17.2 Periodic Inspection, Testing, and Maintenance Frequencies

Once the system is in service, Module 5.II and Module 5.V both expect you to know that inspection, testing, and maintenance (ITM) is a program, not a single annual visit. This Independent OpenExamPrep section is the frequency half. Chapter 18 is how you troubleshoot a failure during that visit. Chapter 19 is maintenance contracts and false-alarm reporting. The Uniform Code points to NFPA 72, National Fire Alarm and Signaling Code; the frequencies below are commonly taught NFPA 72 Table 14.3.1 (visual) and Table 14.4.3 / 14.4.3.2 (testing) practice. They are not a reprint of the copyrighted tables and not New York-invented intervals. Always confirm the adopted edition and any authority having jurisdiction (AHJ) or occupancy rule that is more frequent.

Three verbs you must not mix

Inspection is a visual look: damage, obstruction, missing covers, corrosion, lamps, labeling, and whether the device is still where the as-built says it is. Testing physically exercises a function: smoke entry, heat, pull, flow, sound, flash, relay transfer, communicator signal. Maintenance is the work that keeps or restores operation — cleaning, adjustment, replacement — following the manufacturer’s published instructions and site conditions. NFPA 72 does not give one universal “maintenance every X months” table the way it gives inspection and testing tables.

The building owner (or the owner’s designated representative) is responsible for seeing that ITM happens. In New York, a firm that services or maintains the system generally needs the Article 6-D license (Chapter 2). Qualified people perform the work; a trainee can assist under the licensee’s supervision, but the record still has to show who did what (Chapter 17.3).

Visual inspection — frequency is by device, not by slogan

Students look for “inspect the fire alarm yearly” and miss the table structure. Commonly taught Table 14.3.1 practice is that frequency depends on the component, whether the system is monitored off-premises, and what the AHJ has adopted.

Examples installers are commonly taught (label these as Table 14.3.1 practice, then verify the edition):

• Control equipment visual inspection is often semiannual.
• Initiating devices (smokes, heats, manual stations, and similar) visual inspection is often annual in course materials; some later-edition rows use semiannual for groups of initiating devices — the idea to remember is device-type row, not a single building-wide slogan.
• Batteries are commonly taught as monthly or quarterly visual depending on chemistry (vented lead-acid is typically more frequent than sealed). Look for corrosion, leakage, swelling, date codes, and connection tightness. Do not treat a New York classroom number as a substitute for the adopted table row.
• Waterflow and supervisory (tamper) devices: older course notes often said quarterly visual; the 2022 edition of NFPA 72 is widely taught as moving those visuals to semiannual. New York’s 2025 Uniform Code points to NFPA 72—22, but the Department of State exam bulletin names the Uniform Code and NFPA standards rather than reprinting edition years. If a question names an edition, use that edition’s row.
• Unmonitored control-unit fuses, lamps/LEDs, primary power, and trouble signals are commonly taught as weekly visual checks; when those conditions are supervised off-premises, the visual interval for those indicators is often much less frequent (monthly or annual in many table rows). The teaching point: monitoring changes the visual calendar for those status checks.

That table is a study aid, not a substitute for the copyrighted NFPA tables.

Functional testing — what installers must actually operate

Smoke detectors. Commonly taught practice is an annual functional test using a listed method (listed aerosol or equivalent — not a blast of unknown dust). Sensitivity must remain within the listed and marked range. A typical course sequence is: calibrated sensitivity within 1 year after installation, then every alternate year, with a possible extension to a longer interval (commonly taught up to 5 years) after consecutive in-range results — if the adopted edition allows it. Detectors found outside range are cleaned, recalibrated if listed for that, or replaced.

Listed self-test exception (commonly taught): not every detector needs a technician with a sensitivity meter every cycle if the unit (or the fire alarm control unit (FACU) method) provides a listed calibrated sensitivity test, including some automatic self-test methods. Canned smoke that merely makes the LED blink is not a calibrated sensitivity test. If the listing does not include that method, you still owe the sensitivity check on the Chapter 14 schedule.

Heat detectors. Restorable types (rate-of-rise, restorable combination, restorable fixed-temperature) are functionally tested with a listed heat source on the testing calendar (commonly annual). Non-restorable fixed-temperature detectors are destroyed by a heat test. Commonly taught practice is to replace them 15 years from the date of manufacture, unless the adopted edition’s laboratory sampling method is used instead. Do not heat-gun a non-restorable detector “to see if it works.”

Waterflow. Functional testing is commonly semiannual under Table 14.4.3 practice (actual flow or a listed method). Coordinate with the sprinkler contractor and NFPA 25 so you do not double-count or skip the switch.

Manual stations. Commonly annual mechanical operation, then restore and reseal if the station uses a break-glass or similar seal.

Notification appliance circuits (NACs) / audible / visible appliances. Commonly annual functional operation. You are proving the circuit and the appliance, not redesigning candela in the hallway — but a dead strobe is a failed test, not a “next year” item.

Batteries (conservative, labeled practice). Treat batteries as their own row, not as “part of the annual.” Commonly taught NFPA 72 practice includes more frequent visual checks than for smoke detectors — monthly visual appears in many course notes for some chemistries, quarterly for others — plus charger tests and discharge- or load-related tests that are often annual (with additional load-voltage checks on shorter cycles for some types). Sealed lead-acid replacement is typically a manufacturer 3-to-5-year schedule in addition to those NFPA 72 tests; replacement on age does not excuse skipped load tests, and a passing visual does not prove capacity. Record date codes. If a question is edition-specific, do not invent a New York-only month count.

Building a calendar without copying the book

A workable field habit:

1. List every device type on the as-built.
2. Assign visual and test intervals from the adopted Table 14.3.1 and Table 14.4.3 rows.
3. Overlay manufacturer maintenance (cleaning, firmware, battery replacement age).
4. Overlay AHJ or insurance frequencies that are stricter.
5. Notify monitoring before tests; restore afterward (Chapter 19 covers false-alarm fallout if you skip this).

In practice

A school contract sold as “annual fire alarm inspection” that never includes the semiannual waterflow functional test is incomplete under commonly taught Table 14.4.3 practice, even if every smoke detector gets canned smoke each August. The owner still owns the gap; the licensed firm that wrote the contract owns the explanation.

On the exam

Wrong answers collapse ITM into one annual visit, treat canned smoke as sensitivity, heat-test non-restorable detectors, or skip battery load tests because the jars “look fine.” Right answers name device-type frequency, listed methods, and edition/AHJ control.