On a typical industrial site, compressed air is the most expensive utility per unit of useful work delivered, and the one with the lowest measurement coverage. Plants with sub-metered electrical, water and even steam circuits routinely run the compressor house as a black box: one meter on the compressor, no flow meters, no pressure logging, no audit history, and an unexamined assumption that everything downstream of the receiver works as designed.
It almost never does.
The cost of a millimetre
The benchmark figure circulated in industry is that a poorly maintained system loses [RANGE] of generated air to leaks. A system that has never had a leak survey is often worse — we have measured sites at [PERCENTAGE] leak fraction, meaning nearly half the electricity entering the compressor heats the factory and produces nothing useful.
The framing that lands with plant managers is per-aperture cost. A continuous leak from a 3 mm aperture at 7 bar costs roughly [VALUE] per year in electricity at current tariffs with the compressor base-loaded. A 6 mm leak costs four times that. A facility with a dozen identifiable leaks — couplings, a worn filter-regulator-lubricator, a condensate drain bleeding continuously — pays [VALUE] a year for air that performs no work. This is operating expense, every month, on a line the plant cannot see because the compressor is not metered against output.
Why nobody measures it
- It feels free at the point of use. The operator opening a blow-gun has no cost signal in hand; the compressor is in another building.
- It is treated as a utility, not a process. Process equipment gets a maintenance schedule and a performance conversation. The compressor gets an oil change.
- There is no local benchmark. “We run two 75 kW compressors” is a description, not a performance statement, because no specific-energy figure exists to compare against.
What a one-day audit actually does
- Walk the full distribution with an ultrasonic detector during a non-production window, when the only flow is leakage. Tag and estimate every leak.
- Sum the leak rate; multiply by compressor specific energy at the actual operating point, by annual hours, by tariff.
- Log compressor pressure, power and outlet temperature for at least 24 hours, ideally across a weekend. The ratio of unloaded to loaded power is one of the most informative numbers on any compressor.
- Inventory end uses: open blowing, inappropriate substitution where a low-pressure blower would serve, and pressure-band misallocation where one 7 bar network serves work needing 4 bar.
The recoveries
- Leak elimination consistently recovers [RANGE] of compressor energy, often within four to eight weeks because the fixes are torque-and-replace work.
- Pressure optimisation recovers a further [RANGE]; every 1 bar of unnecessary pressure costs roughly [PERCENTAGE] of compressor energy.
- Substitution of inappropriate uses recovers another [RANGE] and typically improves the task it replaced.
- Sequencing controls on multi-compressor installations recover [RANGE] by ensuring one machine modulates while others run fully loaded or off.
The principle. Compressed air is not difficult engineering. It is uninspected engineering. We have closed programmes recovering [PERCENTAGE] of annual compressor energy on capital under [VALUE] — payback inside [TIMEFRAME]. The audit is the active ingredient; the only reason this remains a 20–30% leak industry is that nobody walks the line with the right instrument.