Sound insulation of precast walls: measured values

Sound insulation of walls is one of the few properties of a building that cannot be improved after handover. The facade colour can be changed, services rerouted, partitions moved — but if you can hear the neighbour, the only real remedy is rebuilding the element. That is why the decision about what the separating walls are made of is taken once and for good, before production even starts.

This article explains where the sound insulation of an element comes from, how to read the ratings in technical data, and which elements suit which use. Every figure below comes from laboratory measurements of our own elements, not from calculations.

Where sound insulation comes from

Sound passing through a wall sets it vibrating. The heavier and more homogeneous the element, the harder it is to excite — hence the basic rule of building acoustics, the mass law: doubling the mass per unit area gives roughly 6 dB more insulation.

That explains why concrete performs better here than lightweight construction, and why wall thickness matters more than any finish. But mass is not everything. Homogeneity counts too: an element without voids, gaps or penetrations through which sound could bypass the material. In a precast wall the services are embedded during production rather than chased in afterwards — no bypass routes are created that nobody will inspect later.

The third factor is the joints. The best-chosen element stops working if its connection to the floor slab or the perpendicular wall is not tight. Ratings quoted in data sheets are therefore meaningful only when measured on an element assembled the way it is assembled on site.

How to read the ratings

Test reports and technical data use the notation R'w (C; Ctr). It carries three different pieces of information.

R'w is the weighted apparent sound reduction index, determined to PN-EN ISO 717-1. A single figure in decibels, derived by comparing the measured third-octave curve with a reference curve. This is the value you compare between products.

C is the spectrum adaptation term for pink noise — speech, music, television. You add it to R'w when the element has to protect against domestic noise from the flat next door.

Ctr applies to low-frequency noise — road traffic, heavy vehicles, bass. It is usually more negative than C, because low frequencies pass through elements most easily. For external walls on a busy street it is therefore the sum R'w + Ctr that matters.

The practical consequence: an element with R'w = 59 dB and Ctr = −4 dB delivers an effective 55 dB against traffic noise. Comparing bare R'w values between products with different Ctr can therefore mislead.

Measured values for our elements

In May 2025 the KFB Acoustics laboratory tested five of our wall elements to PN-EN ISO 10140-2:2021-10. The measurements were made on elements assembled to our installation manual — joints grouted with expansive concrete and a movement joint at floor level, that is, in a configuration matching the building site.

Report numbers 0166-25P-T42-1-001a to -005a, KFB Acoustics Sp. z o.o., Domasław.

Two conclusions are worth drawing straight away.

Lightweight concrete is not acoustically worse — it is lighter. A 180 mm wall in LC16/18 reaches 55 dB, more than 160 mm of normal-weight concrete (54 dB). At equal thickness normal-weight concrete wins by four decibels, but the comparison only holds once the same sections are set against each other.

Moving from LC16/18 to LC20/22 does not change the rating. Both classes give 55 dB at 180 mm. The denser lightweight class is therefore chosen for load-bearing capacity, not for acoustics — only the switch to normal-weight concrete affects that.

Selection table

Proven in practice: at the guesthouse complex in Ośno Lubuskie the external and partition walls were built in lightweight concrete 12 and 15 cm thick, while the walls between the rented rooms were made of C30/37 at 18 cm.

What these numbers do not cover

The figure measured in the laboratory applies to the element alone. On site, flanking paths are added: sound also travels through the floor slab, the perpendicular walls and shared structural members. Insulation achieved in the building is therefore usually lower than in the laboratory, and the gap depends on how the joints are detailed and whether acoustic movement joints were provided.

The second matter is impact sound — footsteps, furniture being moved, dropped objects. It travels through the structure rather than the air, and it is countered by the floor build-up: a screed on a resilient layer, separated from the walls by edge strips. A wall rated R'w = 59 dB will not help if the screed on the slab is laid rigidly.

The third is services. Electrical boxes cast into the element in the factory do not weaken it, but a socket chased right through a separating wall does. When ordering elements it is worth agreeing the position of the boxes on both sides of the wall so that they do not sit back to back.

What this means when ordering

Sound insulation is a property ordered together with the element, exactly like concrete class or fire resistance. It is settled at shop-drawing stage — before the mould is made — not made good later with layers on site. If the design does not clearly mark which walls are separating walls, it is worth resolving that before production: changing the thickness of an element afterwards means a new element.

We make the test reports available on request, together with the third-octave band curves, if the designer needs them for the acoustic calculation of the whole building.