Compressed air is one of the most expensive utilities a plant buys, and it's also one of the least monitored. A quarter-inch leak in a fitting at 100 PSI can quietly cost thousands of dollars a year in wasted compressor energy, and unlike a water leak, it doesn't leave a puddle for anyone to notice. Most plants only find their air leaks during an annual walk-down survey with a handheld ultrasonic gun, which means a leak that started the day after last year's survey runs undetected — and unpaid for — for up to twelve months. A demo can show what continuous leak detection finds that an annual survey misses.
Compressed Air & Gas Monitoring
Find Compressed Air and Gas Leaks Before the Annual Survey Would
iFactory fuses ultrasonic sensor data with your historian to locate and value leaks continuously, ranking each one by energy cost so repair priorities are obvious.
Why Air Leaks Are the Cheapest Fix Nobody Prioritizes
Compressed air systems are frequently the single largest electricity draw in a plant that doesn't involve a core production process, and studies of typical industrial air systems have repeatedly found that a meaningful share of generated air — often a fifth or more in an unmanaged system — never reaches a productive use point at all. It escapes through worn fittings, cracked hoses, leaking quick-disconnects, and open drain valves. None of that waste shows up as a single dramatic cost line; it shows up as a slightly higher compressor duty cycle that nobody questions because it's always been that way.
The reason leaks persist despite being individually cheap to fix is detection, not economics. Fixing a leaking fitting is usually a five-minute job with a wrench. Finding that specific fitting among the thousands of joints, valves, and connections across a plant's compressed air network is the actual bottleneck, and it's a bottleneck that an annual walk-down survey with a handheld ultrasonic detector only partially solves, since it captures a single point in time and misses everything that develops in between.
How Continuous Ultrasonic Detection Works
01
Fixed Ultrasonic Sensors
Sensors placed at key distribution points listen continuously for the high-frequency signature of turbulent gas escaping through a small opening.
02
Historian Correlation
Ultrasonic readings are cross-referenced against compressor load data from the historian to separate a genuine leak signature from normal demand fluctuation.
03
Leak Localization
Time-series AI narrows the likely leak location using signal strength differences across multiple sensor points in the same zone.
04
Cost-Ranked Alert
Each detected leak is valued in estimated annual energy cost, so maintenance can prioritize the largest leaks first rather than working a flat list.
Leak Detection Coverage: Annual Survey vs Continuous Monitoring
| Aspect | Annual Handheld Survey | Continuous AI Monitoring |
|---|
| Detection frequency | Once per year, typically during a scheduled walk-down | Continuous, around the clock |
| New leak lag time | Up to a full year before discovery | Typically within days of onset |
| Coverage during off-hours | Limited to when the survey team is on-site | Includes nights, weekends, and unmanned shifts |
| Cost valuation | Estimated manually per finding | Automatically calculated from live compressor load data |
| Prioritization | Based on surveyor judgment during the walk | Ranked automatically by estimated annual cost |
Beyond Compressed Air: Process Gas Leaks
Compressed air gets most of the attention because it's the most common application, but the same ultrasonic principle applies to a wide range of process gases moving through valves, flanges, and seals under pressure. A slowly leaking process gas connection produces the same kind of high-frequency turbulence signature as a compressed air leak, which means the same sensor network and detection approach extends to nitrogen purge lines, steam traps, and other pressurized gas systems without requiring an entirely separate monitoring program.
Steam trap monitoring deserves particular mention here, since a failed-open steam trap behaves acoustically much like a gas leak and carries a similarly invisible energy cost — continuously venting live steam that was paid for but never used productively. Plants that treat steam trap surveys the same way they treat compressed air surveys, as an annual walk-down rather than a continuous check, tend to have the same blind spot between survey dates that compressed air systems have.
24/7
continuous listening instead of a once-a-year walk-down survey
$-Valued
every detected leak converted to an estimated annual energy cost automatically
Ranked
repair list ordered by cost impact instead of walk order or convenience
Find Out What's Leaking Right Now
See How Much a Live Leak Survey Would Find
Most plants are surprised by how much has accumulated since the last walk-down. Book a demo and see what continuous monitoring would surface in your system.
Building the Financial Case for Leak Monitoring
1
Pull your compressor run hours and total kWh consumption over the last twelve months as the baseline for what a leak reduction program would offset.
2
Review the findings from your most recent handheld survey, if one exists, and estimate how many months of continued leaking occurred before the next scheduled walk.
3
Calculate the fully loaded electricity cost per unit of compressed air generated at your plant's specific rate structure, since this varies significantly by region and contract.
4
Identify the distribution zones with the oldest piping or highest fitting density as the first candidates for sensor placement, since older infrastructure typically carries the highest leak concentration.
Frequently Asked Questions
How does ultrasonic detection tell a real leak apart from normal system noise?
Turbulent gas escaping through a small opening produces a distinct high-frequency ultrasonic signature that's acoustically different from the lower-frequency mechanical noise of compressors, motors, and normal plant activity. Cross-referencing the ultrasonic reading against historian data on compressor load and demand further helps confirm whether a signal correlates with an actual constant leak rather than a transient operational event.
How many sensors does a typical plant need for adequate coverage?
It depends on the size and layout of the distribution network, but most plants start with sensors at major distribution points and branch headers rather than trying to cover every individual fitting from day one. Coverage typically expands over time as the highest-value zones are identified from the initial rollout.
Support can help map out a sensor placement plan based on your facility layout.
Can this replace our annual compressed air audit entirely?
Continuous monitoring covers the gap between audits far more effectively than the audits themselves, but many plants keep a periodic manual walk-down as a secondary check, particularly for zones without fixed sensor coverage. The two approaches work well together, with continuous monitoring catching what develops between scheduled audits.
Does the cost valuation account for changing electricity rates?
Yes, the cost estimate is calculated against your plant's actual rate structure and compressor efficiency data rather than a generic industry average, which means the ranked repair list reflects what a leak genuinely costs at your specific facility rather than a rough industry benchmark.
Is this approach useful for steam systems as well as compressed air?
Yes, the same ultrasonic detection principle applies to steam trap failures and other pressurized gas leaks, since a failed-open steam trap produces a comparable acoustic signature to a compressed air leak. Many plants extend the same sensor network to cover steam distribution once compressed air monitoring is established.
A demo can walk through how steam trap monitoring fits into the same platform.
Why Leaks Reappear Even After a Clean Survey
A compressed air system that passed a clean survey last quarter isn't guaranteed to stay leak-free, because the mechanisms that create leaks in the first place don't stop operating just because a survey happened. Thermal cycling loosens fittings over time, vibration from nearby rotating equipment works quick-disconnects gradually looser, and normal wear degrades seals and gaskets on a continuous basis rather than on a schedule that happens to line up with an annual audit. This is precisely why a point-in-time survey, however thorough, only ever captures the leak population that existed on the day someone walked the plant with a handheld detector — it says nothing about what develops in the eleven months until the next one.
Plants that have run continuous monitoring for a full year after an initial clean survey are often surprised at how much new leak activity accumulates well before the next scheduled walk-down would have caught it. That gap is the entire value proposition of continuous monitoring in a single sentence: it isn't about finding leaks a survey would have missed forever, it's about finding them the week they start instead of the year they're finally discovered.
What a Ranked Repair List Looks Like in Practice
High-Priority Leak
A large leak on a main distribution header, estimated at a meaningful annual energy cost, flagged for immediate repair scheduling.
Medium-Priority Leak
A moderate leak on a branch line feeding a secondary process, scheduled for the next planned maintenance window.
Low-Priority Leak
A minor leak on a rarely used drop point, logged and tracked but not urgent enough to warrant an unplanned repair visit.
Resolved Leak
A previously flagged leak confirmed closed after repair, with the acoustic signature verified as gone on the next monitoring cycle.
Stop Waiting for Next Year's Survey
Turn Compressed Air Waste Into a Ranked, Fundable Repair List
See how continuous ultrasonic monitoring locates and values leaks the moment they start, not months after.
