8.2 Three-Year Air-Leakage Options and Annual Air-Pressure Supervisory Testing

Two permitted three-year air-leakage methods

NFPA 25 (2023), 13.4.5.2.9, provides two methods for the dry-system air-leakage test at least every three years. The first uses a 40 psi test pressure for two hours, with no more than 3 psi loss. The second uses normal system air pressure with the air supply shut off for four hours, during which the low-air supervisory alarm must not operate. Do not combine the starting pressure from one method with the duration and acceptance criterion from the other. Identify the selected method in the test record.

The second option is not permission to disable the low-air alarm so that a leaking system appears to pass. Verify that the supervisory function is operational and that the normal pressure is correct for the installed valve. Apply the impairment and notification procedures appropriate to the test. Arrange staffing and access so that an unexpected pressure decline or valve trip can be handled safely.

Conduct and record a defensible test

Use the actual valve procedure and instruments appropriate to the selected method. Avoid opening covers, adjusting regulators, or isolating unrelated equipment without understanding the arrangement. Isolate replenishment so the compressor or generator cannot hide leakage. Record system identification, date, method, starting and ending pressures, elapsed time, relevant temperatures, and alarm observations. Restore the normal supply path afterward and confirm the correct valve positions and normal pressure.

For the two-hour method, a start of 40.0 psi and finish of 37.4 psi give a loss of 2.6 psi, within the 3 psi limit. A finish of 36.7 psi gives 3.3 psi loss and fails that criterion. Do not round 3.3 down to 3 merely because a coarse installed gauge has large divisions. Use suitable test instrumentation and document the actual readings. A successful leakage test does not demonstrate hydraulic capacity or replace a trip test.

For the four-hour method, record whether the low-air signal operates at any time during the interval. A signal after three hours is a failed result even if someone restarts the compressor before the fourth-hour observation. Testing must preserve the event history. Frequent compressor cycling in normal service also merits attention even when a particular test result is acceptable.

Temperature can change pressure

At approximately constant volume, ideal-gas absolute pressure varies with absolute temperature. Use gauge pressure plus local atmospheric pressure and temperature in kelvin or degrees Rankine when explaining that relationship. At sea level, 40 psig corresponds to about 54.7 psia. At 70°F, the absolute temperature is about 529.7°R. A 10°F cooling change would produce approximately 54.7 × 10 / 529.7 = 1.03 psi pressure reduction in an ideal sealed volume.

That calculation explains a possible influence, not permission to subtract an assumed allowance and override the prescribed pass criterion. Actual gas and pipe temperatures can differ from a room thermometer. Let conditions stabilize, document changes, and repeat or evaluate an invalid or inconclusive test through the proper procedure. Heating can mask a leak as cooling can imitate one. Do not create a false pass by selecting only favorable readings.

Find the leak without creating another hazard

Inspect accessible trim, connections, couplings, and equipment using compatible leak-detection fluid or other appropriate instruments. Ultrasonic detection can help locate escaping gas in a noisy facility, but range, background ultrasound, access, and instrument skill affect performance. It is not immune to all noise and cannot guarantee detection of every concealed leak from floor level. Confirm the suspected location before repair.

Do not increase pressure beyond the prescribed test or component rating to make a leak louder. Do not inject arbitrary tracer gases or chemicals into a listed system. Refer specialized methods to qualified personnel using an approved procedure. Isolate and depressurize the affected equipment for repair, manage any impairment, and retest under the selected criterion. A nitrogen system also presents oxygen-displacement hazards in enclosed spaces.

Annual dry-system air-pressure supervisory testing

NFPA 25 (2023), 13.4.5.2.6, requires annual testing of dry-system air-pressure supervisory devices. Do not substitute the quarterly provision applicable to a different system type. The low-air signal must give warning before the dry valve operates. The actual setting and high-pressure supervision, where provided, follow the applicable installation requirements, listed equipment, and approved arrangement; this guide does not impose one pressure offset on every valve.

Coordinate the test with the receiving station, reduce pressure through the prescribed means, and record the actual pressure and signal received. Stop before an unintended trip where the procedure calls for it. Check the correct system identification, restoration signal, and receiving-station status. A local switch click alone does not prove that the intended remote indication was received. Restore normal pressure and all isolations, and notify the responsible parties of completion or any unresolved defect.

Source checked October 19, 2026: NFPA 25 (2023), 13.4.5.2.6 and 13.4.5.2.9.