Transformer Sound Level Determination: Testing Methods, Standards, and Applications

Transformer sound level determination is a standardized acoustic testing procedure used to measure and verify the noise emitted by power transformers, distribution transformers, and reactors in accordance with IEC 60076-10, ISO 3744, and GB/T 1094.10. By quantifying A-weighted sound pressure and sound power levels, manufacturers, utilities, and accredited third-party laboratories can demonstrate compliance with contractual noise limits, environmental regulations, and product specifications. This article explains the measurement principles, sample types, test procedures, calculation methods, and practical applications of transformer noise testing.

Technical Overview

Transformer noise, commonly perceived as a low-frequency hum, originates from several physical mechanisms inside the active part and cooling system of the unit. The dominant source is magnetostriction, the periodic dimensional change of grain-oriented electrical steel laminations under alternating magnetization, which vibrates the core at twice the supply frequency and its harmonics (100, 200, 300, and 400 Hz for a 50 Hz system). Additional contributions come from electromagnetic forces acting on windings and structural joints, tank and magnetic shield resonance, as well as broadband aerodynamic noise generated by cooling fans and oil pumps.

Sound level determination is the standardized process of measuring this emitted noise and converting the results into objective acoustic quantities. The principal international standard is IEC 60076-10 (Power transformers – Part 10: Determination of sound levels), supported by the application guide IEC 60076-10-1. In China, the equivalent national standard is GB/T 1094.10. These documents reference the general acoustic methods of the ISO 3740 series (sound pressure based determination of sound power levels) and ISO 9614 (determination of sound power levels by sound intensity). The key deliverables of the test are the A-weighted average sound pressure level and the A-weighted sound power level, which are compared against guaranteed or specified limits.

Two measurement philosophies are recognized: the sound pressure method, which measures sound pressure on a prescribed contour surrounding the transformer, and the sound intensity method, which measures the flow of acoustic energy through a measurement surface and is less sensitive to reflective environments. Both methods require strict control of excitation conditions, background noise, and environmental corrections to produce reproducible results with engineering-grade accuracy.

Testing Samples

Oil-Immersed Power Transformers

Large generator step-up transformers, interconnecting autotransformers, and system power transformers are tested for sound level as part of routine and type testing. Because these units produce significant low-frequency noise and are frequently installed near residential boundaries, sound power verification is a contractual requirement. Tests are performed in both no-load (excitation) and load-current conditions where specified, with cooling equipment operating in defined modes such as ONAN, OFAF, or ODAF.

Distribution Transformers

Pole-mounted, pad-mounted, and compact substation transformers with ratings typically from 50 kVA to 3150 kVA are measured to verify compliance with utility noise specifications and eco-design requirements, such as those mandated by European regulations for small, medium, and large power transformers. The compact dimensions of these units allow the measurement contour to follow the principal radiating surface closely.

Dry-Type Transformers

Cast-resin and vacuum pressure impregnated (VPI) transformers installed indoors in commercial buildings, data centers, and industrial facilities must satisfy strict room noise criteria. Since the active part is not enclosed in an oil-filled tank, natural-convection or forced-air cooling noise can be more prominent, and the measurement surface is defined around the enclosure faces of the unit.

Reactors and Special Transformers

Shunt reactors, series reactors, smoothing reactors for HVDC applications, and converter transformers are also subjected to sound level determination. Reactors may exhibit distinctive tonal spectra due to winding vibration and magnetic clearance forces, so spectrum analysis in octave or one-third-octave bands is often added to the basic A-weighted measurement.

Testing Items

A complete transformer sound level determination program typically includes the following measured and derived quantities:

• A-weighted average sound pressure level (LpA) on the specified measurement contour.
• A-weighted sound power level (LWA), calculated from the measured sound pressure level and the measurement surface area.
• No-load (excitation) sound level, measured with rated voltage and frequency applied to one winding and other windings open-circuited.
• Load-current sound level, measured with rated current circulated to capture winding vibration noise, where specified.
• Sound level contribution of cooling equipment (fans and pumps), measured separately or in combination with excitation noise.
• Octave-band or narrow-band frequency spectrum analysis to characterize tonal components at 100 Hz and its harmonics.
• Background noise level and environmental correction factor (K) used to validate and correct the measurement results.

Testing Methods

Sound Pressure Method

The sound pressure method described in IEC 60076-10 and GB/T 1094.10 follows the engineering principles of ISO 3744. A measurement contour is established at a horizontal distance of approximately 0.3 m from the principal radiating surface of the transformer, which is an imaginary surface enveloping the tank and major attachments. Microphone positions are distributed along this contour at prescribed heights: for transformers of lower height, a single contour at half the height is used, while taller units require additional contours at fractions of the total height, with positions spaced no more than 1 m apart along the perimeter. At each position, the A-weighted sound pressure level is measured with a class 1 integrating sound level meter, and the energy-average level is computed.

Sound Intensity Method

The sound intensity method, based on ISO 9614-1 and ISO 9614-2 and also accepted by IEC 60076-10, uses a two-microphone intensity probe to measure the sound intensity vector normal to a measurement surface enclosing the transformer. Because the method inherently rejects steady background noise and reflections outside the measurement surface, it is well suited to factory halls with moderate reverberation and to on-site measurements where environmental control is limited. The sound power level is obtained directly by integrating the normal intensity over the measurement surface.

Test Conditions and Procedure

To ensure validity and reproducibility, the test is performed under controlled conditions:

• The transformer is excited at rated voltage and rated frequency, with supply waveform distortion within the limits specified by the standard.
• The tap changer position, cooling mode, and operating state of fans and pumps are defined and documented.
• Background noise is measured before and after the test; the difference between the combined level and the background level must be sufficient, and corrections of 1 to 3 dB are applied when the difference falls within the permitted range, otherwise the measurement is invalid.
• An environmental correction factor K, accounting for room absorption and reflections, is determined and applied in the sound power calculation; the magnitude of K is limited by the accuracy grade of the method.
• Wind speed, ambient temperature, and nearby reflecting objects are recorded, and measurements are taken with the microphone positioned away from obstructions.

Calculation of Results

The average A-weighted sound pressure level is calculated as the energy average of all microphone positions: LpA = 10 × log10 [(1/N) × Σ 10^(LpAi/10)]. The A-weighted sound power level is then determined from LWA = LpA + 10 × log10(S/S0), where S is the area of the measurement surface in square meters and S0 = 1 m². Background and environmental corrections are incorporated according to the standard, and the final result is reported together with the measurement uncertainty, typically within approximately 2 to 3 dB for engineering-grade methods.

Applications

Transformer sound level determination serves a wide range of engineering and regulatory purposes:

• Factory acceptance testing to verify compliance with guaranteed sound power levels stated in purchase contracts and technical specifications.
• Demonstration of conformity with eco-design and energy-related product regulations that impose maximum permissible noise levels for distribution and power transformers.
• Environmental impact assessment of substations, where measured transformer sound power data are used as source input for boundary noise calculations in accordance with ISO 1996.
• Product development and benchmarking, including evaluation of core materials, step-lap core designs, tank vibration damping, and low-noise cooling systems.
• Diagnostic support, since changes in the tonal spectrum or overall level can indicate core clamping relaxation, loose structural parts, or cooling system degradation.
• Independent verification by accredited third-party laboratories for type testing, customer audits, and dispute resolution.

FAQ

Why does a transformer hum at 100 Hz or 120 Hz?

Magnetostriction causes the core laminations to expand and contract twice per cycle of the magnetizing current, so the fundamental acoustic tone appears at twice the supply frequency: 100 Hz on 50 Hz systems and 120 Hz on 60 Hz systems, accompanied by harmonics at higher multiples. Winding vibration under load current produces tones in the same frequency family, while fans and pumps add broadband noise.

What is the difference between sound pressure level and sound power level?

Sound pressure level describes how loud the noise is at a specific location and varies with distance and environment, whereas sound power level characterizes the total acoustic energy emitted by the transformer and is an intrinsic property of the source. Standards require the measured sound pressure data to be converted into a sound power level, which allows fair comparison between units of different sizes and verification against contractual limits.

At what distance are the measurements taken?

Per IEC 60076-10 and GB/T 1094.10, microphone positions are located on a contour approximately 0.3 m from the principal radiating surface of the transformer, with the contour height or heights determined by the overall height of the unit. Some specifications additionally require measurements at 1 m or 2 m from the tank for compatibility with older data or particular customer requirements.

Can sound level tests be performed on site after installation?

Yes, on-site measurement is possible using the sound intensity method or carefully corrected sound pressure measurements, because ISO 9614 based techniques tolerate moderate background noise and reflections. However, results obtained in the field generally carry higher uncertainty than laboratory or factory measurements and must be interpreted with documented ambient conditions.

What noise levels are typical for transformers?

Typical A-weighted sound power levels range from roughly 50 to 65 dB(A) for small distribution transformers, 60 to 75 dB(A) for medium power units, and 75 to 90 dB(A) or more for large generator step-up transformers with forced cooling. Contractual limits and regional regulations define the applicable maximum values for each installation.

Which standards should be specified when ordering a sound level test?

The measurement itself should reference IEC 60076-10 or GB/T 1094.10, with the method clause identifying either the sound pressure or sound intensity procedure. Supporting references include ISO 3744 for engineering-grade sound pressure methods, ISO 3746 for survey-grade measurements, and ISO 9614-1 and ISO 9614-2 for sound intensity based determination of sound power levels.