Transformer Noise Guarantees: Sound Power, Sound Pressure and Acceptance

Make acoustic bids comparable by specifying the quantity, operating state and measurement method before translating factory data into a site noise study.

1. Name the quantity before naming the limit

A guarantee of “65 dB” is incomplete. Identify whether it means A-weighted sound power level LWA or A-weighted sound pressure level LpA, the reference quantity and the operating condition. Sound power describes acoustic energy emitted by the source; pressure describes the acoustic field at a measurement position. Their numerical values are not interchangeable. IEC 60076-10 defines pressure and intensity methods from which transformer sound power is determined.

Factory source acceptance and environmental noise acceptance at a neighbour's property are separate deliverables. A site receiver limit cannot be inserted into a factory purchase order without an acoustic model connecting the source, location and environment. This guide addresses specification clarity rather than setting a universal permissible noise level.

2. Specify all operating states

Request the rated voltage, frequency, tap position, loading state, cooling stage and fan or pump arrangement for each guarantee. Core excitation noise, winding load noise and cooling equipment can contribute differently. A quiet naturally cooled result cannot represent a forced-cooling duty unless the combined operating state is explicitly assessed. Harmonic currents and excitation conditions deserve attention when the supplied transformer serves converters.

Define whether the factory report covers the complete assembly or separate source contributions. Ask for the relevant frequency-band data where the project study needs tonal assessment. The IEC application guide explains acoustic characteristics and why factory and in-service measurements can differ. Its public description supports these subjects, not an unseen numerical tolerance.

3. Illustrative combination of source levels

Suppose an illustrative acoustic study has independent source sound-power contributions of 70 dB(A) from the transformer and 67 dB(A) from its operating cooling equipment. Using the same sound-power reference and logarithmic energy addition:

LWA combined = 10 log10(10^(70/10) + 10^(67/10)) = 71.76 dB(A)

The combined result is approximately 71.8 dB(A), not 137 dB(A), and not the larger value alone. This assumes contributions may be combined on an energy basis for the stated condition; it is not a measured product result. It does not predict pressure at 10 m, indoor reverberation or a boundary compliance margin. Coherent tonal interactions and the measurement method require separate treatment.

4. A comparison sheet that prevents false equivalence

Offer or requirementClarification neededProcurement consequence
LWA guaranteeCooling state, load and methodCompare source power on matching conditions
LpA guaranteePositions, surface, environment and correctionsDo not compare directly with LWA
Property-boundary limitReceiver, time period and local assessment ruleRequires a site propagation study
Separate fan dataInstalled arrangement and simultaneous operationInclude actual cooling duty

Do not apply a generic distance correction to a near-field factory pressure number. A point-source free-field approximation has assumptions that may fail for a large transformer, reflective enclosure or multiple sources. The study must account for directivity, ground, barriers, building reflections and other plant equipment. An indoor kiosk may change the field even if the transformer's source sound power is unchanged.

5. Write an executable acceptance clause

The purchase specification should identify the selected standard edition, declared quantity, weighting, applicable measurement method, test configuration and reporting requirements. Agree how measurement uncertainty and any contractual margin are treated. Record ambient background conditions, measurement geometry, instruments and calibration, operating settings and corrections applied. Avoid an automatic pass based only on a brochure number.

Where factory access or background conditions prevent the agreed method, resolve an alternative before testing. A later site reading should be evaluated against its own agreed criterion, with energised neighbouring equipment and cooling stages recorded. Do not force unlike pressure measurements to appear comparable by renaming them sound power.

6. Join the purchase and site study

Provide the acoustic consultant with equipment layout, elevations, cooling schedules and guaranteed source data. Reserve space for realistic mitigation and maintenance access; a barrier that obstructs cooling changes another engineering constraint. Connect the forced-air cooling review to the noise duty and incorporate evidence in the factory acceptance plan.

Keep the accepted report and source model together. When adding fans, moving the transformer or changing the enclosure, revisit the site study. A successful factory test supports the declared equipment guarantee; it does not by itself establish compliance with every local receiver rule.

7. Acoustic procurement FAQs

Can identical dB numbers mean different performance? Yes. LWA and LpA are different quantities, and operating conditions also matter.

Can pressure at one distance prove a boundary limit? Only through a justified measurement or propagation assessment for the actual geometry and environment.

Should fans be included? Include the cooling stage required by the guaranteed duty, or specify separate contributions and the accepted combination method.

8. References