Compressed air escaping through a quarter-inch fitting produces an ultrasonic signature loud enough to detect from across a noisy production floor, yet completely inaudible to a walking technician relying on their own ears. That gap between what leaks actually cost and what plants can hear is why compressed air systems routinely waste twenty to thirty percent of generated volume through leaks nobody has found yet. Ultrasonic detection closes that gap by picking up the high-frequency turbulence of gas escaping under pressure, the same physical principle that also reveals a steam trap that has failed open or a bearing developing friction long before it generates measurable heat. Facilities ready to quantify their own leak load can book a demo to see ultrasonic survey data mapped against their compressed air and steam systems.
PREDICTIVE MAINTENANCE · ULTRASONIC INSPECTION · 2026
Hear the Leaks Your Plant Can't
Ultrasonic detection for compressed air leaks, steam trap failures, and early bearing faults, quantified and mapped across the whole facility.
What Unfound Leaks Are Actually Costing
20-30%
Of generated compressed air volume is typically lost to leaks in an unmanaged system
$1-8K
Annual energy cost of a single quarter-inch compressed air leak running continuously
15-30%
Of steam traps in a typical plant are failed and leaking live steam at any given time without a testing program
40kHz+
Typical ultrasonic frequency range used to detect turbulent gas flow and early friction, well above human hearing
How an Ultrasonic Leak Survey Actually Runs
A structured survey moves through the same sequence every time, converting inaudible ultrasonic signals into a ranked list of leak points with an estimated cost attached to each one.
1
Airborne ultrasound is scanned across fittings, valves, and connections while the system is pressurized under normal operating conditions.
→
2
Each detected signal is heterodyned down to an audible frequency and logged with its location and relative intensity.
→
3
Intensity readings are converted into an estimated flow rate and annualized energy cost for each individual leak point.
→
4
Leaks are tagged in priority order so repair crews fix the highest-cost points first instead of working randomly.
Ultrasonic Detection vs Traditional Methods
| Detection Method | Works in Noisy Areas | Typical Limitation |
|---|---|---|
| Ultrasonic Detection | Yes, unaffected by ambient plant noise | Requires direct or near line-of-sight to the leak point |
| Soap Solution Testing | Yes, but slow across large systems | Impractical for surveying an entire facility regularly |
| Manual Listening | No, masked by machinery noise | Only catches the largest, loudest leaks |
| Pressure Decay Testing | Yes, but isolates a zone rather than a single point | Identifies that a leak exists, not exactly where |
Three Applications, One Detection Principle
Compressed Air Leaks
Fittings, quick connects, hoses, and valve packing generate a continuous ultrasonic hiss the moment pressure escapes, detectable from several feet away even in a loud production area.
Steam Trap Failures
A trap that has failed open produces a distinct sustained ultrasonic flow signature, easily distinguished from the intermittent cycling pattern of a properly functioning trap.
Early Bearing Faults
Rising friction inside a bearing generates ultrasonic energy well before it produces measurable heat or a vibration-detectable amplitude change, giving an earlier warning window on slow-speed equipment in particular.
Why Ultrasonic Detection Outperforms the Ear in a Plant Environment
The human ear is genuinely sensitive to sound in the range of roughly twenty hertz to twenty kilohertz, and the turbulent flow generated by gas escaping through a small orifice under pressure produces most of its energy well above that upper limit, typically concentrated between thirty and one hundred kilohertz depending on the size and shape of the leak path. This means that even in a perfectly silent room, a meaningful portion of a compressed air leak's acoustic signature would remain completely inaudible to an unaided technician, and in a working production environment filled with machinery noise across the audible spectrum, any low-frequency portion of the leak signature that might otherwise be heard gets masked entirely. Ultrasonic instruments solve both problems simultaneously by listening specifically in the frequency band the leak actually dominates, then heterodyning that signal down into an audible range the technician can interpret through headphones.
Directionality is the second major advantage ultrasonic instruments bring over unaided listening. High-frequency sound waves travel in much more direct, narrow paths than lower-frequency sound, which means a technician sweeping an ultrasonic sensor across a suspect area can pinpoint the exact location of a leak within inches, even among a cluster of closely spaced fittings, valves, and connections. Lower-frequency audible sound tends to reflect and diffuse throughout an enclosed space, making it far harder to trace back to a precise origin point using hearing alone, especially in the kind of steel-and-concrete industrial environment most manufacturing facilities operate within.
The same directional sensitivity that makes ultrasonic detection effective for finding leaks also makes it valuable for confirming a repair actually worked. Rather than relying on a visual inspection or an assumption that tightening a fitting solved the problem, a technician can return to the exact point with the same instrument and confirm the ultrasonic signature has genuinely dropped to background levels, closing the loop on the repair with objective data rather than a visual check that cannot detect a leak that has simply become quieter without fully stopping.
Steam trap testing benefits from this same directional precision in a slightly different way. A functioning steam trap cycles between an open and closed state as it discharges condensate, producing an intermittent ultrasonic pattern, while a trap that has failed open produces a continuous signature and a trap that has failed closed produces silence even though condensate is backing up behind it. Distinguishing these three states reliably requires listening precisely at the trap body itself, since ambient steam line noise can otherwise obscure the difference between a healthy cycling pattern and a genuinely failed trap, particularly on smaller trap sizes where the acoustic difference between states is more subtle.
Managed Leak Program vs No Detection Program
Regular Ultrasonic Surveys
Leaks found and cost-ranked on a recurring schedule
Steam trap failures caught before energy loss compounds
Compressed air system operates near its designed efficiency
Leaks accumulate silently across the whole distribution system
No Detection Program
Failed steam traps waste energy for months undetected
Compressor capacity is consumed compensating for leak load
Running a Survey That Actually Gets Leaks Fixed
The gap between finding leaks and actually fixing them is where many otherwise well-run ultrasonic programs lose most of their potential value. It is common for a facility's first survey to identify well over a hundred individual leak points across a large compressed air distribution system, and without a clear prioritization and ownership structure, that volume of findings can overwhelm a maintenance team already managing a full work order backlog. Structuring survey results by estimated annual cost rather than presenting a flat list in the order they were found gives maintenance planners a natural way to triage, typically fixing the top ten to twenty percent of findings by cost first, since that segment usually accounts for the large majority of the total leak load in dollar terms.
Tagging convention matters more than most facilities expect on the first survey. A leak point identified only by a verbal description or a rough sketch location is far harder for a repair crew to relocate weeks later than one marked with a physical tag showing an identification number that ties directly back to the survey report, its estimated cost, and its priority ranking. Facilities running recurring surveys benefit from maintaining a persistent leak point numbering system across survey cycles, since this allows year-over-year comparison of whether the same physical locations keep reappearing, often pointing to a systemic issue like a fitting type or installation practice that keeps failing rather than a series of unrelated random leaks.
Steam trap surveys benefit from a parallel tracking discipline, since a trap population across a large facility can easily number in the hundreds, and without individual trap identification tied to a maintenance history, it becomes difficult to distinguish a trap that has failed for the first time from one that has failed repeatedly and might be better served by a different trap type or size for its specific application. Facilities that combine trap identification with basic application data, orientation, condensate load, and operating pressure, generate far more useful long-term data than those tracking pass or fail status alone, since that additional context often explains why certain traps in certain applications fail at a meaningfully higher rate than others across the same steam system.
Common Mistakes That Undermine an Ultrasonic Program
Surveying Only Once a Year
Treating a leak survey as an annual audit rather than an ongoing program allows new leaks to accumulate silently for months between visits, eroding much of the savings the previous survey achieved.
No Repair Verification Step
Assuming a visual repair fixed the leak without a follow-up ultrasonic check leaves facilities unable to confirm whether the fix actually worked or the leak simply became quieter without fully stopping.
Flat, Unprioritized Findings Lists
Handing maintenance teams a long list of leak locations with no cost ranking overwhelms already-stretched crews and often results in only the easiest, not the costliest, leaks getting fixed first.
No Persistent Tagging System
Identifying leaks only by verbal description rather than a permanent tag number tied to survey history makes it difficult to track repeat failures at the same location across multiple survey cycles.
Where Leak Detection Priorities Shift by System Type
Compressed air, steam, and rotating equipment each present a different leak or fault profile, and matching survey frequency and technique to the specific system type improves both detection accuracy and the return on the time spent surveying.
Compressed Air Distribution
Quick-connect fittings, hose assemblies, and point-of-use valves accumulate the largest share of leak points over time, making these connection types the natural starting focus for a first survey.
Steam Distribution Systems
Trap populations at process heat exchangers and tracing lines tend to fail at a higher rate than traps on simple drip legs, warranting closer attention during a structured testing cycle.
Slow-Speed Rotating Equipment
Bearings on slow-speed equipment generate weaker vibration signatures at the same fault stage compared to high-speed machines, making ultrasonic friction detection a particularly valuable early-warning layer here.
Facility Teams Ask
How is a leak's dollar cost actually calculated from an ultrasonic reading?
The ultrasonic intensity reading at a fixed distance from the leak point correlates to an estimated orifice size and flow rate, which is then combined with the facility's known compressed air generation cost per unit volume to produce an annualized dollar figure for that specific leak. This calculation lets a survey rank dozens or hundreds of leaks found across a facility by actual cost rather than by how loud or accessible each one happens to be. Facilities can walk through this cost-ranking methodology through support.
How often should a compressed air and steam trap survey be repeated?
Most facilities run a full ultrasonic survey annually at minimum, with high-leak-rate systems or plants with aggressive energy reduction targets moving to semi-annual or quarterly surveys until the leak rate stabilizes at a low baseline. Steam trap populations in particular benefit from more frequent testing, since trap failure rates climb steadily with age and a plant that has never tested its traps often finds a much higher failure percentage on the first survey than on subsequent ones. Facilities can plan a survey cadence suited to their system size by reaching out through support.
Can ultrasonic detection find leaks that are too small to hear even with the equipment?
Ultrasonic instruments are specifically built to detect turbulent flow at frequencies well above the range of human hearing, which means many of the leaks they catch would never be found by ear even if the surrounding plant were completely silent. Sensitivity and detection distance scale with leak size, so very small leaks require the technician to work closer to the suspected point, but the underlying physical signature exists at nearly any leak size once pressure differential is present.
Why does ultrasonic detection work for bearing faults if it was designed for gas leaks?
Both applications rely on the same underlying physics, since friction between two surfaces and turbulent gas escaping through an orifice both generate broadband ultrasonic energy, just from different physical mechanisms. A bearing beginning to lose lubrication or develop surface wear produces a rising ultrasonic amplitude that a trained ear or trending software can pick up well before the friction generates enough heat to register on a thermal scan. This is why some reliability programs use the same handheld ultrasonic instrument for both leak surveys and routine bearing route checks. Teams can book a demo to see both applications running on the same platform.
What happens after a leak or failed steam trap is identified in a survey?
Every finding is tagged with its location, estimated cost, and a repair priority ranking, which feeds directly into the maintenance work order queue so the highest-cost leaks get scheduled first rather than whichever one a technician happens to notice. Repair verification is typically done with a follow-up ultrasonic check at the same point to confirm the fix actually resolved the signature rather than assuming a visual repair was successful. Facilities can review a sample survey-to-repair workflow through support.
