# Leak testing

Selecting leak-test methods, defining gas and pressure conditions, interpreting leak-rate units, and verifying system sensitivity.

Reviewed: 2026-11-01

Tags: inspection, nondestructive-testing, leakage, pressure, vacuum

## Define the leakage requirement first

Specify the medium, direction, upstream and downstream pressures, temperature, duration and leak limit. A pressure change needs the isolated volume and elapsed time to support a gas-throughput estimate.

**Illustrative calculation:** assume the complete isolated test volume is 2.00 L, temperature and volume remain constant, and a 0.060 mbar pressure decrease over 60 s is wholly due to outward leakage. Then q = 2.00 × 0.060 / 60 = **0.0020 mbar·L/s**, equivalent to **0.00020 Pa·m³/s**. The conversion follows from 1 mbar = 100 Pa and 1 L = 0.001 m³. These invented values demonstrate units; they do not establish an acceptance limit or detection capability.

Pfeiffer's **13 August 2026** explanation emphasizes that leakage rate depends on the fluid and operating conditions. Gas quantity is commonly represented as pressure-volume throughput per time, using units such as **mbar·L/s** or **Pa·m³/s**. A pressure change alone cannot be compared across different enclosed volumes. [Pfeiffer: understanding leak rates](https://www.pfeiffervacuum.com/au/en/knowledge/understanding-leak-rates.html)

Distinguish a containment test from a pressure-strength or proof test. A leak result at one pressure does not establish structural capacity or demonstrate that a seal behaves identically under reversed pressure. Use the governing product or engineering specification for both the test conditions and acceptance criterion.

## Choose the method and the result it provides

| Method | Observation | Principal interpretation question |
| --- | --- | --- |
| Bubble or immersion check | Visible bubble formation at accessible locations | Were the specified pressure, observation time, surface condition, and locations covered? |
| Pressure or vacuum decay | Pressure change over a timed interval | Can leakage be separated from thermal, volume, and fixture effects? |
| Tracer-gas sniffing | Escaping tracer sampled near a location | Was the probe exposed to enough of the escaping gas for long enough? |
| Vacuum tracer-gas test | Tracer entering an evacuated component or chamber | Does the pumping arrangement deliver a representative signal within the test time? |
| Integral collection or accumulation | Combined leakage into a defined enclosure | Are collection efficiency, mixing, background, and timing controlled? |

Pfeiffer distinguishes **local** leak finding from **integral** measurement of overall leakage and recommends considering the pressure gradient encountered in service when selecting the method. A vacuum test and an externally pressurized sniffer test reverse the tracer's direction through the component. Their results need the corresponding conditions. [Pfeiffer: tracer-gas test methods](https://www.pfeiffervacuum.com/no/en/knowledge/vacuum-technology/knowledge-book/7-leak-detection/7.2-leak-detection-with-tracer-gases.html)

**ISO 20485:2017**, published November 2017, covers leak detection using tracer gas and a tracer-specific detector. Its scope does not supply a universal leak-rate limit for all products. [ISO: tracer-gas method scope](https://www.iso.org/standard/68190.html)

## Pressure-decay controls and a worked example

ATEQ describes a pressure-decay cycle as preparation/filling, stabilization, measurement, and venting. Stabilization matters because pressure and temperature fluctuations can influence the result. The complete isolated circuit includes the part, connections, fixture seals, and associated volume, so validate the setup as well as the component. [ATEQ: pressure and vacuum decay testing](https://www.ateq-leaktesting.com/products/leak-tester/)

For an ideal rigid enclosed volume at constant temperature, the magnitude of pressure-volume loss rate can be estimated as **q = V × |Δp| / Δt**. Apply this only when the measured pressure change represents gas loss rather than thermal or mechanical change. Pfeiffer gives the corresponding volume-based relationship for pressure-rise testing. [Pfeiffer: pressure-volume rate relationship](https://www.pfeiffervacuum.com/au/en/knowledge/understanding-leak-rates.html)

If the same pressure change occurs because the gas cools, the calculation would falsely attribute it to leakage. Establish an adequate cycle on representative parts, including production temperature variation and fixture operation. Report the assumed or measured volume and pressure conditions.

## Tracer concentration, background, and response

Tracer concentration belongs in the setup. Do not compare a diluted-gas response with a pure-tracer limit unless the instrument and procedure apply the appropriate correction. INFICON's helium-use guidance explains that diluted helium requires concentration control and suitable detector sensitivity. It also cautions that lowering fill pressure can change the leak itself: some paths open at higher pressure. A simple mathematical pressure correction may then be misleading. [INFICON: reducing helium consumption](https://www.inficon.com/en/news/2023-12-at-how-to-save-helium-in-vacuum-leak-testing)

For sniffing, define scan speed, probe distance, access, and dwell from a demonstrated procedure. Air movement, tracer distribution, and background affect the gas reaching the detector. For vacuum or integral testing, check the worst response time in the actual connection and pumping arrangement. A short instrument response measured at its inlet does not demonstrate the same response through a large component or restrictive line.

Pfeiffer describes a calibrated test leak as a known gas rate under defined conditions and notes that external test-leak location and auxiliary pumping affect system response. Use a reference arrangement representative of the inspection path. Internal calibration alone does not verify every fixture, hose, valve, or sampling location. [Pfeiffer: calibrated leaks and system response](https://www.pfeiffervacuum.com/no/en/knowledge/vacuum-technology/knowledge-book/7-leak-detection/7.2-leak-detection-with-tracer-gases.html)

## A bounded detector example

INFICON's undated **UL1000/UL1000 Fab** product page lists minimum detectable helium rates of **<5 × 10⁻¹² mbar·L/s in vacuum mode** and **<5 × 10⁻⁸ mbar·L/s in sniffer mode**. It separately lists maximum inlet pressures of **15 mbar in GROSS**, **2 mbar in FINE**, and **0.4 mbar in ULTRA** modes. These are mode-specific detector specifications, not a guarantee that a production installation achieves those limits or a pressure prescription for the test component. [INFICON: UL1000 specifications](https://www.inficon.com/en/products/leak-detectors/ul1000-fab)

Demonstrate the complete system's ability to discriminate acceptable and reject conditions at the required cycle time. Include background, recovery after a large leak, fixture leakage, repeatability, and reference-leak uncertainty. A fine display or low catalog minimum is insufficient by itself.

## Gas results versus service-fluid leakage

COSMO Instruments explains that idealized gas/liquid comparisons assume a smooth capillary, while real leakage paths are irregular. Surface tension, contamination, and liquid remaining in a path can change or suppress later gas leakage; deformable seal paths can also change with pressure. Consequently, there is no universal conversion from a measured helium or air rate to “waterproof” service performance. [COSMO: gas and liquid leak comparisons](https://www.cosmo-k.co.jp/en/leak-test/gases-liquids/)

Validate any correlation against representative parts, leak mechanisms, media, and pressure/temperature conditions. Keep permeability through a material, leakage through a discrete path, and fixture or background contributions separate where the test objective requires that distinction. An integral signal may combine more than one contribution.

## Reporting and acceptance

Record component/fixture identity, method and procedure revision, medium and tracer fraction, pressures and their absolute/gauge basis, temperature, test volume where relevant, stabilization and measurement time, background, reference-leak identification and checks, measured rate and units, inspected coverage, and acceptance criterion.

Report a bounded conclusion such as meeting the stated limit under the recorded conditions. Investigate failed system checks before accepting affected results. Establish pressure, vacuum, gas-handling, and depressurization arrangements from the approved equipment and component procedure rather than deriving them from this overview.

## Related reference

[Inspection and quality documents](../../sourcing/documents/index.md)