Leaks

Compressed air and steam: the two utilities that leak the most and alarm the least

Compressed air is 10 to 30 percent of plant electricity and leaks waste 20 to 40 percent of it. Failed steam traps can lose 20 percent of boiler output. Neither failure ever trips anything.

9 min read · Updated 2026-08-19 · Milvian Group research

A dense network of stainless steel process pipes

Short answer

Compressed air systems consume 10 to 30 percent of a plant's total electricity and avoidable leaks frequently waste 20 to 40 percent of the air generated, reaching 50 percent of compressor capacity in untreated facilities. On the steam side, roughly 20 percent of steam leaving the boiler can be lost through failed traps without a maintenance programme, with a further 10 to 20 percent lost in distribution. Neither failure produces an alarm, because both systems simply work harder and keep delivering.

The shared pathology

Compressed air and steam are different utilities with the same failure characteristic: they are both distributed pressurised systems whose control loops respond to a leak by producing more. When air leaks, the compressor runs longer to hold pressure. When steam leaks, the boiler fires longer to hold header pressure. The plant never notices, because the plant never runs short.

That is the whole problem in one sentence. The control system is designed to hide exactly the fault you want to find.

Compressed air, the most expensive utility in the building

10 to 30%of plant electricity consumed by compressed airPublished plant assessments
20 to 40%of generated air wasted through avoidable leaksPublished plant assessments
<10%leakage in a well maintained systemCompressed Air Challenge
up to 50%of compressor capacity in untreated facilitiesPublished leak studies

Compressed air is expensive per unit of useful work because the conversion chain is lossy at every stage: electrical energy to mechanical work to pressurised air, most of the input ending as heat, then distribution losses, then the work performed at the tool. Which makes leaks unusually costly, because a leak wastes the entire chain.

Why leak surveys do not solve it permanently

Ultrasonic leak surveys work. A technician walks the plant, tags leaks, and the plant fixes them. Leakage drops sharply. Then it climbs again, because a plant generates new leaks continuously through vibration, thermal cycling, hose wear and fittings disturbed during maintenance.

A survey is a point in time intervention against a continuously regenerating problem. It produces a sawtooth: sharp drops after each survey, steady climbs in between, and an average far above the achievable minimum.

The other compressed air losses worth metering

  • Artificially high system pressure. Set high years ago to overcome a pressure drop somewhere, and never reduced. Every additional bar costs energy across the entire system and increases leak flow through every existing hole.
  • Inappropriate use. Compressed air used for cleaning, cooling or moving product where a blower or a fan would do the same job at a fraction of the energy.
  • Poor part load control. Multiple compressors running unloaded, or sequencing that keeps a large machine trimming when a small one should be.
  • Condensate drains stuck open. A timer drain that fails open passes air continuously and is one of the most common single point losses.

Steam, and the three ways one loss is paid for

Steam losses have been quantified repeatedly and consistently. Approximately 20 percent of steam leaving a boiler plant can be lost through leaking traps in systems without a preventive maintenance programme. Distribution losses waste 10 to 20 percent of generated steam in poorly maintained systems. Steam and condensate leaks can contribute as much as 19 percent of overall plant energy consumption, and typical surface and bottom blowdown runs at about 5 percent of total steam production.

What makes steam distinctive is that a single loss is paid for three separate times. The fuel that generated the steam. The treated water that has to replace the lost condensate. The chemicals to treat that make up water. Most plants account only for the first.

The four thermal measurements that close the loop
MeasurementWhat it revealsWhat it costs you to not have it
Fuel inputEnergy purchased, and boiler efficiency as output over inputBoiler efficiency remains a nameplate figure rather than a measurement
Steam flow at header and major usersWhere thermal energy actually goes across the plantNo attribution, so no unit level thermal intensity
Condensate return ratioTrap and distribution health, as a single high signal numberThe clearest early warning of trap failure is unavailable
Feedwater make up and blowdownThe water side of the thermal loop, and its costThermal loss and water loss appear as unrelated problems

Why these two belong in one platform with water and power

A failed steam trap presents as three separate symptoms in three separate departments: rising fuel consumption to energy, rising make up water volume to utilities, and rising chemical consumption to the water treatment contract. In most plants those three people never compare notes, and each individual movement is small enough to attribute to something else.

On one platform they are a single event with one cause and one work order. That is the actual argument for an integrated utility layer, and it is the reason a compressed air monitoring product and a steam monitoring product bought separately deliver less than the sum of their parts.

How much does a compressed air leak cost?

It depends on hole size, system pressure and running hours, but the system level figures are the more useful frame. Compressed air typically consumes 10 to 30 percent of plant electricity, and avoidable leaks waste 20 to 40 percent of the air generated. In a well maintained system leakage should be under 10 percent.

The reason to think at system level is that individual leak calculations always understate the total, because plants consistently underestimate how many leaks they have.

How do you find compressed air leaks without a survey?

Measure air demand during genuine production stops. Any consumption when nothing is running is leakage, which gives you a continuously tracked leak rate rather than a snapshot.

That does not replace ultrasonic surveys, which are still how you locate individual leaks. It tells you how bad the problem currently is and when a survey is worth commissioning, which is what turns leak management into a managed process.

How much steam is lost through failed traps?

Approximately 20 percent of the steam leaving a boiler plant can be lost through leaking traps in systems without a preventive maintenance programme. Distribution losses waste a further 10 to 20 percent in poorly maintained systems, and steam and condensate leaks can account for as much as 19 percent of overall plant energy consumption.

What is the single most useful steam system metric?

Condensate return ratio. It is a single number that responds to trap failures, distribution leaks and unreturned condensate simultaneously, and it links the thermal loss directly to the water and chemical cost of replacing it.

A sustained fall in return ratio is the most reliable trigger for a trap survey available.

References

  1. Compressed air system leaks fact sheet, Compressed Air Challenge
  2. Finding and fixing leaks, Compressed Air Best Practices
  3. Best practices: steam and condensate leaks, Plant Engineering
  4. Steam and condensate leakage, causes and corrections
  5. How much is steam leakage costing your plant, Thermodyne

Next step

Not the whole campus. One unit, one quarter.

Pick a self contained, high intensity unit with a complete water and thermal story. We instrument it end to end as a lighthouse the rest of the site can see, and the numbers are yours either way.