How Loud Is an Air Compressor? Two Regulated Figures

Directive 2000/14/EC splits outdoor machinery into two regimes. Some equipment must simply carry a marked guaranteed sound power level; other equipment must additionally stay below a permissible sound power level. Compressors below 350 kW are in the second group, which makes them one of the more strictly regulated machines in a workshop or on a site.

For compressors up to 15 kW net installed power the permissible sound power level is 99 dB in stage I, applying from January 2002, and 97 dB in stage II, applying from January 2006. Above 15 kW the limit is expressed as a formula in installed power rather than a fixed figure.

Because these are limits on sound power, they are enforceable against the machine without reference to any particular site. That is a deliberate design: sound power can be determined in a test environment and reproduced, whereas a level at a distance depends on the site.

The FHWA's construction noise work takes the opposite approach. Its reference inventory lists equipment with a specification limit and an actual measured maximum level, both expressed in dBA slow response at 50 feet, together with the number of field samples behind each figure.

For an air compressor the specification level is 80 dBA and the measured level is 78 dBA, based on 18 samples gathered during a major construction programme. The inventory also assigns an acoustical usage factor of 40 percent, reflecting that the machine does not run continuously.

These are modelling reference values for predicting construction noise at a distance. They are not a specification of what any individual compressor emits, and they describe large site compressors rather than the portable unit in a home garage.

One is sound power in dB relative to 1 pW; the other is A-weighted maximum sound pressure at 50 feet. There is no way to move between them without knowing the acoustic environment, the directivity of the source and the measurement geometry — none of which either source publishes for this purpose.

This is worth dwelling on, because compressor noise is one of the most commonly misquoted figures on the internet, usually by presenting a sound power limit as though it were a level at a few metres. Doing so overstates what a bystander experiences by a wide margin.

The honest position: the EU tells you what the machine may radiate, the FHWA tells you what a site compressor measured at 50 feet, and neither tells you what your ears receive at arm's length.

Oil-free direct-drive units — the pancake and hot-dog compressors sold for nail guns — are typically the loudest per unit of output, because the motor drives the pump without a belt and there is little damping in the assembly. Belt-driven, oil-lubricated compressors run more slowly and are usually quieter for the same delivery.

Rotary screw machines behave differently again: continuous rather than reciprocating, with a steadier spectrum and often an acoustic enclosure. Site-towed diesel units add engine noise to the compression noise entirely.

None of those distinctions is captured in a single category figure, and the two regulated numbers above were not written to distinguish between them.

The cycle is what makes compressors annoying, and it is also what makes them measurable in a useful way. Compare filling from cold against topping up, tank in the garage against tank outside the door, hard floor against a rubber mat.

Uncalibrated, our meter reports an A-weighted digital signal level in dBFS rather than sound pressure. A one-time local calibration against a trusted reference reading turns that into an estimated dBA figure. Neither is a sound power determination, and neither can check a declared figure on a machine plate.