation.
Related feature: day/night sensitivity. Addressable panels can schedule lower thresholds during unoccupied hours (catching smoldering fires earlier when nobody is watching) and higher thresholds during occupied hours (reducing nuisance alarms from coffee burns and perfume). The schedule is programmed into the panel and tied to occupancy type.
Kitchen Strategy
Kitchens are the single biggest nuisance-alarm source in commercial fire alarm systems. The NFPA 72 rules:
- No spot-type smoke detector within 20 ft of a cooking appliance (NFPA 72 §17.7.4.2.3).
- Use a heat detector in the kitchen — 135°F fixed-temperature or rate-of-rise.
- Use photoelectric smokes outside the 20-ft buffer (in the corridor or dining area). If the buffer cannot be achieved, specify detectors listed to the 2019 UL 268 8th Edition, which adds cooking-nuisance immunity testing.
- For commercial hoods, the hood's own UL 300 wet-chemical suppression system handles the hood/duct. Fire alarm duct detectors monitor the HVAC supply, not the hood itself.
Residential kitchens (apartment units) follow a slightly different rule set per NFPA 72 §29: photoelectric smokes required at least 10 ft from the cooking range, or 6 ft with cooking-nuisance-immunity listing. Ionization detectors are increasingly prohibited in residential installations near kitchens because of their historically poor nuisance-alarm performance.
Duct Smoke Detectors vs Area Smoke Detectors
A duct smoke detector is a specialized photoelectric sensor mounted in a housing that clamps onto HVAC ductwork. A sampling tube projects into the duct, and air is diverted through the detector chamber via Venturi action. NFPA 72 §17.7.5 and NFPA 90A §6.4 require duct detectors on HVAC systems moving more than 2,000 CFM, primarily to shut down fans and prevent smoke distribution — not as a substitute for area smoke detection.
The distinction matters because duct detectors respond slowly (they depend on airflow) and can miss a developing fire in an unoccupied mechanical room. Every occupied space still requires area smoke detection per the occupancy rules in NFPA 101 regardless of whether that area is served by ducts with duct detectors. See the dedicated Duct Detectors article for installation specifics.
Frequently Asked Questions
What is the difference between photoelectric and ionization smoke detectors?
Does NFPA 72 require photoelectric specifically?
How long do smoke detectors last?
Can I put a smoke detector in a kitchen?
What is an aspirating (air-sampling) smoke detector?
How often must sensitivity be tested?
References
1. NFPA 72 (2022), §17.7 — Smoke detector spacing and placement.
2. NFPA 72 (2022), §14.4.5 — Sensitivity testing intervals.
3. UL 268 — Smoke Detectors for Fire Alarm Signaling Systems (2019 revision added polyurethane foam test and cooking nuisance immunity).
Open the discussion panel to comment, flag an inaccuracy, add field experience, or ask a question. Approved contributions earn SRP and may be incorporated into the article.