Leak Detection

Checkline Europe's Technical Support 10 Jul 2026

Leak detection is the work of finding escaping gas, air, vacuum or pressure loss before it turns into wasted energy, unsafe plant or unplanned downtime. Its defining fact is that a leak you cannot hear is rarely silent: an escaping flow radiates sound well above the range of human hearing, which is why one method family — ultrasonic detection — dominates practical plant leak work, from compressed-air audits to valve, steam-trap and enclosure testing.

A leak survey is not the same thing as a leak-rate measurement. The instrument reports a signal shaped by the pressure across the leak, its geometry, the distance and angle of the probe, and the ultrasonic background around it — not a calibrated volume per unit time. Keeping that distinction clear is what separates fast, repeatable location work from the quantified leak testing a tight specification eventually demands, and it decides when an ultrasonic survey has answered the question and when it has only narrowed the search.

1. Why a Leak You Cannot Hear Still Makes a Sound

When gas escapes through a restriction, the turbulence at the opening generates sound across a wide band that reaches well into the 20–100 kHz ultrasonic region. Human hearing falls away around 20 kHz, so the leak can be effectively silent to someone standing beside it, while a detector tuned to that band picks it up cleanly, converts it to an audible cue by heterodyne conversion, and shows a signal level for comparison. Because ultrasound has a short wavelength it stays directional and localises well, so an operator can separate one leaking fitting from its neighbour in a dense plant run.

The same physics is used in three modes, and they should not be blurred. In airborne mode the leak radiates straight into open air, so the probe scans exposed fittings, couplings and accessible seals. In contact, or structure-borne, mode the useful signal travels through metal and is picked up by touching the part — the way valves, steam traps and bearings are checked. And in the transmitter-tone method an ultrasonic source is placed inside a sealed space so an operator can scan the outside for escaping sound — the way an enclosure with no pressure of its own, such as a cabinet, hatch or vehicle cabin, is tested. Recognising which mode a job needs is the first practical decision in leak detection.

2. Which survey mode does the job need?

Ultrasonic detection carries almost all practical plant leak work. Ultrasonic Leak Detection covers the airborne, contact and transmitter-tone modes together with probe and accessory choice, survey technique and the reporting discipline that keeps repeat surveys comparable — the method behind compressed-air cost-of-leakage audits, hydraulic and pneumatic fault-finding, refrigeration and HVAC screening, valve and steam-trap checks, and automotive weather-seal and enclosure testing. What that page does not claim to do is quantify a leak: ultrasound locates and ranks, and where a specification fixes a very small leak rate, tracer-gas (helium) sniffing, pressure-decay or vacuum-decay testing, and bubble confirmation take over as the follow-up step. Those are complementary confirmation methods, chosen by how small the target leak is rather than as separate instrument routes, so an ultrasonic survey is where the work both begins and is documented.

And when the question becomes which detector to buy, the leak detection selection guide lays out the choice.

3. Standards for Ultrasonic Leak Surveys

Leak-detection standards matter because they define what "tight enough" means in a given context, and they guard equally against over-buying for a simple compressed-air survey and under-specifying when a result has to satisfy an auditor. ASTM E1002 is the reference for ultrasonic gas leak detection and the survey conventions behind most compressed-air and gas inspection, and EN 1779 ties the choice of method to the leak-tightness criterion the job actually needs. Those confirmation methods — tracer-gas, pressure-decay and bubble testing — carry their own standards in turn. But what really holds an ultrasonic survey together is procedure: sweep the background before starting, record the pressure, temperature and running state, function-check the transmitter and receiver before any enclosure test, and report against a documented threshold rather than reading a sound level as though it were a leak rate.

4. Frequently Asked Questions

1. How do I compare this year's compressed-air survey against last year's?

Record the things that make surveys comparable: the same reference threshold on the instrument, the running pressure and load, and a fixed route so the same points are checked each time. An ultrasonic level is relative, not an absolute leak rate, so a survey's value over time comes from repeating the conditions, not from the raw reading. Tag each logged leak to its location and give it a severity rank, and the next survey shows plainly what was fixed and what is new.

2. Why does my compressed-air survey miss leaks behind insulation?

Lagging absorbs and scatters the airborne signal, so a strong leak on bare pipe can become a weak, directionless trace once it is insulated. Use contact inspection at exposed fittings and valves, and treat an insulated run as suspect only when the process evidence supports it rather than guessing through the lagging.

3. When should I use transmitter-tone mode instead of airborne mode?

Use it when there is no pressure-driven leak to hear — vehicle cabins, electrical cabinets, hatches, refrigerated containers and clean-room doors. A tone source inside the sealed space lets you scan from outside, but function-check the transmitter and receiver together first, because a frequency mismatch can make a leaking seal look sound.

4. Does a higher reading always mean a larger leak?

No. The signal level depends as much on how close and how square the probe is to the leak, and on the background around it, as on the size of the leak itself — so a louder trace is not automatically a bigger one. Treat the reading as a way to rank repairs, not to size a leak, unless a separate leak-rate method has actually measured it.

5. Do I need traceable calibration on a leak detector?

For informal fault-finding, often not. But for a report that has to stand up to an energy audit, a safety review or a customer acceptance test, you need a documented function check and a calibration approach — ideally against a reference leak of known rate — so the record is more than the operator's impression on the day.

5. Glossary

Airborne ultrasound	High-frequency sound in the 20–100 kHz region radiated through air by an active leak, above the range of human hearing.
Contact (structure-borne) mode	Ultrasonic inspection through a waveguide or probe pressed against a component, used when the useful signal travels through the structure rather than into open air.
Transmitter-tone method	A technique in which an ultrasonic source is placed inside a sealed enclosure and a receiver scans the outside for escaping sound, used where there is no pressure differential to drive a leak.
Heterodyne conversion	The process of shifting received ultrasound down into an audible signal the operator can hear while a level is displayed.
Pressure differential	The difference in pressure that drives flow through a leak path; without it, airborne and bubble methods may show nothing.
Leak rate	The quantity of gas or fluid passing through a defect per unit time, commonly expressed in mbar L/s or standard cubic centimetres per second.
Localisation	Narrowing the search from a general area to the specific fitting, seal, valve or component producing the signal.
Tracer gas	A marker gas, commonly helium, used to confirm very small leaks under sensitive methods when an ultrasonic survey cannot prove the rate.
Calibrated leak	A reference artefact with a known leak rate, used to check instrument response and support traceability of a survey result.