Sound power level and sound pressure level are not the same number, and confusing them is the usual reason a "quiet" fan turns out loud on site.
Power vs Pressure, Simply
Sound power level (Lw, in dB) is a property of the fan itself, like its electrical power; it does not depend on where you stand. Sound pressure level (Lp, in dB(A)) is what a person hears at a point, and it falls with distance and is shaped by the room's reflections and absorption. The same fan in a hard tiled plant room sounds louder than in an absorptive plant room, yet its sound power is unchanged, because the room only changes what arrives at the ear, not what the fan emits.
This is why a spec quoting only dB(A) without distance and room context is nearly useless: it hides where and how the number was measured. Two fans with identical sound power can read very differently on a supplier's brochure purely because one was measured in a better chamber, which is why the buyer must insist on the underlying power data.
Knowing how a number was produced matters as much as the number. Sound power is derived rather than measured directly: a laboratory takes sound pressure at defined positions in a reverberant or anechoic room, or uses an in-duct method, and calculates the power the source must be emitting. Standardised methods exist precisely so two laboratories reach the same answer, and a data sheet should name the method it used. A-weighting is the other thing to check. The A filter discounts low frequencies to match human hearing, so a fan with strong 63 Hz and 125 Hz content looks quiet in dB(A) yet still transmits a rumble through structure that occupants describe as noise even when the meter reads well.
Estimating On-Site Noise
To estimate pressure at a listener, start from the sound power, subtract the spreading loss (about 6 dB per doubling of distance in free field, or 3 dB in a reflective indoor "hemi" condition with a directivity factor), then add or subtract the room effect. A fan rated 85 dB sound power at 3 m in a reverberant room may read close to that; in free field at 10 m it would be much lower, because distance and room together set what the ear experiences.
| Item | Specification | Note |
|---|---|---|
| Sound power Lw | Fan property | Independent of position |
| Sound pressure Lp | At listener | dB(A), distance-based |
| Spreading | 6 dB / double | Free-field loss |
| Room effect | ยฑ several dB | Absorption vs reflection |
| Octave data | 63โ8,000 Hz | Needed for real estimate |
| Inlet, outlet, casing | Rated separately | Different paths need different data |
| Low-frequency content | 63 and 125 Hz bands | Discounted by A-weighting |
For a real estimate you also need the octave-band spectrum, because a fan loud at 1,000 Hz annoys more than one loud at low frequency, and any silencer must target the right band. A single-number rating hides which octave dominates, so the designer cannot size the attenuator correctly without the spectrum.
A fan radiates by more than one path, and a specification quoting a single figure hides which one dominates. Inlet-side and outlet-side sound power travel along the duct to the occupied space, while casing-radiated sound enters the plant room and then the structure, and each path needs its own treatment: attenuators for the duct, lagging or an enclosure for the casing, isolators for the structure-borne route. Attenuators are not free either. Every silencer adds pressure drop the fan has to make up, and at high duct velocity the silencer itself regenerates noise, so an oversized attenuator in a fast duct can end up louder than a shorter one in a slower duct.
Buyer Checklist
- Ask for Lw: Request sound power with octave bands, not a single dB(A) that hides the measurement.
- Fix the point: State the listener distance and room when comparing fans, or the comparison is meaningless.
- Model the room: Use the room constant to convert power to pressure at the actual ear position.
- Size silencers: Target the octave bands that dominate the spectrum, not the overall level.
- Beware the chamber: Check whether a quoted dB(A) came from a favourable test room.
- Ask which path: Get inlet, outlet, and casing-radiated data separately so treatment goes where the noise actually is.
- Watch regenerated noise: Keep duct velocity moderate, since a silencer in a fast duct can generate noise of its own.
Why the Distinction Matters
Judging a fan by a bare dB(A) risks either over-specifying silencers and adding cost, or under-protecting occupants and breaching comfort limits. Keruite publishes sound power data with octave bands so the buyer can convert to the actual site pressure level rather than guess from a single flattering number on a brochure.
Acoustic disputes are expensive because remedial attenuation is retrofitted into ductwork that is already installed, so the data has to be right at selection. Fans built in an 80,000 m² plant by a manufacturer with more than 30 years of production are measured for airflow and sound in an in-house performance laboratory, so octave-band power data accompanies the performance curve instead of being estimated afterwards. ISO 9001:2015 covers the panel and rotor fabrication that determines how closely a production unit matches the tested one, CCC certification covers the electrical build, and with fans in more than 50 countries data is issued in both metric and imperial conventions.
Our engineers respond to selection enquiries within 24 hours, including sound power data and on-site noise estimates for fan selections.