Last edited 31 Jan 2021

Main author

ParkerJonesAcoustics Other Consultant Website

Sound insulation testing


[edit] Sound testing to comply with Part E for residential developments

There are many sets of Building Regulations requirements which a new or refurbished residential development must meet. One of these is to achieve a minimum standard of sound insulation performance internally between dwellings. This is an important regulation, as neighbours seek some level of sound privacy from their neighbours, ensuring all can live and sleep in peace.

[edit] What are the sound insulation targets that need to be achieved?

The specific acoustic performance requirements for new builds and conversions are set within Approved Document E ‘resistance to the passage of sound’, as shown in the table.

Requirement E1 – on-site airborne sound insulation performance targets for separating (party) walls and floors.

Sound insulation targets for residential.jpg

Approved Document E classes ‘rooms for residential purposes’ as a room, or a suite of rooms, which is not a dwelling-house or a flat and which is used by one or more persons to live and sleep and includes a room in a hostel, a hotel, a boarding house, a hall of residence or a residential home, but does not include a room in a hospital or other similar establishment used for patient accommodation.

The criteria for refurbishments are slightly more lenient than for new builds. This recognises that it can be more difficult to soundproof an existing structure if it cannot be replaced, compared to implementing a construction from scratch.

This relates to sound insulation internally, between residential properties which share a separating wall or floor. It relates to separate properties, not internal walls/floors within a dwelling, which carry a much lower requirement for sound insulation and which are not subject to sound testings.

Standards for noise ingress from outside (i.e. road traffic) are not covered under Approved Document E, with the local planning authority having more of an input into what it considers to be a suitable internal noise-level target.

[edit] What is airborne sound insulation and a DnT,w + Ctr?

Airborne sound insulation relates to sound that travels via the air through separating structures between rooms. For example, airborne sound across a lightweight floor would pass through the plasterboard ceiling, into the floor void, through the floorboards and into the room. It relates to noise sources such as speech, music, television sound, and any noise source that radiates into the air, rather than directly into the structure.

The DnT,w + Ctr relates to a level of on-site sound insulation. In simple terms, put a noise source in one room (the ‘source room’), measure the noise level, then go next door (the ‘receiver room’) and measure again, then calculate the difference. This would be a ‘D’. A Dw is similar, but with the ‘w’ denoting a weighting across frequency which accounts for the way humans perceive different frequencies. A DnT,w is a Dw which is then ‘normalised’ to account for the reverberation time of the room (the ‘nT’), recognising that the level difference is dependant on how reverberant the receiver room is, and that every room will have a slightly different reverberation time, hence we need to normalise the result to assess every sound test on a level playing field against a fixed target. A DnT,w + Ctr then adds the addition of a low-frequency correction (the ‘Ctr’), as low-frequency noise is difficult to control and therefore the regulations add this correction to improve the standard of sound insulation against low-frequency noise such as music.

Here is an outline expanation. Take a source level in one flat of 100 dB(A), then a receiver level in the neighbouring flat of 55 dB(A), the level difference is 45 dB, hence our sound insulation performance should be somewhere in the region of 45 dB DnT,w + Ctr.

The higher the DnT,w + Ctr, the better the level of airborne sound insulation.

[edit] What is impact sound insulation and an L’nT,w?

Impact sound insulation relates to noise generated by actual impact on a structure. This is typically only assessed for intermediate floors, to consider noise such as footsteps, chair scrapes and any other object being dropped or impacting on the floor.

The L’nT,w is determined by measuring sound levels within a room where the floor above is being excited by an impact. An acoustic consultant will use what is called a ‘tapping machine’ to do this, which drops a series of weights/hammers on the floor in regular succession at a fixed force (if they varied the force every time they tested, they would get completely different results).

The lower the L’nT,w, the better the level of impact sound insulation.

[edit] What does sound testing actually involve?

For compliance with the Building Regulations, the acoustic testing must be conducted by a qualified acoustic engineer who is accredited to conduct sound testing by an organisation such as Ukas or the Association of Noise Consultants. Such an engineer will be using expensive equipment, including a Class 1 sound-level meter which costs several thousand pounds… so do not try this with a noise meter bought for a few pounds.

Airborne sound insulation tests are conducted by placing a loudspeaker, or two loudspeakers in the ‘source’ room. These loudspeakers are used to play white or pink noise at a very high amplitude, to the tune of 95 – 110 dB(A) within the room (the acoustic engineer will be wearing hearing protection). The reason for using this type of noise, is that it is constant in amplitude, with an even spread of sound energy across frequency, keeping the test consistent. The engineer will use their sound-level meter to make a series of measurements in the source room. They will then go to the adjacent room (with the loudspeaker still on) and take a series of measurements within the ‘receiverroom.

After turning off the loudspeaker, two other tests will be made in the receiver room. One is to measure the reverberation time (to get the ‘nT’ correction in DnT,w + Ctr), which can be achieved by generating an impulse in the room, using the sound-level meter to measure/calculate the time taken for sound to decay by 60 dB. The impulse can be made by popping a balloon, shooting a starter pistol, or using the loudspeaker to generate white noise and then abruptly turning it off. The reverberation time affects how much sound energy bounces around the room, therefore affecting sound levels in the receiver room, which affects the level difference from the sound source next door. The second test, is to measure background noise. If measuring a high level of sound insulation, it might be expected that background noise from other noise sources will be at a similar level to the level in the receiver room from the noise source next door, therefore a correction for this must be made.

Putting these together, the measurements of level difference, reverberation time, and background noise, enables the calculation of the result DnT,w + Ctr.

Impact sound insulation tests as alluded to earlier, are conducted using a ‘tapping’ machine. This machine is placed on the floor and set to work. The acoustic consultant will then move downstairs to measure in the room below. As per airborne sound tests, the acoustic engineer will measure reverberation time and background noise in the receiver room. Putting this together enables the consultant to calculate the L’nT,w.

[edit] How many sound tests are needed?

In large-scale residential developments, there will be a certain level of repeatability. Does every single wall and floor need to be tested? No. Assume, for example, that we have a building containing 10 very similar flats, with the same constructions and construction detailing between them. For this, Approved Document E indicates that we would only need 1 ‘set’ of tests, as it suggests that at least ‘one set of tests for every ten dwelling-houses, flats or rooms for residential purposes in a group or sub-group’.

[edit] So, what defines a ‘set’ of tests?

Sets of tests in flats with a separating floor and a separating wall.
Sound Test sets 1 no text.png

One set of tests should compromise six tests (four airborne, two impact):

Sound Test sets 2 no text.png

Sets of tests in flats with separating floors but without separating walls.

One set of tests should compromise four tests (two airborne, two impact):

Sets of tests in dwelling-houses with separating walls but no separating floors.
Sound Test sets 3 no text.png

One set of tests should compromise two tests (two airborne):

  • One airborne test across a wall between (where possible) a pair of living rooms.
  • One airborne test across a wall between (where possible) a pair of bedrooms.

[edit] What condition should the site be in for sound testing?

Sound testing should ideally be conducted after completion of the building works, or as close to completion as possible. A failed acoustic test can often be down to constructions not being completed, penetrations and air gaps not being sealed or doors not being in place.

An acoustic consultant will also ask for quiet. If asked to come to site a few weeks before completion and there are still a lot of construction works and snagging going on, the acoustic testing results may be adversely influenced by this noise.

The acoustic consultant may welcome the provision of 240V power on site. While some acoustic engineers use battery-operated equipment, many do not, and having to carry a transformer with an extension lead can mean that testing takes longer, which may mean more site visits, and therefore more cost for the client.

Lastly, something that is often overlooked, is that impact sound insulation testing should be conducted in residential developments without the floor finish. A carpet floor finish is a very good dampener of impacts upon the floor. But in a few years, what if the occupant decides they would like to change this to hard flooring? Answer: the level of impact sound insulation could get much worse. Therefore, to safeguard against this, impact sound insulation testing must be conducted without the floor finish (unless it can be argued that the finish is integral and cannot be changed later down the line). It’s for this reason that party floors usually have an acoustic/resilient underlay within the floor construction, or beneath the floor finish.

[edit] About this article

This article was written by ParkerJones Acoustics in January 2020.

[edit] Related articles on Designing Buildings Wiki


Designing Buildings Anywhere

Get the Firefox add-on to access 20,000 definitions direct from any website

Find out more Accept cookies and
don't show me this again