Other Brands

Media Hub

Media Hub

CEO Forums

Podcasts

Power Panels

Tech Talks

Technology Showcases

Webinars

Event Interviews

Other Brands

Logo

Trustworthy coverage of the transformer and transformer-related industries.

Visit Website
Logo

Trustworthy coverage of the transformer and transformer-related industries.

Visit Website

Quiet by design: Understanding and Reducing Transformer Noise

Production of a low-noise transformer starts with a careful evaluation of the customer’s needs and special requitements.

Introduction
Transformer noise was recognized as a problem as early as 1940 [1], and the magnetostriction phenomenon in core steel sheets was identified as the primary source. Over the following decades, the electrical steel improved.

With thinner steel sheets, grain orientation, and domain refinement, core losses and noise were reduced, making winding noise increasingly important [2]. In parallel, noise standards were developed [3], [4] and countries adopted legislation or guidance [5], [6], [7], [8] to limit excessive environmental noise. In the past two decades, the research on the topic has accelerated due to general increase in public-health awareness and the documented links between noise and health outcomes [9]. As cities grow and the distance between people and power transformers shrinks, demand for low-noise transformers is increasing.

Noise sources
There are three main noise sources in a power transformer: the core, the windings and auxiliary equipment, such as cooling fans and pumps. The core generates noise through the magnetostriction phenomenon and Maxwell forces. Magnetostriction is the change in electrical-steel dimensions due to a change in the magnetic flux; Maxwell forces (magnetic pressure) arise from the magnetic field and act on magnetized laminations. Both mechanisms cause core vibrations, which cause pressure fluctuations in the transformer oil. These travel from the core surface to the internal tank surface and excite the tank walls. Finally, the excited tank walls radiate airborne sound, which is then perceived as unwanted transformer noise.

The same mechanism applies for the pressure fluctuations from the winding surface, where the vibrations are the result of electromagnetic forces between current-carrying conductors.

Each of the three sources has a characteristic sound spectrum. Core noise, with a highly nonlinear magnetostriction phenomenon, produces a distinct tonal spectrum, the familiar transformer hum. The noise spectrum is composed of multiple harmonics of the line frequency, 60 Hz, with prominent components at 120 Hz, 240 Hz, and 360 Hz (100 Hz, 200 Hz, and 300 Hz in the 50 Hz world), in some cases even higher harmonics can dominate. Winding noise is dominated only by double the line frequency, higher harmonics are rare and can be an indication of an internal issue. By contrast, cooling fans emit a broadband noise spectrum without significant tonal spikes. An example of core and windings noise spectrum for 50 Hz line frequency is shown in Figure 1.

Noise prediction
A challenging task emerges when a power transformer supplier needs to guarantee the noise level of a not yet built power transformer. The task is even tougher if only one or two units of a new design will ever be produced. That is common in our factory, many different designs, but small production runs.

The first approach to the problem is based on past measurements and empirical data. It is well known that core noise correlates with core mass and magnetic flux density. Noise also depends on steel grade and supplier, design specifics, the clamping frame design, and more. Winding noise is caused by electromagnetic forces, which correlate well with transformer leakage reactive power. The empirical approach quickly hits its limits, where some transformers deviate significantly from the predicted noise values. Deeper insight is needed. To that end, we use advanced numerical tools to model the dynamic response of individual components and complete transformers to the specific excitation mechanisms, such as magnetostriction or electromagnetic forces.

Figure 1: A noise spectrum example for core and winding noise.

Figure 1: A noise spectrum example for core and winding noise.

To model the magnetic core with its specific laminated structure and thus orthotropic stiffness properties, we use a custom, two-step numerical approach [10]. First, we run a static analysis to determine the interlaminar pressure distribution; then we bond the laminations with beam elements with the varying cross-sections simulating the stiffness due to the interlaminar friction. This approach yields a physically representative model of the core within the frame work of linear structural dynamics.

Figure 2: a) fully coupled structural model of a single phase, b) The resulting sound pressure field and the tank displacement field at a specific frequency.

For the windings, we first solve a 2D axisymmetric electromagnetic model to obtain excitation forces. We then apply those forces to a fully coupled 3D structural model with solid-fluid interaction between the windings, the oil and the tank [11] as shown in Figure 2. These models augment empirical predictions and provide a deeper understanding of the specific design characteristics and the general strategies to reduce transformer noise.

Low-noise transformer
Production of a low-noise transformer starts with a careful evaluation of the customer’s needs and special requirements. Many customers prohibit external noise-reduction measures, which have a direct impact on the design. It is also crucial to define the operating conditions in which the noise level is guaranteed; controlling core noise alone is very different from controlling the combined noise of the core, the windings and the fans. The latter is far more challenging, because any design change can affect one or all three main noise sources.

With this in mind, we use our numerical prediction tools in an iterative design process to develop a design that meets all customer needs. To reduce core noise, we use a narrow selection of the world’s best electrical steels, lower core induction, apply vibration isolation measures, use step-lap core joints with trimmed excess material, refine the clamping frame and the clamping pressure. Reduction in winding noise involves resonance prevention, lowering excitation forces, adjusting winding geometry, changes in winding type, increasing clamping pressure and using high-stiffness materials. If cooling fans are unavoidable, their noise can be reduced either by using many small low-noise fans or a few larger fans operating at reduced RPM. For further reductions in noise levels we optimize the tank design to balance stiffness and noise radiation efficiency.

Once the design phase is completed, the job is not yet done. Low-noise designs require special attention on the shop floor; each component must be manufactured to a higher standard than if noise were not a constraint. The same core can be cut and stacked in a way that meets loss targets, but not noise. Similarly, the same windings can be wound, dried and dimensionally stabilized in either a noisy or a quiet way.

To verify both our models and the production process, we use a nonstandard noise test in which the supply frequency is stepped and the noise is measured at each step, as shown in Figure 3. This approach reveals resonances that can be compared with predictions. When the optimal low-noise design is built correctly, the result is not only a quiet transformer, but also a high-quality one.

Figure 3: Transformer resonance identification using microphones and supply frequency sweep.

Reduction in winding noise involves resonance prevention, lowering excitation forces, adjusting winding geometry, changes in winding type, increasing clamping pressue and using high-stiffness materials.

Case study

Two 50 MVA low-noise units with sound pressure level guarantees, as specified in Table 1, were installed at a transformer station in Bludenz. The same table also presents the measured values for one of the units.

The required noise levels were consistently achieved by applying all the previously described techniques. As shown in Figure 4, only the large number of fans and the special tank stiffener design are noticeable from the outside, due to low-noise design.

Table 1: Guaranteed and measured noise values.

ParameterGuarantee [dBA]Measured [dBA]
LpA,Un* (no fans)4543.6
LpA,Sn** @ 40 MVA (no fans)4845.9
LpA,Sn** @ 50 MVA (with fans)5251

*: Core noise.

**: Log sum of core, winding and optionally fan noise

Figure 4: Two 50MVA low-noise units in transformer station Bürs in Austria.

Over the past two decades, power-transformer noise standards and national noise regulations have been updated; meanwhile, world population has grown by about 20% and became more noise-aware.

Conclusion
Over the past two decades, power transformer noise standards and national noise regulations have been updated; meanwhile, world population has grown by about 20% and became more noise-aware. The increasing demand for transformers has also been accompanied by the need for lower noise.

Transformer noise is a complex, multiphysics problem spanning electromagnetics, structural dynamics and acoustics. Understanding the underlying noise generation mechanisms is the first step toward a low-noise transformer. The next is applying that knowledge to optimize the design; the final step is successful manufacturing of the transformer and achieving guaranteed noise levels.

Notably, a low-noise transformer is made from high-quality materials and manufactured to a higher degree of precision, which raises overall product quality.

References
[1] “Limitation of Transformer Noise,” Nature, vol. 145, no. 3687, pp. 1029–1029, Jun. 1940, doi: 10.1038/1451029a0.
[2] E. Reiplinger, “Study of noise emitted by power transformers based on today´s viewpoint,” presented at the CIGRE Internation Conference on Large High Voltage Electric Systems, Paris, 1988.
[3] “IEC 60076-10: Power transformers – Part 10: Determination of sound levels.” 2016.
[4] “NEMA TP 80050-2013 (R2024): Transformers, Step Voltage Regulators, and Reactors.”
[5] EPA 550/9-74-004, “Information On Levels Of Environmental Noise Requisite To Protect Public Health and Welfare With An Adequate Margin Of Safety.” Mar. 1974.
[6] Bundesministerium für Umwelt, Naturschutz und nukleare Sicherheit, “TA Lärm — Technische Anleitung zum Schutz gegen Lärm (Sechste Allgemeine Verwaltungsvorschrift zum Bundes Immissionsschutzgesetz).” Aug. 26, 1998. [Online]. Available: https://www.verwaltungsvorschriftenim-internet.de/bsvwvbund_26081998_IG19980826.htm
[7] European Parliament and Council, “Directive 2002/49/EC of the European Parliament and of the Council of 25 June 2002 relating to the assessment and management of environmental noise.” Jun. 25, 2002. [Online]. Available: https://eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=CELEX:32002L0049
[8] Gouvernement de la République française, “Décret n° 2006-1099 du 31 août 2006 relatif à la lutte contre les bruits de voisinage et modifiant le code de la santé publique (dispositions réglementaires).” Aug. 31, 2006. [Online]. Available: https://sante.gouv.fr/IMG/pdf/decret_bruits_voisinage_2006_1099.pdf
[9] G. Leventhall, P. Pelmear, and S. Benton, “A Review of Published Research on Low Frequency Noise and its Effects,” Jan. 2003.
[10] M. Pirnat, G. Čepon, and M. Boltežar, “Introduction of the linear contact model in the dynamic model of laminated structure dynamics: An experimental and numerical identification,” Mech. Mach. Theory, vol. 64, pp. 144–154, Jun. 2013, doi: 10.1016/j.mechmachtheory.2013.02.003.
[11] M. Pirnat and P. Tarman, “Low noise Transformer technology,” INTER-NOISE NOISE-CON Congr. Conf. Proc., vol. 259, no. 9, pp. 816–824, 2019.

Miha Pirnat holds a PhD in Mechanical Engineering and works as a Transformer Noise Expert at Kolektor ETRA in Slovenia. His work focuses on the analysis, modeling, and reduction of noise in power transformers, using techniques such as finite element analysis and vibration diagnostics. He collaborates with design and R&D teams to optimize transformer structures for acoustic performance and compliance with IEC and IEEE standards. Miha was a member of a CIGRE working group A2.54 on transformer acoustics and has authored or co-authored over 10 scientific papers in the field of vibroacoustics and transformer mechanics.



This article was originally published in the December 2025 issue of the Transformer Critical Components magazine.

View Magazine

Like this article? Share on

Subscribe image

Subscribe to our Newsletter

Subscribe to our newsletter and stay ahead with the latest innovations, industry trends, and expert insights in power systems technology. Get updates on cutting-edge solutions, renewable energy advancements, and essential best practices delivered straight to your inbox.