FESI DOCUMENT A4

Acoustics in buildings

October 2007

to be ordered:

Thermal Insulation Contractors Association

Mr. Ralph Bradley
Tica House
Allington Way
Yarm Road Business Park
Darlington
DLI 4QB
Tel : +44 (0) 1325 734140 direct
Fax: +44 (0) 1325 466704
www.tica-acad.co.uk
ralphbradley@tica-acad.co.uk

Acoustics in Buildings
Contents
A4-0

Intention....................................................................................................................................... 3

A4-1

Introduction.................................................................................................................................. 3

A4-2
Airborne sound insulation between rooms.................................................................................. 4
A4-2.1
General........................................................................................................................................ 4
A4-2.2
Terms and definitions.................................................................................................................. 5
A4-2.2.1 General principles ....................................................................................................................... 5
A4-2.2.2 Acoustic insulation indices depending on frequency .................................................................. 6
A4-2.2.2.1 Relations between the different criteria of acoustic insulation between rooms .......................... 8
A4.2.2.2.2 Relations between the acoustic insulation between rooms and the sound
reduction index of the separating wall ........................................................................................ 8
A4-2.2.3 Single numbers quantities........................................................................................................... 9
A4-2.3
The direct sound transmission through the separating wall...................................................... 10
A4-2.3.1 The influence of the dimensions of the reception room ............................................................ 10
A4-2.3.2 The influence of the sound reduction index of the separating wall ........................................... 12
A4-2.4
The transmissions by the lateral walls ...................................................................................... 12
A4.2.4.1 Simplified method: Statistical evaluation of the sound reduction loss resulting from
lateral transmissions................................................................................................................. 12
A4-2.4.2 The limits of the simplified method............................................................................................ 13
A4-2.4.3 Calculation of flanking transmissions........................................................................................ 17
A4-2.4.4 Special case of light double-partition walls ............................................................................... 18
A4-2.5
The parasitic transmissions....................................................................................................... 19
A4-2.5.1 Expected parasitic transmissions.............................................................................................. 19
A4-2.5.2 Accidental parasitic (spurious) transmissions ........................................................................... 20
A4-2.5.3 Which precautions to take?....................................................................................................... 22
A4-2.6
Acoustic insulation between rooms divided by a wall of different surface elements ................ 23
A4-2.6.1 Method of pre-calculation of the sound reduction index of composite walls............................. 23
A4-2.6.2 Sound reduction index limit of a composite partition wall of two different elements................. 24
A4-2.6.3 Acoustic insulation between two rooms separated by a wall with an opening ......................... 25
A4-2.6.4 Some considerations on doors ................................................................................................. 26
A4-3
A4-3.1
A4-3.2
A4-3.3
A4-3.4
A4-3.4.1
A4-3.4.2
A4-3.4.3

Acoustic insulation against outside noise ................................................................................. 27

General...................................................................................................................................... 27
The direct transmission of noise through a facade ................................................................... 27
The flanking transmissions in case of facades ......................................................................... 28
Parasitic transmissions in case of facades ............................................................................... 28
The expected parasitic transmissions....................................................................................... 28
Accidental parasitic (spurious) transmissions ........................................................................... 29
Some considerations on windows............................................................................................. 29

A4-4

A4-4.4.2
A4-4.5
A4-4.6

Transmission of the noise created in a room to the outside Environment

protection .................................................................................................................................. 32
General...................................................................................................................................... 32
General principles ..................................................................................................................... 32
Determination of the representative point sources ................................................................... 32
Calculation of the sound power level of a representative point source .................................... 33
The case of a segment consisting of structural elements of the building envelope
(group of plane radiators) ......................................................................................................... 33
For an open segment ("great openings" as given in chapter A4-4.3) ....................................... 33
Determination of the correction of the directivity of a representative point source ................... 34
Example for the application....................................................................................................... 34

A4-5
A4-5.1
A4-5.2
A4-5.3

Impact sound insulation ............................................................................................................ 39

General...................................................................................................................................... 39
Principles................................................................................................................................... 39
The level of impact sound ......................................................................................................... 40

A4-4.1
A4-4.2
A4-4.3
A4-4.4
A4-4.4.1

A4-5.3.1 Terminology to be used............................................................................................................. 40

A4-5.3.2 Efficiency of floor coverings ...................................................................................................... 41
A4-5.4
The prediction of the impact sound pressure level ................................................................... 42
A4-5.4.1 Lateral transmissions ................................................................................................................ 42
A4-5.4.2 Effect of the room dimensions................................................................................................... 42
A4-5.4.3 Behaviour of naked floors regarding impact sound .................................................................. 43
A4-5.4.4 Floor coverings.......................................................................................................................... 44
A4-5.4.4.1 Thin floor coverings................................................................................................................... 44
A4-5.4.4.2 Hard coverings directly placed on top of a soft under-layer...................................................... 44
A4-5.4.4.3 Floating floors............................................................................................................................ 45
A4-5.4.5 Example for the application of the prediction method ............................................................... 46
A4-5.5
Improvement of the impact sound insulation ............................................................................ 47
A4-5.6
Horizontal or diagonal transmissions of impact noise............................................................... 48
A4-6
A4-6.1
A4-6.2
A4-6.2.1
A4-6.2.2
A4-6.2.3
A4-6.3
A4-6.3.1
A4-6.3.2
A4-6.3.3
A4-6.4
A4-6.4.1
A4-6.4.2
A4-6.4.3
A4-6.4.4

Noise levels of building equipment ........................................................................................... 49

Maximum sound pressure level at the reception areas ............................................................ 49
The different types of energy transmission to be considered ................................................... 50
Transmission of airborne sound by the structure of the building .............................................. 50
Transmission of airborne sound by or across the ducts ........................................................... 51
Transmission of structure-borne sound .................................................................................... 51
The main equipment parts of a building and the types of transmission which they
produce..................................................................................................................................... 51
Ventilation systems ................................................................................................................... 51
Heating installations .................................................................................................................. 52
The installations for water supply.............................................................................................. 53
Some principal precautions to take to limit the noise of certain installations ............................ 54
Equipment installed at the outside ............................................................................................ 54
Heating systems in occupied places ......................................................................................... 55
Basement heating stations........................................................................................................ 55
Electrical transformation installations........................................................................................ 55

A4-7

Symbols, quantities, units ......................................................................................................... 56

A4-8

Standards and bibliography ...................................................................................................... 57

A4-0

Intention

The FESI Document A4 "Acoustics in buildings" is one of a series of six papers on acoustical building
problems together with their solutions.
The total block of acoustical documents will comprise the following titles:

A2
A3
A4
A5
A6
A7

"Basics of acoustics"
May 2000
"Product characteristics Acoustic insulation, absorption, attenuation" September 2002
"Acoustics in buildings"
"Acoustics in rooms"
"Industrial acoustics"
May 2004
"Acoustic measurements"

It is the aim of these documents to provide the theoretical basis and the consequential practical applications for designers. Numerous tables with useful characteristics are given which are normally widely dispersed in the related literature.
A4-1

Introduction

In case of airborne sound, the sound waves travel away from the source. When they hit an obstacle, a
wall, a ceiling, a partition, a facade, part of the energy driving it is reflected by the obstacle, another part
sets the obstacle into vibration and is partially absorbed. The rest is radiated from the obstacle in the form
of airborne sound, especially to the space on the other side of the obstacle, opposite the location of the
source.
What happens to the energy which stays on the side of the source has been presented in document A5
Acoustics in rooms. This document deals with the vibration of the obstacle and its consequence for the
energy it transmits on the side opposite to the side of the source. Actually, this is the phenomenon governing the problems associated with the acoustic insulation of airborne sound between rooms (the sound
insulation between two rooms in one building), the insulation between the exterior and the room (insulation of the facade) and between a room and the outside area (protection of the environment against noise
produced in a building).
The partitions in a building are not only excited by airborne sound waves. They can also be excited by
impact sound (that which is produced by walking on a floor, falling objects, the movement of chairs ...). In
the vicinity of the impact, a relatively strong energy is transmitted to one point of the partition. This is the
problem which is especially examined in the area of impact sound insulation.
There are also sources that produce airborne sound in a room where they are located and which in addition send direct energy to the partitions on which they are mounted. This is especially the case with building equipment.
National regulations respectively standards give the requirements related to these different problems. The
requirements can take two different forms which correspond to the approaches following these different
philosophies:

either the requirements take the form of the obligation to use certain materials; performance of walls,
of coverages or of the equipment to be used, with finally the description of techniques to be used in
combining these elements with bordering building elements. This type of requirement is best illustrated by the German standard DIN 4109.
or, alternatively, the requirements are given in the form of demanded results: One demands for example that the acoustic insulation between two places (two flats, two classrooms, two hotel rooms or
hospital rooms ...) must at least reach a certain value. In this case it is for the contractor to decide the
means able to reach this demanded result, which can be checked by measuring at the end of the
work and leads to liabilities in case of non-conformity. This type of requirements is best illustrated by
the French regulation regarding the acoustic quality of buildings.

For this second form of requirement, the designer, but also the contractor, need to know the acoustic
fundamentals of buildings, the former to determine the solutions most likely to meet the required results,
the latter to understand that what is demanded of him and to optimise the solution which might appear in
certain cases to be superfluous. In all contracts, the contractors must make sure that the written demands

do not hold for the same problem both an obligation to obtain certain results and the obligation to use
certain means. In case their means described are sadly insufficient, it is nevertheless the contractor who
is held responsible for measured bad results.
The regulations concern themselves most frequently with new buildings. For existing buildings, in which
acoustic problems had been proven, it is necessary to make a diagnosis of the existing situation to be
able to decide which are the predominant acoustical transmission paths. In this area, an improvisation
could be ineffective and costly. This is going to be the final chapter of this document.
Three series of different kinds of standards exist in Europe:

ISO 717, containing single values to characterise the acoustic properties of building components,
ISO 140, dealing with measurements by frequency (one third octave band or octave band),
EN 12354, dealing with the prediction of the acoustic insulation results to be expected from different
measurements.

The European harmonisation has necessitated for certain countries an adaptation of their regulations,
st
their techniques, their measurements, which is given in the vocabulary. From 1 January 2000, it is compulsory to use the regulations defined in the standard ISO 717. Particularly, certain symbols which need
to be used might seem very complex. For example, to characterise the acoustic insulation given by an air
intake in a room, one uses the symbol DnTeA,tr. The symbol D corresponds to a difference in the acoustic
pressure level, "nT" indicates that this difference is standardised, which means to say calculated for a
room where the reverberation time has a reference value, "e" to signify that this is the performance of an
element (the air opening), "A" to express the isolation by a unique value using the A-weighted noise
measurement, and "tr" to define the range of emission noise to be used, to know the noise of conventional traffic.
That the key of the symbols to be used is known, does not simplify these symbols, but instead it is allowed to simply determine the meaning of the considered values.
Finally, one could declare that the acoustic in buildings is treated by the series of standards EN 12354:
Estimation of acoustic performance of buildings from the performance of elements
Part 1: Airborne sound insulation between rooms
Part 2: Impact sound insulation between rooms
Part 3: Airborne sound insulation against outdoor sound
Part 4: Transmission of indoor sound to the outside
Part 5: Sound levels due to service equipment
Part 6: Sound absorption in enclosed spaces
These standards give the current possibilities to forecast a performance, but in a slightly too theoretical
form for contractors which are not acousticians. In the FESI document A4, we tried to extract from these
standards the useful elements and to give simplified methods of pre-calculation including the limits of their
utilisation. In case of doubt or in specific problems, one must take recourse to the appropriate part of the
standard EN 12354. Many practical examples will be given in the subsequent chapters.
The entire area of acoustics in buildings has several interactions. It is not possible to deal with one of
these phenomena without considering the consequences, favourable or unfavourable, for the others.
Equally, one must not deal with acoustical problems without also considering the interactions of solutions
with the other functions of a building: stability of the building, thermal insulation, ventilation, fire protection
... To examine the influences of one function on the other functions is the starting point for an optimisation
of solutions.
A4-2

Airborne sound insulation between rooms

A4-2.1

General

The Figure 1 repeats the Figure 19 from chapter 8 of FESI document A2 "Introduction to the following
documents obstacle to propagation of airborne acoustical waves.

When an acoustic wave hits a wall, as we have recalled in the Introduction, part of its energy is reflected,
another part absorbed by the wall and a third part is transmitted to the other side of the obstacle away
from the source (Figure 1).

Transmission

Reflection

Absorption

Figure 1: Reflection, absorption and transmission of the energy coming from a source a and hitting a partition b
The partition is characterised by three coefficients: r, reflection coefficient which is the proportion of the
energy reflected to the incoming energy. , absorption coefficient which is the proportion of the energy
absorbed and , transmission coefficient which is the proportion of the energy transmitted by the partition.
One can also write "r + + = 1" since the total amount of energy hitting the wall is transformed into reflected energy, absorbed or transmitted energy.
In the context of the acoustic insulation of airborne noise between rooms, it is the coefficient of transmission which interests us. This is materialised by the measurement of the sound reduction index R of a
wall:
R = 10 log(1/) = - 10 log () dB.
This is the special characteristic of a wall made of one or several materials composed in a certain manner. One speaks of the sound reduction index of a 16 cm concrete wall, of a double 98/48 wall with mineral wool, etc.
This is the transmission of acoustic energy as a result of the vibration of the partition to which it has been
excited by the sound waves arriving from the source.
A4-2.2

Terms and definitions

A4-2.2.1

General principles

Once a noise is produced by a source in a room, called "emission room", the acoustic waves hit the partitions which are then set into vibration, thus becoming sources of noise in the neighbouring room, called
"reception room". One uses normally the index 1 for the acoustic values characterising the emission and
the index 2 for those in the reception room.
In the emission room, all the walls of the room are set into vibration, especially the walls dividing the
rooms and the lateral walls, which the two rooms have frequently in common. Thus, the sound level L2 in
the reception room is composed of the direct transmission through the separating wall, characterised by
its sound reduction index R, but also by the transmission by the lateral walls. Especially where a lateral
wall is linked to the partition wall and is common to both rooms, the vibration of the lateral wall in the
emission room is partially transmitted to the lateral wall in the reception room and partially to the wall of
separation to which it is linked. Equally, the vibration of the separating wall is partially transmitted to the
lateral wall of the reception room. Thus, there are three paths of transmission at the junction of one lateral
wall and one separating wall. Generally, there are four lateral walls, which gives 4 3 = 12 paths of lateral

transmission to which the direct transmission through the separating wall must be added, gives 13 paths
of transmission between two rooms (Figure 2).

EMISSION

L1

RECEPTION

L2
A,T

Figure 2: Direct transmission through the separating wall and the transmission paths by the lateral walls. At each connection of the separating wall with the lateral walls, there are
three lateral paths of transmission, which gives 12 paths at a total of four lateral walls
between two rooms
The direct and lateral transmission paths are sometimes augmented by parasite transmission paths due
to (holes ?), openings in the separating wall, to defaults of tightness of junctions between the walls, to
crossing through (passage?) of pipes or ducts.
One notes that the acoustic energy tries to exploit all possible paths of transmission to travel from the
emission room to the reception room. The separating wall is by no means the only one to be considered
to ascertain acoustic insulation between the rooms. A consequence is that if the acoustic insulation existing between two rooms is not sufficient, one must not concentrate on the improvement of the acoustic
quality of the separating wall, but it is necessary to make an analysis to establish the hierarchy between
the transmission paths and then treat the most important ones. In this area, improvisation is frequently
misleading and costly.
All the energy transported by the direct transmission, the lateral and the parasite transmission paths is
consumed more or less rapidly by the absorbing materials contained in the reception room. The level L2
in the reception room is thus dependent on the walls forming that room, on the possible shortcomings of
these walls or of their connections and the absorption characteristics of the room. These absorption characteristics depend most frequently on the furniture and decoration of the room, factors dependent on the
occupant and not on the builder.
To be able to identify the insulation aims for the designer and to be able to compare the performance of
the insulation between different rooms of one building or between different buildings, it is necessary to
neutralize the factors that are outside the responsibility of the designer. It is thus that the standards for
definitions and for measurements provide corrective terms to use in order to calculate the insulation performances which would be obtained in case the reception rooms have reference absorption characteristics. The characteristics used are either the equivalent absorption area, either the reverberation time.
A4-2.2.2

Acoustic insulation indices depending on frequency

The acoustic insulation between rooms corresponds to the difference between the sound pressure level
L1 in the emission room and the sound pressure level L2 in the reception room.
One discerns three types of insulation between the rooms (see ISO 140-4/1998):
The level difference D between rooms:
D = L1 L2 in dB

(1)

This insulation depends on the characteristics of the reception room. For this, one might as well consider
2
the equivalent absorption area A (m ) of the room or the reverberation time T (s).

The normalised level difference Dn between the two rooms uses a corrective function for the equivalent
absorption area:
Dn = L1 L2 10 log (A/A0) in dB

(2)

A is the equivalent absorption area in m established in the reception room during the test.
2
A0 is a reference equivalent absorption area, set equally at 10 m , if nothing else has been stated in a
regulation or in a contract.
The correction term used corresponds to an evaluation of the difference of the level L2 if the equivalent
absorption area changes from A to A0 (in this approach, the constant R of a room is assimilate to the
equivalent absorption area (see also FESI Document A5).
It is determined by using the SABINE formula ( T = 0,16

V
) after having measured the reverberation time
A

in the reception room.

The standardised level difference DnT between two rooms uses a correction function for the reverberation
time in the reception room.
DnT = L1 L2 + 10 log (T/T0) in dB

(3)

T is the reverberation time in seconds, established in the reception room during the test.
T0 is a reference reverberation time, set equally at 0,5 seconds, unless a different statement is made in
the contract.
st

Note: In some European countries, up to 1 January 2000, this standardised insulation was called "normalised acoustic insulation". The new terminology in use is a consequence of European harmonisation.
One must not confuse the acoustic insulation between two rooms with the sound reduction index R of the
separating wall as measured in a laboratory. The actual insulation reflects the direct transmission through
the separating wall and the transmissions by the lateral walls, which both rooms have in common. The
sound reduction index of a wall is measured when that wall is placed between two rooms designed so
that no lateral transmission occur (Figure 3).

EMISSION

RECEPTION

1
S

A,T

Figure 3: For the laboratory measurement of the sound reduction index of a wall, the lateral
transmission paths are neutralised
The sound reduction index of a wall renders the sound transmission factor of that isolated wall R = 10
log (1/).
The formula for the index R is obtained by writing that the energy transmitted through the separating wall
conditioned by this factor is consumed by the absorption in the reception room.
R = L1 L2 10 log (A/S) in dB

(4)

where L1 and L2 are the sound pressure levels in dB in the emission, respectively the reception room.
2
A is the equivalent sound absorption area in the reception room (m )
2
S is the surface of the separating wall (m )
It is always possible to use this sound reduction index in a construction in place, even if there are lateral
transmissions on top of the direct transmission. But then, one has to deal with an apparent sound reduction index R' of the separating wall. With all acoustic values, the dash indicates that there is also lateral
transmissions. The standard ISO 140-4 gives a definition of that "apparent" index.
R' = L1 L2 10 log (A/S) in dB

(5)

The sound reduction index measured in a laboratory, where no lateral transmissions occur, could be written as:
R = 10 log (W 1/W 2)
where W 1 is the acoustic energy hitting the wall on the emission side, and W 2 the acoustic energy transmitted by the separating wall (W 1/W 2 = 1/). In an actual construction, in the presence of lateral transmissions, the apparent sound reduction index could be written as follows:
R' = 10 log (W 1/(W 2 + W 3))
where W 3 is the acoustic energy transmitted by all the elements outside the separating wall.
A4-2.2.2.1

Relations between the different criteria of acoustic insulation between rooms

One could compare the standardised insulation DnT and the normalised isolation Dn: In both cases, the
unprocessed difference L1 L2 is the same, and from this it could be deduced, using the SABINE formula
2
and taking 10 m and 0,5 seconds as reference values for the equivalent absorption area and the reverberation time, that:
DnT = Dn + 10 log ((V/V0) in dB

(6)
3

where V is the volume of the reception room and V0 is a reference volume set equally at 31,5 m , rounded
3
to 30 m , which is the size of a normal bedroom.
Equally, using the SABINE formula and the fact that in all cases the levels at the emission and the reception rooms are the same, one could deduce the following relations:
Dn = R' 10 log (S/A0) in dB

(7)
2

S being the surface of the separating wall, common to both rooms (m ).

or Dn = R' 10 log (S/10) in dB if A0 = 10 m
DnT = R' + 10 log ((0,16 V/(S T0))

(8)
(9)

S being the surface of the separating room common to both locations and V the volume of the reception
3
room in m
or DnT = R' + 10 log (0,32 V/S) if T0 = 0,5 s
A4.2.2.2.2

(10)

Relations between the acoustic insulation between rooms and the sound reduction
index of the separating wall

For a first step, one works with the hypothesis that there were no flanking or parasite transmission. In this
case, nothing would remain, but the direct transmission through the separating wall and one could measure its sound reduction index. In the three equations for R, Dn and DnT, the difference L1 L2 is the same,
which allows to continue with the following equations which correspond to the value of direct transmission
through the separating wall with its sound reduction index R:
The standardised acoustic insulation DnT and the sound reduction index R of the separating wall are
linked by the equation

DnT = R + 10 log (0,32 V/S) dB

(11)

where V is the volume in the reception room and S the surface of the separating wall which the two rooms
have in common.
The normalised acoustic insulation Dn and the sound reduction index R of the separating wall are linked
by
Dn = R 10 log (S/10) dB

(12)

where S is the surface of the separating wall which both rooms have in common.
In a second step, one arrives at the real situation in taking note of the fact that in case of an actual acoustic insulation, flanking transmissions do exist. These flanking transmissions cause an increase of the
sound pressure level in the reception L2 and as a consequence a diminishing acoustic insulation compared to that associated with the direct transmission. Two approaches can be chosen:

Either one expresses the insulation loss caused through lateral transmissions by taking off a certain
quantity of dB from the DnT and the Dn obtained.
DnT = R + 10 log (0,32 V/S) Tf dB
(13)
and Dn = R 10 log (S/10) Tf dB
(14)
Also, R' = R Tf
where Tf is the assessed loss of acoustic insulation caused by flanking transmissions. And in any
case it represents the level of lateral transmissions.
Equally, one could express:
DnT = R' + 10 log (0,32 V/S)
and Dn = R' 10 log (S/10)
Or one could calculate each of the thirteen passes of transmission between the two rooms by taking
into account the sound reduction indexes of all walls which the two rooms have in common and the
transmission coefficients at the junctions between the separating wall and the lateral walls associated
with the insulation of vibrations between the wall in the emission room and the wall in the reception
room. But these assessments become very complicated and it is preferable to use calculation software based on the European standard EN 12354-1 ( Estimation of acoustic performance of buildings
from the performance of elements part 1 : Airborne sound insulation between rooms)

Actually, two soft wares are available, both based on the application of the standard cited above; the one
BASTIAN (proposed by G+H in Germany) corresponds to the application of the standard with its informative annexes and especially to the models given as examples for the vibration insulation at the junctions.
The other ACOUBAT (made by CSTB in France) uses the standard in so far as it concerns the development of a determination method after an adaptation of the models in the informative part, using numerous
measurements of the vibration insulation at the junctions. Due to this adaptation, the second logical approach gives results very near to those one could obtain through measurements in place.
Note: In some European countries, acoustic regulation obliges the designer to achieve results in the area
of acoustic insulation between rooms. It is his responsibility to establish a construction which allows
for these results and measurements afterwards are frequently executed to verify the conformity to
the regulation.
These softwares allow for additionally, if this is needed a calculation in a relatively exact fashion of the
value of acoustic insulation loss Tf mentioned above.
In this chapter, only direct and flanking transmissions have been discussed. Parasitic transmissions will
be presented in chapter A4-2.5.
A4-2.2.3

Single numbers quantities

The European directive regarding building products demands to characterise the performances of products and of constructions by single quantities. These are defined in the standard EN ISO 717.
Before the publication of these standards, already most European countries expressed their results of
acoustic insulation by using single values. Disadvantageously, the criteria used were not identical and
there was no obvious link between the different methods.

10

Two principal methods competed with each other: the one which consists, as in France, in expressing the
overall level in dB(A) in the emission and the reception room and finding the difference, and the other
method which consists of a comparison of the level difference curve or the sound reduction index, as a
function of the frequency, and a reference curve.
For European harmonisation, it is the method with the reference curve which was retained. Thus, regardless of their earlier traditions, all countries must now use this method. Luckily, the European standards
have established certain terms of adaptation which allow to form a bridge between the two methods.
The differences are even aggravated when one realises that the analyses by frequency intervals, in use
in France up to the 1 January 2000 and the European analyses do not use the same frequency range. In
France, the laboratory measurements have been realised with third-octaves bands from the third-octave
centred on 100 Hz to the third-octave centred on 5000 Hz, which means six octave intervals from 125 Hz
up to 4000 Hz. In Europe, from 1 January 2000 onward it is required to make analyses between 100 Hz
and 3150 Hz in third-octaves and between 125 Hz and 2000 Hz in octaves. This has consequences for
the global single values.
The same rules must be used to express the sound reduction indexes: the R pink or road traffic in dB(A)
used in France are invalid and replaced by the Rw (C ; Ctr) in dB with
for pink noise at the emission, RA = Rw + C in dB
for traffic noise on roads at the emission, RA,tr = Rw + Ctr in dB
Assessment of a standardised acoustic level difference between two rooms
DnTA = DnT,w + C = RA + 10 log (0,32 V/S) Tf dB

(15)

Assessment of a normalised acoustic insulation between two rooms

DnA = Dn,w + C = RA 10 log (S/10) Tf dB

(16)

Note: The calculation of C and as a consequence of DnTA and DnA are realised taking into account the A
weighting. On the other hand, the results are given in dB and not in dB(A).
A4-2.3

The direct sound transmission through the separating wall

In all these methods, regardless whether simplified or complete, the expression of the direct transmission
of noise from one room to the neighbouring room is the same (see chapter A4-2.2.3).
If one uses the notion of normalised acoustic insulation: Dn (direct) = R 10 log (S/10) dB.
If one uses the notion of the standardised acoustic insulation: DnT (direct) = R + 10 log (0,32 V/S) dB.
A4-2.3.1

The influence of the dimensions of the reception room

For the normalised acoustic insulation, it is the surface of the separating wall, which the two rooms have
in common, that is important.
On the other hand, for the standardised acoustic insulation, it is the ratio V/S of the volume V of the re3
2
ception room in m to the surface S of the separating wall common to both rooms in m . Where the rooms
are exactly adjacent or above one another, the ratio V/S is directly related to the depth of the reception
room, taking perpendicularly from the separating wall.
Example:

Figure 4 represents a vertical cut through a building where only the sizes of the rooms vary.
The separating walls and the lateral walls are the same for all the rooms.

11

2m

DnTA - 2

3,2 m

5m

DnTA

DnTA + 2

(V/S = 3,2 m)

(V/S = 5 m)

2,5 m

(V/S = 2 m)

DnTA - 1
(V/S = 2,5 m)

Figure 4: Standardised level difference (DnTA) as a function of the dimensions of the reception
room. It is principally dependent on the ratio of the volume to the surface of the separating wall common to both rooms (V/S)
This example shows that the standardised acoustic insulation DnTA can vary by 4 dB when one considers
the smallest possible or respectively the largest possible reception room.
On the other hand, the normalised acoustic insulation is the same when one considers the horizontal
insulation (the surfaces of the separating walls are identical) and vary by 4 dB if one considers the vertical
acoustic insulation (Figure 5).
2m

3,2 m

5m

DnA

DnA

D nA + 1

DnA - 1

DnA - 3

2,5 m

DnA

Figure 5: Normalised level difference (DnA) as a function of the dimensions of the reception room.
It depends principally on the surface of the separating wall between the two rooms.
This variation of the direct transmission through the separating wall with the dimensions of the reception
room must be used for the rapid verification, whether in a building the acoustic insulation required by the
regulation or by the contract could be obtained with the design performances of the walls.
For a given objective of normalised acoustic insulation, one verifies whether it could be obtained by a
separating wall of the largest possible surface. For an objective in standardised acoustic insulation, one
uses the theoretical performance of the reception room of the smallest depth (V/S at a minimum).
One could also use the variation of the insulation as a function of the dimensions of the reception room by
increasing the performance with a consideration of the relative position of the emission and reception
rooms to each other.
The following example demonstrates what happens when the rooms are shifted.

12

Emission

Reception

Surface S of the
separating wall
common to both
rooms

Figure 6: For the rooms "shifted" the surface S of the separating wall is smaller than the total
wall surface in the reception room. The relation V/S of the volume of the reception room
to the surface of the separating wall common to both rooms is superior to the depth of
the room. The standardised or normalised insulation between the two rooms is higher
than in the case of these two rooms being either adjacent to or directly on top of each
other
A4-2.3.2

The influence of the sound reduction index of the separating wall

To determine this sound reduction index, it is useful to return to chapter A3-3 "Isolation of airborne noise"
in FESI Document A3 "Product characteristics Acoustic insulation, absorption, attenuation".
In this document, we give the possibilities to determine the sound reduction indices R of simple walls
(formed by a single element such as the walls or floors of concrete, walls of blocks or bricks, the separations of plaster elements or of plaster bricks), double walls (formed by several simple walls separated by
air cavities or by soft materials) and walls equipped with a lining complex, with the consequences more or
less favourable for the acoustic performance of the wall with the lining in front.
A4-2.4

The transmissions by the lateral walls

Knowing the dimensions of the rooms (study of the drawings) and the nature of the separating walls (see
the description), one has the means to determine the direct transmission between the rooms.
One now needs to examine the lateral transmissions which, let us remember, correspond to a dozen
paths of transmission as a result of three paths per junction.
The prediction methods envisaging the calculation of all transmission paths have been the aim of a recent
study which was laid down in the European standard EN 12354-1 "Calculation of acoustic performance in
buildings, deduced of the acoustic performance of building elements Part 1: The acoustic insulation of
airborne sound between rooms".
Due to the complexity of the problem, which is difficult to master for people not specialising on building
acoustics, it appeared to be necessary to develop a practical approach sufficiently simple for the use by
designers having at least a basic knowledge of acoustics.
How did one do before the introduction of the computer softwares?
One possessed a simplified method, based on the statistical exploitation of measurement results. This
method could still be used on the condition that the room configurations to be studied would be the same
as those which allowed for their establishment.

13

A4.2.4.1

Simplified method: Statistical evaluation of the sound reduction loss resulting from
lateral transmissions

This method has been established in the 60s for residential buildings, in an era when the buildings traditionally consisted of simple and heavy separating walls between rooms that needed isolation, not covered
by acoustic reinforcement and rigidly linked to lateral walls, being either simple or double, heavy or light,
covered or not covered with linings.
The equations used to appreciate the acoustic insulation are those given in chapter A4-2.2.3.
DnTA = RA + 10 log (0,32 V/S) Tf in dB
DnA = RA 10 log (S/10) Tf in dB
(19)
R'A = RA Tf in dB

(17)
(18)

The separating walls being simple and heavy of a surface weight m, one evaluates their sound reduction
index with the aid of the law of mass (see chapter A3-3 of document A3).
Regarding the insulation loss Tf due to flanking transmissions, the exploitation of a great number of
measurement results has allowed for the following assessment:

If the lateral walls are heavy: Tf = 5 dB (the flanking transmissions are two times as important as the
direct transmissions).
If the lateral walls are light and rigid, such as walls, linked to the separating wall made of plaster
2
blocks or thin bricks, one needs to add an insulation loss of 1 dB per 10 m of such walls.

Tf = 5 + Sr/10,
2

where Sr is the surface in m of the light and rigid wall linked to the separating wall in the reception room.
Honeycomb walls linked to the separating wall do not create any insulation loss and are disregarded
when calculating Sr.
In case a lateral wall is coated by a material which is acoustically ineffective, such as EPS, XPS or polyurethane rigid foam, the surface of the coating is taken into account when calculating Sr in the preceding
formula.
In case the lateral wall is covered by a lining improving the sound reduction index of its support such as a
composite with mineral wool or elastified polystyrene foam and a gypsum coat, the lateral insulation loss
is decreased by N dB, N being the number of the lateral walls covered by these composites.
This leads to a general formula which is apparently more a practical recipe than a scientific calculation:
Tf = 5 + Sr/10 N in dB
Below we are giving examples of the use of this formula and we demonstrate the limits of its validity.
A4-2.4.2

The limits of the simplified method

These limits will reveal themselves through the study of the following examples:

height under
ceiling 2.5 m

height under 3,43 m

ceiling 2.5 m

3,83 m

Figure 7: Sketch of two adjacent room of different flats

2,70 m

14

In this operation one should have obtained a standardised level difference between the flats of 50 dB, to
be in conformity with the acoustic regulations at the time of that construction.
These are two adjacent rooms, separated by a brick wall of 20 cm thickness, covered by plaster on both
sides. The facade is also of 20 cm bricks. In each of these flats, the internal walls of the flat linked to the
separating wall are made of thin hollow bricks of 5 cm, the floors are of 14 cm concrete.
This general description reflects well the type of construction that has served as the basis for the establishment of the formula above.
In order to make easy this example, we calculate only the standardised level difference.
DnTA = RA + 10 log (0,32 V/S) Tf in dB with Tf = 5 + Sr/10 N in dB

(20)

To obtain the sound reduction index RA of the separating wall: we use the mass law
3

RA = (40 log m) 47, with m = 0,20 1250 = 250 kg / m2 (the weight per volume of the bricks is 1.250 kg/m )
Thus, RA = 49 dB.
On the sketch in Figure 7, the relation V/S equals the depth d of the reception room, since the rooms are
exactly adjacent to each other ( V = S.d ) The acoustic insulation is the higher, the more important the
depth of the room is. We consider the least favourable case, that of the smallest possible room for which
the V/S is equal to 2,7 m.
From this one deduces 10 log (0,32 V/S) = - 0,6 dB.
There are no lateral walls covered by a composite on the basis of mineral wool. Therefore, N = 0.
On the contrary, in the reception room, there is a light and rigid masonry wall linked to the separating wall:
2
An internal wall of plaster bricks with a surface of 2,7 2,5 m2 , which gives Sr/10 = 0,7 m .
The foreseeable standardised level difference is then the following:
DnTA = 49 0,6 5 0,7 + 0 = 42,7 dB, rounded to 43 dB

(21)

The result of the measurement given in Figure 8 corresponds to a single value of standardised acoustic
insulation of 42 dB, a value quite near to the forecast but far from the expected value.
60

DnT (dB)

50

40

30

20
125

250

500

1000 2000 4000

Frequency (Hz)

Figure 8: Result of the measured standardised level difference DnT as a function of the frequency
The designer should have stated before starting the work that the performance foreseen led to a grave
non-conformity. Sadly, nobody made the calculation which we did just perform.

15

Whatever the reason, it was necessary to improve this insulation. To do this, the designer, believing he
could master the problem in this way, put a composite coverage on the separating wall, which was composed of 5 cm mineral wool and a plaster board.
The evaluation of the efficiency of such system allowed to hope for an improvement.
R = 30 RA/2 = 30 24,5 = 5,5 dB (see Table 3 of document A3)

(22)

But the result of the measurement had been disappointing with a standardised level difference 45 dB. It
has therefore been decided to replace the first cover by a process thought more effective, composed by a
plaster board mounted on a metallic frame, disconnected from the wall, and 80 mm of mineral wool. In
this case, R = 35 RA/2 = 10,5 dB. The result was 46 dB!
However, if one uses the formula for prediction, an insulation of 51 dB should have been achieved. In
effect, in this formula, only the acoustic attenuation index of the separating wall is modified by the coverage. The term V/S is not changed, since the coverage had been placed in the larger of the two rooms, the
emission room, the lateral transmissions had not changed since the surface of the light and rigid lateral
wall had remained the same. The foreseeable insulation in this method is therefore 42,7 + 10,5 = 53,2 dB,
which is a deviation of 7,2 dB from the result of the measurement.
In fact, this prevision method should not have been used since the separating wall was covered.
To use it nevertheless, amounts the presumption that the covering of the separating wall would have the
same effect on the lateral transmissions as on the direct transmission. In reality, the coverage of the
separating wall only affects one of the three lateral transmission paths at each junction (see Figure 9).
The two others remain unchanged.

Wall without lining

Wall with lining

Figure 9: The coverage of the separating wall only influences one of the three transmission paths
via the lateral wall
In this case, the direct transmission correspond to 48.4 dB (49 0,6) and the loss due to lateral transmissions is equal to 5.7 dB.
An analytical approach to the foreseeable result after the covering of the separating wall could possibly
be the following:
The acoustic insulation before the coverage (42.7 dB) is the result of a combination of direct acoustic
insulation Dd due to the transmission through the separating wall and a lateral acoustic insulation Dlat due
to the sum of the different transmissions via the lateral walls.

DnTA

Dd = RA + 10 log (0,32 V/S) = 49 0,6 = 48,4 dB

(23)
(
Dlat / 10 )
(
Dlat / 10 )

= 10 log 10
+ 10
= 42,7 dB (expected value below)

-4,27

Thus, one deduces Dlat = - 10 log (10

-4,84

10

) = 44 dB

(24)

16

Dlat - DnTA
3
2,5

dB

2
1,5
1
0,5

10

9,5

8,5

7,5

6,5

5,5

4,5

3,5

0
Dd direct - DnTA en dB

Figure 10: Determination of Dlat (corresponding to all flanking transmissions) when Dd (direct level
difference) and DnTA (final result) are known.
This lateral insulation is a combination of three lateral insulations (the three paths of flanking transmission). If one forms the hypothesis that the three lateral insulations are identical (this is the only possible
hypotheses without making exact calculations), each of the lateral insulations should be
Dlat1 = Dlat2 = Dlat3 = 44 + 10 log 3 = 44 + 4,8 = 48,8 dB

(25)

If the coverage placed on the separating wall would increase the acoustic attenuation index by 5,5 dB,
after the coverage the standardised overall acoustic insulation is the result of a combination of the four
insulations below:
direct insulation, Dd + 5,5 = 53,9 dB
lateral insulation 1, Dlat1 = 48,8 + 5,5 = 54,3 dB
lateral insulations 2 and 3, not changed by the coverage, Dlat2 = Dlat3 = 48,8 dB
The estimate of the standardised overall acoustic insulation after the coverage of the separating wall,
therefore, is:
-5,39

DnTA = - 10 log (10

measured.

+ 10

-5,43

-4,88

+ 10

-4,88

+ 10

) = 44,7 dB, which is a value close to the 45 dB actually

Combination of two level differences

Value to be added to the

smallest level difference
(D2)

10

-1

-2

-3
D1 - D2

Figure 11: Quantity to add at the smaller level difference D2 when we have two level differences to
combine D1 and D2
The use of the computer software explained earlier allows to avoid a relatively complex calculation.

17

Nevertheless, to approach the problem logically, one has every reason to make a sketch of the different
transmissions such as the one in Figure 9.
This example shows that the simplified method gives results quite near to those actually measured, as
long as one remains in the area of constructions with single, heavy and non-covered separating walls
between the rooms which must be isolated. Once the separating walls are double walls or once they are
light, the simplified method cannot be used. Let us nevertheless mention that its area of application covers more than 80% of the actual cases in residential buildings or hotels. On the other hand, in school
buildings, hospitals and offices, the required acoustic insulation is frequently very light, the floors and
facades are frequently heavy for structural reasons and the foreseeable acoustic insulation is frequently
near the direct acoustic insulation in the separating wall, the acoustic attenuation index of which is frequently lower than the one of the lateral walls.
A4-2.4.3

Calculation of flanking transmissions

The calculation method is described in the standard EN 12354-1 "Prevision of the performance of buildings on the basis of the performance of building elements Acoustic insulation between rooms".
In order to give an idea of the difficulty to the calculations needed, we give the case of lateral transmissions at a junction between two heavy walls, forming an "T" or an "X".
In all these cases, if one considers the flanking transmission of the wall i versus the wall j, the normalised
lateral insulation Dn,i,j takes the form of:
1/2

Dn,i,j = (Ri + Rj)/2 + Ri + Rj + Dv,ij 10 log (Si Sj) /A0

(26)

where Ri and Rj are the sound reduction indexes of the walls i and j. Ri and Rj are the efficiencies of
the coverages (linings?) met at the transmission path via the walls i and j. Dv,ij is the vibration insulation at
the junction. Si and Sj are the surfaces of the walls i and j. A0 is the reference equivalent absorption area
2
(10 m ).
1/2

With Dv,ij = Kij 10 log (lij/(Si Sj) ) in dB

where Kij is the vibration attenuation index at the junction and lij the length of the junction in m.

Figure 12: Types of junctions

Example of the "T" junction
2
2
K2,2 = 6,7 + 14,1 M + 5,7 M , where M = log (m1/m2), m1 and m2 being the surface masses in kg/m of the
walls 1 and 2 (index 1 is the one of the separating wall, perpendicular to the lateral wall).
2
K1,2 = K2,1 = 6,7 + 5,7 M
Example of an "X" junction
2
K2,2 = 8,7 + 17,1 M + 5,7 M
2
K1,2 = K2,1 = 8,7 + 5,7 M
Note: In the informative annex E of the standard EN 12354-1, the difference between the constant terms
Kij for the junctions in "T" form or in "X" form is 3 dB, despite the fact that in the above formulae it is
only 2 dB. In effect, if one considers all the transmission paths for acoustic energy being used by
the same sound pressure in the emission room, the number of transmission paths in a T junction is
equal to two thirds of the number of possible transmission paths in an X junction. This would justify
a difference of 1,8 dB (this being 10 log of 3/2) which is near 2 dB between the two vibration at-

18

tenuation indices. This difference of 2 dB has been confirmed by the measurement of these indices. This shows that one should consider the values given in the informative annexes of the standard as reference values and not as certified values.
If one considers the four junctions normally found in reality, one realises that a complete calculation of
acoustic insulation is long and cumbersome. This is why it appears to be preferable to use an approach
which permits a fast and automatic calculation. One of the conditions respected by the two approaches
cited above, is that the sum of entries for this formula is not too large. For these two approaches, it suffices to enter the dimensions of the rooms (see the design) and the nature and thickness of the walls
considered (see the description of the construction). The basis for the entries of these elements into the
formula are sufficient to cover about 90% of the cases actually occurring. For the remaining 10% it is possible to complete the basis of the calculation by the introduction of the characteristics of the missing elements.
Differently put, of great importance for the calculation of thirteen transmission paths between two rooms is
the possibility to form a hierarchy between these different paths, and, immediately decide which of those
are the most important ones. If the acoustic insulation foreseen is not sufficient, it is easy to determine
which of the walls must be modified.
A4-2.4.4

Special case of light double-partition walls

In case the separating walls are double and light walls, the problem of flanking transmission between two
rooms is more complex than in the case of single and heavy separating walls. Figure 13 shows the different flanking transmission paths which need to be measured. If lateral walls have a weaker performance
than the separating wall, the flanking transmission becomes preponderant and limits the acoustic insulation.
a
b
1
c
4
3

Figure 13: Different paths of transmission with double walls

The double wall:
a frame
b gypsum board
c mineral wool
The acoustic transmissions:
1 direct transmission through the wall
2 transmission in the floor d on the emission side to the floor on the reception side
3 transmission over the part of the separating wall on the side of the emission to the floor on
the side of the reception
4 transmission from the floor on the side of the emission to the part of the separating wall
facing the reception
For example, if two rooms of a depth of 3,2 m are divided by a separating wall on a double frame of the
type 160/110 (index RA or Rw+C of 59 dB) and if ceiling and floor are of 20 cm concrete and the facade is
of 15 cm concrete with an outer insulation (which has no influence on the acoustic insulation between the
rooms), the direct transmission through the separating wall corresponds to a standardised sound level

19

difference of 59 dB, the flanking transmission through ceiling, floor and the facade to an insulation of 51,5
dB. This leads to a rounded acoustic insulation of 51 dB. The replacement of the separating wall by a wall
of better performance 200/150 (RA = 64 dB) leads to an acoustic insulation of 52 dB. If one replaces the
wall 200/150 by a wall 240/190, which allows for a sound reduction index of 67 dB, the overall acoustic
insulation is 52 dB. In effect, when improving the acoustic reduction index of the separating wall without
changing the elements, but by increasing the gap between these elements, then the flanking transmission
remains the same. An improvement of the entire system can only be achieved by improving the performance of both, the separating and the lateral walls.
In this area, the optimisation of systems cannot be achieved unless one makes a very complex calculation or by using the approaches conform with the European prevision methods.
A4-2.5

The parasitic transmissions

As we have said earlier, the acoustic energy seeks all the transmission paths allowing it to travel from the
emission room to the reception room. We have examined the direct transmission through the separating
wall and the lateral transmissions via the lateral walls which the two rooms to be isolated have in common. But there is a third type of transmission in existence: the parasite transmissions. One example of
th
this type of transmission is that of a hole in the separating wall. If this hole does represent only 1/100000
of the wall, the acoustic attenuation index of the breached wall is limited to 50 dB, incompatible with the
attempt of an acoustic insulation between the two rooms of 53 or 55 dB. In case of a visible breach. The
solution is simple: it suffices to plug it as completely as possible.
One can discern between two types of parasitic transmissions:
Those which are expected, such as passages of pipes through the separating wall, ventilation ducts between rooms of different flats, plenums at ceilings of offices without acoustic barriers over the separating
walls, are called "parasitic".
Those which are accidental through lacking precautions or through insufficient attention during the construction work are called spurious transmissions.
A4-2.5.1

Expected parasitic transmissions

One example for the calculation of an parasitic transmission by a duct is given in document A6 in chapter
A6-2.3.5.1 (see Figure 14 below).

SSd1
g1
gaine
duct

g2
SSd2
gaine
duct

L2

L1

- 10
(S/ S
Ad1/SS
Sg2
- 03 + K(27)
LL22==LL
RdRg 10
loglog
(S A
)
3 +) K
1
d2g1
1 -2 2
0

K0 = 6 dB

K0 = 3 dB

K0 = 3 dB

Figure 14: The constant K0 depends on the placement of the duct in the wall of the room
In this example, the parasitic transmission through the duct leads to the following level difference:

20

L1 L2 = 2 Rd + 10 log (S A/Sd1 Sd2) + 3 K0

(28)

The normalised parasitic level difference Dne = L1 L2 10 log (A/10) = 2 Rd + 10 log (10 S/Sd1 Sd2) + 3
K0
where L2 is the level of radiation through the duct in the reception room in case a level L1 is emitted in the
emission room. Rd is the sound reduction index of the duct wall (dB). S is the cross-section of the duct
2
(m ). A is the equivalent absorption area in the reception room. Sd1 and Sd2 are the surfaces of the duct in
2
the rooms 1 and 2 (m ). K0 is a constant which depends on the position of the duct in the room.
This parasitic level difference being determined, it needs to be combined with the acoustic insulation Dn0
which would exist without the duct. The final acoustic insulation Dn is given by the subsequent equation

Dn = 10 log 10 (Dn0 / 10 ) + 10 (Dne / 10 ) dB

(29)

A similar calculation is to be made once each of the rooms to be isolated has a grid or an opening for
ventilation made in the same duct (this is for example the case with openings to extract the air of two
kitchens placed on top of each other).
The normalised acoustic insulation via the two grids is measured in a laboratory. The acoustic performance is given in the form of a Dne.
Dne = L1 L2 10 log (A/10) dB

(30)

If on the site one needs to respect a standardised insulation, it must not be forgotten to make the transposition from "normalised" to "standardised".
3

In case of a reception room of the volume V (m ), DnTe = Dne + 10 log (V/30) dB (see chapter A4-2.2.3).
A4-2.5.2

Accidental parasitic (spurious) transmissions

Of this one finds several examples, most frequently due to building mistakes. Some of these spurious
transmissions are difficult to find, even to imagine, once the construction is finished.
Amongst the most frequent one could quote the electricity plugs positioned face to face in the separating
wall, badly levelled out troughs of the distancers of the shuttering boards of a cast concrete wall, lacking
air tightness between the wall and the facade, deficiencies which in case of internal insulations are frequently masked by thermal insulation.
In all these cases, the spurious transmissions become obvious if one compares the results of a acoustical
measurements after completion with the expected values from the initial calculation. If one finds results
much worse than the previsions, one must verify the original calculation, and if no mistake is found one
needs to make a diagnosis on the site to find the reason for the loss of acoustic insulation.
To illustrate that investigation, we take a real example, showing the necessity of this diagnosis.
Example:

Two rooms on top of each other are separated by a floor of 16 cm concrete. The facade is
made of prefabricated concrete. Two of the walls of these rooms are made of concrete, the
third is made of plaster blocks (Figure 15).

21

Figure 15: Vertical cut of the junction of one facade of prefabricated concrete with integrated
thermal insulation and a floor
a Floor of 16 cm concrete
b Panel of prefabricated facade
c Skirting (base board)
d Wooden box fixed on a frame
Calculation of the foreseeable acoustic insulation
Since the separating wall between the two rooms on top of each other is heavy, simple and not coated,
one could use the simplified method of pre-calculation.
The sound reduction index of the separating floor:
R A = 40 log (0,16 2.350 ) 47 = 56 dB

(31)

The given dimensions: V/S = 2,5 m which leads to 10 log (0,32 V/S) = - 1 dB.
Losses by flanking transmission: Tf = 5 + 10/10 = 6 dB (the rectangular surface of plaster in the reception
2
room linked to the separating layer is 10 m ).
DnTA = 56 1 6 = 49 dB, which is already below the objective which was 50 dB.
The measurement has given a standardised acoustic insulation of only 42 dB.
The measurement of the acoustic insulation has been conducted with the help of a sound source and a
sonometer. This one is operated by a technician who also has ears and he can use it to give his impression. During this test, the technician could quite clearly perceive that the noise from the flat above appeared to come from the junction of the facade and the separating floor. He removed the wooden box in
the upper part of the window (this box was meant to have the same thickness over the window as the
window breast), he found out that between the concrete facade panel and the end of the separating floor,
there was a gap of more than 1 cm. After having disassembled one part of the skirting in the room above,
he found out that the space between the contact points of the facade panels and the separating floor
were open (this is what had been forgotten during the construction work). This way, a path for the acoustic energy between the two rooms had been created between the facade and the separating floor, with
the only barrier being constituted by the skirting and the wooden box. Before he put those back, he closed
the opening with well stuffed mineral wool. The new test, after this reinforcement had been completed,
gave a result of 48 dB, which was 1 dB below the forecast.
This example renders several lessons:

It is not enough to make a good choice of separating walls and lateral walls. Additionally, attention
must be paid to avoid creating spurious transmission paths.
An acoustic reinforcement of the separating floor would have been totally useless, the deficiency in
the acoustic insulation having nothing to do with any weakness on that part.

22

Since the acoustic insulation was abnormally weak, considering the materials used, one needed to
make a diagnosis with a view to find the predominant transmission paths.
The statistical prevision method gives a good evaluation of the acoustic insulation, since the separating floor was simple and heavy. But as is the case with all prevision methods, it could not possibly
take into account unexpected spurious transmission paths.
The measuring technician must note his impression, because several month after the measure he
didnt remember what he had felt.

If a weakness of this type cannot be localised as was the case in this simple example, then a combination
of acoustic and vibration measurements (impact sound measurements) can help to identify the spurious
transmission path.
A4-2.5.3

Which precautions to take?

It is necessary that the isolation foreseen between rooms is not degraded by spurious transmissions.
Walls constructed of plane or hollow concrete blocks, of brick or of stone, must be covered at least at one
face. It is not important whether the coverage is of gypsum or cement, given that the tightness of the wall
be assured. If one attempts to increase the surface mass, one should favour a cement cover instead of a
cover made of gypsum.
Walls made of concrete should be homogeneous. This is the great principle. Nevertheless, one must
accept certain local weaknesses, such as are constituted by the holes for the electrical plugs, the
branches drilled for electrical circuits, the hangers of lamps, the canalisation into the floors ... The rule to
observe is to guarantee that at the place of these deficiencies in the surface, the half thickness of the
concrete wall be maintained. For example, a box for an electricity plug in a 16 cm concrete wall is admissible as long as it leaves 8 cm of concrete behind it. This means that one could not put the two boxes on
either side of the wall back to back.
The plumbing shafts should be closed if possible in the walls or separating floors which they cross with
the help of a filling in material with a surface mass equivalent to that of the separation crossed. If this
cannot be done, the acoustic attenuation index of the shaft walls placed in the rooms to be isolated
should at a minimum be equal to the sound reduction index of the separating wall minus 5 dB.
The junctions between masonry walls must be tight. One must absolutely be interested in tightly connecting the heavy partition walls and the heavy lateral walls. This allows to take advantage of the energy
losses caused by the vibration occurring at the junctions. A problem threatens to arise here in case one
seeks to reduce the thermal losses at the junction between a wall or a floor and the facade. In this case a
thermal insulant must be placed between the wall and the facade (Figure 16).
This insulating material has the consequence to destroy the rigid liaison between "internal wall and facade" and to increase the flanking transmission via the facade. This could be compensated by using an
internal thermal insulation layer in the running facade which has also acoustical merits (mineral wool or
elastic polystyrene). On the other hand, the thermal insulation layer at the junction of the internal wall and
the facade could create a parasitic (spurious) transmission if it had a thickness below the space available
or if it was not compressed and not tight.

Figure 16: Positioning of an insertion between a wall and a facade with the aim to limit the thermal
bridge at the junction

23

A4-2.6

Acoustic insulation between rooms divided by a wall of different surface elements

These are foremost separating walls with doors, windows, service openings or other elements.
A4-2.6.1

Method of pre-calculation of the sound reduction index of composite walls

The surface of the wall in question is composed of several elements which have different sound reduction
indices (Figure 17). The problem is firstly the determination of the sound reduction index of the composition.

a
b

a-

partition of full concrete blocks of 10 cm thickness and

8 m2 surface (S1) and of an acoustic attenuation index
of 45 dB (R1)

b-

door of 2 m2 surface (S2) and of an acoustic attenuation

index of 20 dB(R2)

Figure 17: Example for a wall with a door

Each element has its acoustic transmission coefficient . In the emission room, the sound pressure reaching the surface S is equal to I S , where the acoustic intensity corresponds to the sound pressure level in
that room. It is no more than the part which travels towards the reception room. In this way, if one considers all the elements "i" which compose the surface of the separation wall, one gets the overall transmission factor of the composite wall when using the following relation:
S = 1 S1 + 2 S2 + ...
Knowing that Ri = 10 log(1/i), one possesses the means to calculate the overall sound reduction index of
the wall, provided one knows the sound reduction indices and the surfaces of each element.

Rres

= 10 log10

10
S

( R i / 10 )

Si

(32)

To avoid this calculation, a basic abacus has been established for that relation. It allows for the composition of these elements in pairs. We are proposing the abacus of Figure 18, which relates the differences
(R1 R2) of the indices of two elements, the ratio of the surfaces S1/S2 of these two elements and (R1
Rres) the difference of the sound reduction index of the best insulating element (R1) and the resulting index of sound reduction.
An example for application: the composition shown in Figure 17.
The horizontal line from R1 R2 = 25 dB, cut the vertical line for the relation S1/S2 = 4 on the curve R1
Rres = 18. As R1 = 45 dB, R = 27 dB.
One sees, contrary to what is often told, that the overall sound reduction index is not the same as the
index of the weakest element.
If one thinks that the result obtained with the example above is not sufficient, one cannot hope for an improvement by improving the wall without also changing the door.
Let us suppose that one realises an acoustical coating for the wall of the mineral wool type with a gypsum
board 60 + 10. The sound reduction index then comes near to 53 dB (R1 = 30 R1/2).

24

In this case, the new value of R1 R2 is 33, S1/S2 is unchanged and R1 Rres is 26 dB. This leads to Rres
= 53 26 = 27 dB, which is the same value as before the reinforcement.
In fact, due to the initial configuration, one has achieved what is called the limit of the sound reduction
index, which cannot be surpassed without changing also the door.

36
34

30

25

20

32
30

15

28
26

R1-Rres

24

10

22
2

-R
R1

20
18

16
14
12
10
8

R 1- R

6
4
2
0
0,130,160,2 0,250,320,4 0,5 0,630,8 1

1,251,6 2

2,5 3,2 4

6,3 8

10 12,516 20 25 32

S 1/S 2

Figure 18: Determination of the sound reduction index Rres, resulting out of the combination of a
surface element S1 with a sound reduction index R1 and of a surface element S2 which
has a sound reduction index R2
A4-2.6.2

Sound reduction index limit of a composite partition wall of two different elements

If one has two elements, one of which does not let anything through (transmission factor zero) and only
the transmission by the other element remains: then the limit acoustic attenuation index is given by this
formula:
Rlim = R2 10 log(S2/S)

(33)

where S is the overall surface S1 + S2.

In our example, if one retains the initial door, the acoustic reduction index of which was 20 dB,
Rlim = 20 10 log(2/10) = 27 dB!
This notion of a limit acoustic attenuation index (Figure 19) is very important and very useful. Here are
some practical consequences:
If one has to solve certain ventilation problems in a room, one foresees doors with gaps at floor level,
which makes it totally useless to install very costly high performance doors. If the gap is limited to 1 cm
th
(which is very small if one considers the actual experience), it represents roughly 1/200 of the door's

25

surface and has a sound reduction index zero. In this case, the limit sound reduction index of the door
with these gaps is 23 dB. This limit index is obtained when the opening is made into a door, which if it
were totally tight would render an index in the vicinity of 35 dB. Also, it is not necessary that the wall, into
which this door is built, should have an index better than 40 dB for a surface relation in the vicinity of 4.

Rlim-R2 (dB)
14
12
10
8
6
4
2
0
0

0,1 0,2

0,3

0,4

0,5

0,6 0,7

0,8

0,9

1 S2/S

Figure 19: Calculation of the limit sound reduction index of a wall in case the least insulating element (R2) is fixed. S2/S is the relation of the surface of that (least insulating) element to
the overall surface of the wall.
In case one has, in a wall, a permanent opening, or in case the door is open, the sound reduction index
does not become zero. In case of the preceding example, the open door in the wall leads to a reduction
index of 7 dB. For walls with openings one can also use the graph below (figure 20)

Sound reduction index with opening (dB)

area percentage of opening

Sound reduction index without opening (dB)

Figure 20: Sound reduction index of a wall with an opening in relation with the sound reduction
index of this wall without opening and the percentage of the opening area
In case, in a very well performing wall with an acoustic reduction index over 60 dB, one installs electricity
plugs opposite each other so that the local sound reduction index might be between 15 and 20 dB, the
2
overall index could be limited to 50 or 55 dB by this mistake of 5 to 6 cm diameter in a wall of 10 m .

26

A4-2.6.3

Acoustic insulation between two rooms separated by a wall with an opening

The opening could be considered as a source of parasitic transmission between the two rooms. In other
words, one assesses the acoustic insulation, which one would obtain if the separating wall were homogenous, in the absence of an opening. This allows to calculate the direct transmission and flanking transmissions. Once the opening is in place, the flanking transmissions do not change, only the direct transmission varies.
Two cases allow for an assessment with only the calculation of direct transmission through the wall with
an opening:

If the foreseeable isolation is low because of the poor performance of the separating wall without
opening (this is notably the case with separating walls having a small sound reduction index by comparison to the indices of the lateral walls).
In case the acoustic insulation without opening is high and the index of the wall with the opening is
low, the flanking transmissions can be disregarded compared to the direct transmission through the
separating wall and its weak point.

A4-2.6.4

Some considerations on doors

Doors are tested in the laboratory with very precise conditions of work incomparable to their actual employment in a building. Even more important, the measuring results rendered indicate that the joints between the door and its frame and the gap at the threshold in no case exceed 1 mm.
That sort of quality cannot be actually achieved on the site for since several dozen, possibly hundreds of
doors need to be installed. It is therefore prudent to reckon with a loss in the index of at least 2 dB, which
already requires a very painstaking installation. In other words, if the calculation shows that it is necessary to use doors with the index R, it is wise to choose doors which in the laboratory test have shown
indices of at least R + 2 dB.
To preclude severe differences between the reduction indices of doors given in the catalogue according
to a measurement taken in a laboratory, and the index to be found when a new measurement is taken in
the same laboratory on a door taken from the site, it is prudent to choose doors with a certification of the
performance and to require the manufacturer to guarantee that the doors delivered are in conformity with
those taken for the original measurement.
The weak points in the actual installation of doors chosen for their good laboratory performance are very
frequently the more or less tight openings between the door and its sides, and the deficiencies between
the frame of the door and the actual wall. One does not see these deficiencies once the installation is
finished, but one perceives them by measuring the acoustic performance afterwards. One must not have
too big confidence in the covering strips used to mask the joint between the wall and the frame of the
door.
To achieve a reduction index above 38 dB for the combination of the door and its frame, a very thorough
peripheral tightness is required (pressure locks may be needed), including the threshold. Because of the
unevenness of the ground, one cannot count on a sufficient tightness with retractable thresholds. Instead,
one must use thresholds in the form of ground plates, a beam equipped with an elastic sealant against
which the door is pressed.
In certain instances, especially when the doors are used as emergency exits, this system is not accepted
by the security commissions, since they could cause people to stumble and fall.
To obtain a strong insulation in case separating walls are equipped with doors, it is desirable to plan absorbent locks with at least one of the two doors of high acoustic performance.
In certain special cases, it is preferable to use mounting frames in the wall and to adjust the door and its
frame later in this mounting aid to limit the variations in the results essentially connected with the problems of tightness. It is also advised to enlarge the framework, respectively to use a double stop face.

27

RW of door
in dB

RW in dB
of tightly builtin door
leaves

22

25

27

30

32

37

37

42

42

47

47

55

Door leave type

Any construction, but about 15
kg/m2
One or more layers, about 25
kg/m2

Seal of functional joint

Any sealing profile

Remarks

The door must be well adjusted

Soft hose, chamber or lip seal

Multi-layered, about 35 kg/m2 or

multiple partition

Soft hose, chamber or lip seal,

possible double-seam seal

Multiple partition, heavy single

panels
Multiple partition made of very
heavy single panels (e. g. sheet
steel, chipboard with lead
sheet); double doors
Sandwich panels, made of very
heavy single panels using sandfilled chambers, about 150
kg/m2; double doors with sound
lock

Double-seam seal with soft

hose, chamber or lip seal

Inherently stable, densely builtin frame; sufficient contact

pressure
Inherently stable, densely builtin frame; high contact pressure

Double-seam seal with highquality hose, chamber or lip

seal

Very inherently stable, densely

built-in frame (steel frame);
very high contact pressure

Double-seam seal with highquality hose, chamber or lip

seal

Special frame; individual fitting

for each door; very high contact pressure

Table 1: Examples of door constructions to meet pre-set sound insulation indices (left column),
the minimum performance of the door leaves is in the second column
A4-3

Acoustic insulation against outside noise

A4-3.1

General

In most cases, the facade should isolate rooms against the traffic noise. Simple values allowing for the
qualification of standardised level differences between the outside and the interior of buildings, and allowing for the sound reduction index of an element, DnTA, tr or the RA, tr, to be obtained when assuming a
sound emission of the type "traffic" also called "road noise". In case no measuring results are available of
noise reduction indices for a traffic noise at its point of emission, and since no curve of frequency intervals
exists to calculate it, one uses a first approximation that the RA, tr is between 3 and 4 dB below the RA. For
the evaluation of a sound reduction index of a heavy wall as a function of its surface mass m, this leads to
the equation:
RA, tr = (40 log m) 50 dB
The example of facade is one application of the general problem of acoustic insulation between two areas
with direct transmission, flanking and parasitic transmissions and the problem of separating walls with
openings.
A4-3.2

The direct transmission of noise through a facade

The direct transmission is linked to the overall sound reduction index of the facade with exterior window
frames, calculated with the aid of the general formula (32) or with the abacus (Figure 18).
In the case of a relatively low insulation demand of a facade: the results are generally governed by the
windows since the wall of the facade is heavy. In this case, one might use the notion of sound reduction
index limit, Rlim (chapter A6-2.6.2), which allows for a rapid calculation.
In case of light facades, it is always necessary to make a complete calculation (chapter A4-2.6.1).
Regarding the expression of the direct transmission, the same rules apply as for the insulation between
two rooms.
In each frequency interval: Dn (direct) = Rres 10 log (S/10), where S is the surface of the facade in the
2
reception room (m ) and Rres is the overall sound reduction index of the facade, taking into account all its
components.
3

DnT (direct) = Rres + 10 log (0,32 V/S) dB, where V is the volume of the reception room (m ).
Regarding the unique values, it is preferable to use the values DnA, tr = Dnw + Ctr or DnTA, tr = DnTw + Ctr dB.

28

A4-3.3

The flanking transmissions in case of facades

The flanking transmissions depend on the nature of the floors, the walls and the separating walls linked to
the facade.
There is one transmission path less than in case of the acoustic insulation between rooms and in the
overall assessment the flanking transmissions create an acoustic insulation loss much smaller than in
case of the insulation between rooms (1 dB as opposite to 5 dB in the general case).

Figure 21: Special case of facades. The only flanking transmission paths are those via the internal
walls (a) which are linked to the facade
One comes to the following prediction equation:
DnTA, tr = RA, tr, res + 10 log (0,32 V/S) 1 dB

(34)

where RA, tr, res is the index resulting from the combination of the different elements of the facade (facade,
window frames) calculated as has been shown above.
In case of doubt, one can always turn to the computer programs of prediction described in chapter A42.2.2.2.
A4-3.4

Parasitic transmissions in case of facades

A4-3.4.1

The expected parasitic transmissions

Parasitic transmissions are frequently "imposed" by the necessity to create access for the air or to install
casings for rolling screens. These components of the facade are characterised in a laboratory by their
acoustic insulation Dne as a function of the frequency and the unique value Dne,w (C ; Ctr).
These values result from measurements of the normalised level differences which are obtained in a laboratory between two rooms separated by a well-insulating wall into which the equipment to be tested is
installed. It is assumed that the predominant transmission of acoustic energy takes place through this
2
piece of equipment. The result is normalised by an equivalent absorption area of 10 m in the reception
room.
To progress from a normalised insulation Dne to a standardised insulation DnTe, when the reception room
has a volume V, one uses the following equation:

29

DnT, e = Dne + 10 log (0,032 V) = Dne + 10 log (V/32) dB

(35)

The difficulty is linked to the fact that this parasitic transmission travelling through an acoustic insulation
does not have the same character as an acoustic reduction index and that these two values cannot be
combined.
Thus, the attempt consists of a calculation of the acoustic insulation given by the facade, without parasitic
transmissions, and to combine with this insulation the one given by the parasitic transmissions such as air
openings, casings for screens etc. It is therefore necessary to design the elements of the facade so that
they have an insulation superior (by at least 6 dB) to the target laid down in a regulation or in a contractual document.
In a first approximation, one could utilise the indications in the table below, if one wants to obtain a standardised acoustic insulation DnTA, tr for reception rooms with a depth of 3,2 m and with a window surface
of less than two thirds of the total surface.
Nature of the element
Facade outside windows
Windows
Assembly of air openings, casings for screens ...

Performance
RA, tr > DnTA, tr + 10 dB
RA, tr > DnTA, tr 2 dB
Dne global > DnTA, tr + 6 10 log (V/32) dB

We must consider the insulation of the total assembly of equipment. If one has two air openings of Dne =
36 dB, they have together the same performance as a single air opening of Dne 33 dB (two times the energy paths through the two openings compared to the one).

D
More generally, Dne,global = 10 log
ni10 ne,i
(35) where ni is the number of insulating ele

ments Dne,i and m is the number of different insulating elements Dne,i.

A4-3.4.2

Accidental parasitic (spurious) transmissions

Amongst the most important ones, the transmissions due to tightness faults between the opening and the
fixed parts of windows are quoted. These faults can be corrected by putting soft joints on the contact area
of the window frame. Nevertheless, nothing can replace a good adjustment of the elements of the windows. Another source of spurious parasitic transmissions is less visible: these are deficiencies in the
forms of gaps between the window frame and the masonry, frequently masked by covering strips.
A4-3.4.3

Some considerations on windows

The same precaution relative to the acoustic performances taken in A4-2.6.4 for doors must be observed
for windows. Because of the big difference between the laboratory and the site working conditions, it is
advised to choose windows with a laboratory sound reduction index R + 2dB if a sound reduction index R
is required for the window in the building.
The same care must be taken for the treatment of the junction between the masonry and the window
frame.
As for the nature of windows able to achieve an acoustic performance, it does not suffice to determine the
type of glass. This is very important when the window performance is near or up to 35 dB. In this case,
well adjusted wood or reinforced plastic frames should be used. Simple windows with aluminium frame
can achieve a performance just over 35 dB, but it is necessary to confirm the result with a laboratory
measurement.
For having a global sound reduction index near or up to 40 dB, it is advised, because generally less expensive, to use double windows, even with aluminium frames, one of the windows equipped with a double
glass and the other with a simple glass. The best results are obtained when the distance of the glasses of
two windows is between 10 cm and 15 cm.

30

Noise
insulation
class

R'W [dB] of
functionally
built-in
windows

25 20

30 34

35 39

40 44

Design features

glazing:
total thickness (both panes)
pane distance
RW glazing
seal
glazing:
total thickness (both panes)
pane distance
RW glazing
seal
glazing:
total thickness (both panes)
pane distance
RW glazing
seal
glazing:
total thickness (both panes)
pane distance
RW glazing
seal

Simple window
with single glazing

Simple window with

insulating glazing

Composite window
with 2 simple panes

Box-type window
with 2 simple panes

4 mm
27 dB
1 necessary

6 mm
8 mm
27 dB
not necessary

6 mm
no requirement

not necessary

no requirement
no requirement

not necessary

8 mm
32 dB
1 necessary

8 mm
12 mm
32 dB
1 necessary

8 mm
30 mm
1 necessary

no requirement
no requirement

no requirement

37 dB
1 necessary

8 mm
40 mm
1 + 2 necessary

no requirement
no requirement

1 necessary

45 dB
1 + 2 necessary

14 mm
50 mm
1 + 2 necessary

8 mm
100 mm
1 + 2 necessary

31

Noise
insulation
class

R'W [dB] of
functionally
built-in
windows

45 49

50

Design features

glazing:
total thickness (both panes)
pane distance
RW glazing
seal
glazing:
total thickness (both panes)
pane distance
RW glazing
seal

Simple window
with single glazing

Simple window with

insulating glazing

Composite window
with 2 simple panes

18 mm
60 mm
1 + 2 necessary

Box-type window
with 2 simple panes

12 mm
100 mm
1 + 2 necessary

General information for window constructions not possible

Table 2: Examples of window constructions to meet pre-set sound reduction indices

32

A4-4

Transmission of the noise created in a room to the outside Environment protection

A4-4.1

General

This is a problem inverse to that treated in the last chapter. It is now not about the control of the noise
which penetrates into a room, coming from the outside, but instead to ascertain that the noise produced in
a room does not annoy the neighbourhood.
The noise level produced in a room in certain cases may be relatively high and may radiate through the
walls of the room to the outside and so cause a nuisance to the environment. This can be the case when
noise is generated in industrial buildings, in noisy ateliers, in technical installation rooms in office buildings, in musical rehearsal rooms and so on.
This problem is treated in prEN 12354, part 4 "Transmission of indoor sound to the outside". This standard only deals with the calculation of the sound energy radiating through the envelope of a building when
one knows the nature, the dimensions and the acoustical characteristics of that envelope. In reality, the
complete problem consists of three parts:
Determination of the interior noise in the room, knowing the sources of noise and their acoustic energy,
and the characteristics of the room, especially the more or less absorbent qualities of its walls. This can
be done using FESI Document A5 (Acoustics in rooms) and A6 (Industrial acoustics).
Determination of the sound pressure at a point situated at the outside, at a certain distance from the
building. This is the domain of sound propagation of an exterior sound with the calculation of the acoustic
field directly stemming from the source, the acoustic reflections of other buildings, the positioning of obstacles, the effect of the soil, the absorption in the air and so on.
A4-4.2

General principles

The sound radiated by the actual envelope of the building is simulated as if it were radiated by a representative point source. Each point source represents the contribution of a section of the building envelope. The number of point sources needed depends on the distance of the exterior reception point and
the variations of the propagation effects.
The level of acoustic pressure in octave bands at a reception point outside a building is determined by
starting from the contribution of each representative point source following the equation:
Lp = LW + Dc Atot dB (36)
where
Lp is the sound pressure level in dB at the reception point caused by the acoustic radiation of the representative source.
LW is the sound power level in dB of the representative point source.
Dc is the correction of the directivity of the representative point source in the direction of the reception
point, also expressed in dB.
Atot is the total attenuation in dB which is created along the propagation path from the representative point
source to the reception point, through distance, absorption by the air, any obstacles and reflections.
A4-4.3

Determination of the representative point sources

The elements contributing to the overall sound radiation are divided into two groups:

The plane radiators, such as facades, roofs, windows, doors, which are composed out of small elements, such as grids or holes.
The great openings such as open doors or windows and big ventilation openings.

The segments of the building represented by the representative point sources must satisfy the following
conditions:

The total attenuation of the propagation Atot towards the nearest reception point must be the same for
all the elements of a segment.

33

The distance to the nearest reception point must be more than twice the largest dimension of the
segment.
All the elements of a segment are subject to the same interior sound pressure level in the building.
All the elements of the segment must have the same directivity.

Finally, the position of the representative point source, corresponding to a vertical segment, must be two
thirds of the height of the segment and be placed in the middle of its width. In the case of non-vertical
segments, the point source is placed at the centre of the segment.
A4-4.4

Calculation of the sound power level of a representative point source

A4-4.4.1

The case of a segment consisting of structural elements of the building envelope

(group of plane radiators)

The sound power level of structural elements is given by:

LW = Lp,in + Cd R' + 10 log (S/S0) dB

(37)

where
Lp,in is the sound pressure level at 1 or 2 m away from the segment at the inside of the building, in dB.
Cd is the diffusivity term for the internal acoustical field at the level of the segment, in dB.
R' is a sound reduction index appearing at the segment, in dB.
2
S is the surface of the segment in m .
2
S0 is a reference surface (= 1 m ).
Evaluation of the diffusivity term Cd: The measurement of a sound pressure level inside a room, near a
wall is influenced by the more or less absorbing nature of that wall. The provisional standard quoted
above gives a table of values to be used in most common cases.
Place of the measuring point
In front of an reflecting surface in a relatively small room with a diffuse acoustical field
Before an absorbing surface in a relatively small room with a diffuse acoustical field
Large, flat volumes or long halls with many sources (industrial work area) in front of a reflecting surface
Industrial workplace with few dominant sources in front of a reflecting surface
Industrial workplace with few dominant sources in front of an absorbent surface

Table 3: Diffusivity terms Cd for different rooms

Evaluation of the apparent sound reduction index at the segment:
m +n
m S

A0
/ 10
D
i
R' = 10 log10
10 R i / 10 +
10 n,e,i dB
S
i = m +1
i =1 S

where
Ri is the sound reduction index of the element i in dB.
2
Si is the surface of that element in m .
2
S is the surface of the segment in m .
Dn,e,i is the normalised acoustic insulation of a small element i in dB.
2
A0 is the reference equivalent absorption area (= 10 m ).
m is the number of elements on the overall surface of the segment.
n is the number of small elements of the segment.
A4-4.4.2

For an open segment ("great openings" as given in chapter A4-4.3)

i =k

L W = Lp,in + 10 log10

Si

S 10

D i / 10

i =1

where
2
Si is the surface of the opening i in m .
2
S is the surface of the sum total of the openings of a segment in m .

dB (39)

(38)

Cd in dB
-6
-3
-5
-3
0

34

Di is the attenuation of a possible silencer in the opening i in dB (if there aren't any silencers Di = 0).
k is the number of openings in the segment.
A4-4.5

Determination of the correction of the directivity of a representative point source

The correction of the directivity Dc corresponds to the actual directivity of the radiating element or opening, characterised by the directivity index Dl (in principle, 10 log Q with Q = directivity of the source). It
could also be completed by the effect of neighbouring hard surfaces, characterised by the solid angle
index D,
where is the solid angle in which the radiation develops, in steradians.
Dc = Di + D = Di + 10 log (4 /)
If the reflecting surface is the building itself, the solid angle equals 2 and D equals 3 dB. In practice, in
front of plane, Dc varies between 5 dB and +5 dB; one could take the average Dc = 0 dB for radiation
angles between 0 and 90 relative to the normal on the plane.
A4-4.6

Example for the application

Two buildings are 21 m apart. The one to the left is an office building which on the last level has a technical installation with ventilation and air-conditioning. The building on the right is a residential building.
Dimensions of the technical installation: 12 m long, 7.2 m deep and 3 m high.
The technical installation has a great opening at the facade as an air intake, equipped with a grid, which
2
is 6 m long and 1.5 m high. On the roof, a brick air outlet, equipped with a grid of 1 m , makes the communication of the room air with the outside space.
Before the construction of the office building, it was important to verify that the noise coming from the
technical installation would not constitute a source of irritation for the occupants of the residential building.
To ascertain that, an exterior reception point was chosen right opposite the centre of the technical installation on the last level of the residential building. This point is exposed to the noise coming from the grid
of the air intake and is opposite the grid of the air outlet.
In the provisional standard, it is indicated that the maximum dimension of the segment should be below
d
2 , where d is the distance from the reception point to the segment. In our case, the largest dimension
4
of the segment should not be above 7.4 m.
For the facade of the technical installation, one could consider 5 segments marked by the letters E to l in
the sketch of Figure 23. Regarding the roof, more homogenous, 4 segments marked from A to D suffice
(Figure 24). The stars in the two figures indicate the places of the representative point sources.
The origin of the co-ordinates shown in Figures 23 and 24 is the reception point, situated in the axis of the
technical installation at 1 m below the roof of that room and at a distance of 21 m from its facade.
The walls of the technical installation are of 14 cm concrete. The roof is light (perforated steel sheet, mineral wool with multi-layer water protection). The grids of the air intake and the outlet are made of 2 cm
bars and the openings in between are 10 cm.

35

21 m

Figure 22: Example Noise generated by a technical installation

z

F
E

- 1.25

-2
- 2.25

- 3.5

y
4.5

- 4.5

Figure 23: Segmentation of the facade of the technical installation

x
A

27.60

4.5

1.5

- 1.5

- 4.5

Figure 24: Segmentation of the roof of the technical installation

Frequency (Hz)
Nature of walls
14 cm concrete
Roofing

125

250

500

1000

2000

4000

35
24

44
32

52
33

58
40

66
48

72
49

Table 4: Frequency dependent sound reduction indices of different walls in dB

36

Calculation of the sound power level on each of the segments of the roof:
Roof: The segments A, B and C are of the same dimensions and are subject to the same interior sound
pressure level. The segment D has a smaller surface because of the presence of the air outlet.
Segments
1
2
3
4
5
6
7
8
9

All the segments

A, B, C of the surface of
21.6 m2
D of the surface of
19.35 m2

Extension
Lp,in (1)
Cd (2)
R
10 log (S/S0)
Dc (3)
LW + Dc (4)
10 log (S/S0)
Dc (3)
LW + Dc

125 Hz
68
-3
24
13.3
0
54.3
12.9
0
53.9

Middle frequencies at octave intervals

250 Hz
500 Hz
1000 Hz
2000 Hz
68
57
49
39
-3
-3
-3
-3
32
33
40
48
13.3
13.3
13.3
13.3
0
0
0
0
46.3
34.3
19.3
12.9
12.9
12.9
12.9
12.9
0
0
0
0
45.9
33.9
18.9
0,9

4000 Hz
40
-3
49
13.3
0
12.9
12.9
0
0,9

Table 5: Calculation of the sound power level on each of the segments of the roof
Commentary:
(1) The interior sound pressure level results from a calculation starting with the sound power levels of
machines installed in the room, the absorption coefficients of the walls and the spacing conditions of
the room (see document A6).
(2) The diffusivity term Cd is the one foreseen in Table 3 in chapter A4-4.4.1, since the wall is covered
with an absorbent material.
(3) The directivity correction Dc corresponds to the performance used where the segment is plane and
reflecting.
(4) LW + Dc corresponds to the lines 1 + 2 3 + 4 + 5.
Facade: The segments E and l are comparable (same surface, same walls, not absorbent at the inside,
same sound pressure on the inner surface), the segments F and G are similar, the segment H corresponds to the grid of the air intake.
Segments
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

Extension

Lp,in
Segments E, F, G and l
Cd (1)
R
10 log (S/S0)
Segments E and l of the
Dc (2)
2
surface of 9.0 m
LW + Dc
10 log (S/S0)
Segment F and G of the
Dc (2)
surface of 4.5 m2
LW + Dc
Lp,in
Cd (3)
Segment H of the surface R (4)
of 9 m2
10 log (S/S0)
Dc (5)
LW + Dc

125 Hz
68
-5
35
9.5
3
40.5
6.5
3
37.5
68
0
1
9.5
8.5
85

Middle frequencies at octave intervals

250 Hz
500 Hz
1000 Hz
2000 Hz
68
57
49
39
-5
-5
-5
-5
44
52
58
66
9.5
9.5
9.5
9.5
3
3
3
3
31.5
12.5
- 1.5
- 19.5
6.5
6.5
6.5
6.5
3
3
3
3
28.5
9.5
- 4.5
- 22.5
68
57
49
39
0
0
0
0
1
1
1
1
9.5
9.5
9.5
9.5
9
9
9
9
85.5
74.5
66.5
56.5

4000 Hz
40
-5
72
9.5
3
- 24.5
6.5
3
- 27.5
40
0
1
9.5
9
57.5

Table 6: Calculation of the sound power level on each of the segments of the facade
Commentary:
(1) The diffusivity term Cd is the one foreseen in Table 3 when the wall is reflecting.
(2) One is opposite a plane reflecting wall.
(3) One can consider that at the interior of the technical installation the grid of the air intake is similar to
an absorbing product.
th
(4) The openings in the grids represent 4/5 of the overall surface. One applies the equation for the limited sound reduction index (see chapter A4-2.6.2, equation (...)).
(5) The directivity of the grid is determined using the abacus in chapter A4-....

37

9
8

Coefficient de directivit Q

6
5

4
3

(a)

1
0
1

21

Coefficient de directivit Q

7
6

5
4

2
4

(b)

Location of the outlet

0
10

12,5

16

20

25

31,5

40

50

63

80

100

125

160

200

250

315

400

500

630

1 The duct crosses through the wall

2 On the wall far from other walls
3 Near the angle of two walls
4 Near a corner

800 1000

f * (S)1/2

Figure 25: Directive effect Q as a function of the product of the frequency in Hz and the square of
the sound outlet area S in m2
a) is for an oblique direction from the centre of the outlet and the reception point
b) is for a normal direction
In our case, Dc = 10 log (Q), the direction is "normal", the outlet is in the middle of the wall (case 2),
1/2
S = 3 and the values of Q are given in Table 7.
f
f * S1/2
Q
10 log Q

125
375
6,8
8,5

250
750
7,3
9

500
1500
8
9

1000
3000
8
9

2000
6000
8
9

4000
12000
8
9

Table 7: Directivity factors' calculation for segment H

Grid of the air outlet: This grid is on a brick chimney. Its height is 90 cm and its width 1.2 m. The chimney
has an opening of the same dimensions as in the technical installation. The sound power entering the
chimney is equal to the sound pressure level in the room, increased by 10 log (S/S0). This acoustic pressure is partly consumed by the two effects of the elbow (one at the entrance of the chimney and one at its
exit), by the chimney itself (straight length of 4 m) and by the reflection at the entrance. It is this acoustic
pressure which is taken into account in the calculation of the table below.

1
2
3

Segments

Extension

Grid of air outlet

LW, grid
Dc (1)
LW + Dc

125 Hz
60
7
67

Middle frequencies at octave intervals

250 Hz
500 Hz
1000 Hz
2000 Hz
54
46
43
33
8
8.5
9
9
62
54.5
52
42

4000 Hz
34
9
43

Table 8: Calculation of the sound power level of the air outlet

Commentary: As in the case of the grid at the facade, the directivity index has been calculated with the
help of the abacus (Table 7).

38

Calculation of the sound pressure level at the reception point

For each representative point source, the sound pressure level produced at the reception point is
2
Lp = LW + Dc 10 log (4 d ), where d is the distance of the representative point source from the reception point.
2

With d = x + y + z , x, y and z being the co-ordinates of the point source, the point of reception being at
x = 0, y = 0 and z = 0.
Point source
A
B
C
D
E
F
G
H
I
Grid of the outlet

x
24.6
24.6
24.6
24.6
21
21
21
21
21
26.7

y
4.5
1.5
- 1.5
- 4.5
4.5
0
0
0
- 4.5
5.25

z
-1
-1
-1
-1
-2
- 1.25
- 3.5
- 2.25
-2
1.5

d2
626
608
608
626
445
443
453
446
445
743

10 log (4 d2)
39
39
39
39
38
37
38
37
38
40

Table 9: Calculation of 10 log (4 d )

Frequency (Hz)

125

Point source
A
B
C
D
E
F
G
H
I
Grid of the outlet

15
16
16
15
3
0
0
48
3
27

250

500

1000

2000

4000

7
-5
8
-4
8
-4
7
-5
-6
- 25
-9
- 28
-9
- 28
48
37
-6
- 25
22
15
Overall level

- 20
- 19
- 19
- 20
- 39
- 42
- 42
29
- 39
12

- 38
- 37
- 37
- 38
- 57
- 60
- 60
19
- 57
2

- 38
- 37
- 37
- 38
- 62
- 65
- 65
20
- 62
3

Frequency (Hz)
dB(A)
2
3
3
2
- 11
- 14
- 14
41
- 11
19
41 dB(A)

Table 10: Calculation of the expected frequency-dependent sound pressure level in dB / dB(A) at
the reception point
This level of 41 dB(A) is liable to be perceived by the occupants of the residential building which are most
exposed, especially at night, in the summer, when the windows of the living areas could be opened.
Note:

These calculations are frequently realised with the help of a programme integrated into a calculation table. When the programme is well designed, it is not more complicated to calculate the level
originating from 10 representative point sources than it would be to calculate the levels stemming
from 2 or 3.

However, it is possible to limit the number of calculations by making a first approximation of the level expressed in dB(A).
Knowing the spectrum of the noise at the interior of the location and the sound reduction index of the
walls as a function of the frequency level in dB(A) for an emission noise can be calculated and an estimate of the sound pressure level at the exterior, opposite each wall can be made. Using an average distance of the walls to the reception point, one has an estimation, possibly rough, but frequently sufficient,
of the level which would be perceived at the reception point. For the complete calculation, one does only
require the elements giving a significant value of sound pressure level.

39

Hz
Emission noise Lp,in
R concrete, 14 cm
Transmission level Lp,ext
R roofing
Transmission level Lp,ext

125
68
35
33
24
44

250
68
44
22
32
36

500
57
52
5
33
25

1000
49
58
-9
40
9

2000
39
66
- 27
48
-9

4000
40
72
- 32
49
-9

dB(A)
61
18.5 (1)
31 (2)

Table 11: Example used on the preceding case

Commentary:
(1) Even if one increases this level by 10 log S, where S is the overall surface of the concrete parts of the
2
facade (in this case 27 m ), one obtains a sound power level of 33 dB(A), which at a minimum distance of 21 m (distance of the listening point from the wall, respectively the facade), one gets a sound
pressure level of less than 0 dB(A).
2
(2) Even if one increases that value by 10 log S, where S is the total surface of the roof (be it 86.4 m ),
one obtains a sound power level of 50 dB(A), which at a minimum distance of 21 m (distance from the
listening point to the roof, respectively the facade) would give a sound pressure level of less than 10
dB(A).
Thus, one can avoid the calculation of the point sources A, B, C, D, E, F, G and I. Only the point sources
corresponding to the grid of the air intake on the facade and the air outlet on the chimney on the wall
should merit a complete calculation.
A4-5

Impact sound insulation

A4-5.1

General

In buildings, the bigger part of regulations and recommendations concern themselves principally with
impacts on floors, produced in the generally accessible parts of residential areas. The origin of these impacts foreseen are walking individuals, the fall of objects, the pushing of chairs ...
The impact on a floor relates to the wall an energy quite superior to that associated with the impact of an
airborne acoustical wave on that wall. The impact energy is propagating through the entire floor and the
walls connected with it with a speed which is the speed of sound in the materials concerned.
The impact noise is not only transmitted to the room situated just below the room where the impact noise
is produced (vertical transmission), one must also consider the transmission to the adjacent room at the
same level (horizontal transmission) or at the next lower level (diagonal transmission) seen from above
(Figure 25).

Emission room

b
a

Figure 26: Impact sound is propagated not only as vertical transmission (a) in the room situated
below the emission room, but also as diagonal transmission (b) or as horizontal transmission (c)

40

A4-5.2

Principles

The impact sound insulation is defined by the sound level measured in a room when the floor of another
room is excited by a machine generating standardised impacts. This machine carries five cylindrical
hammers of 500 g, falling from a height of 4 cm on the floor, making ten impacts per second. This is not
totally representative of the impacts made by a walking person, but it has the advantage of allowing for a
repeatable test. Let us recall that the repeatability is indispensable to allow for a pre-planning of the dispositions to take to reach an objective fixed by a regulation or an contractual agreement. Let us also note
that the hierarchy of solutions tested by the impact machine is generally close to those experienced by
actual impacts produced by occupants of locations.
As with airborne sound, there is a direct transmission through the floor separating the two rooms located
above each other and lateral transmissions by the walls linked to that floor in the reception area (Figure
26).

Figure 27: Transmission of impact sound. There is only one lateral transmission at the junction of
the floor and the walls, separating walls or facade
1 transmission direct through the floor 2 lateral transmission
A4-5.3

The level of impact sound

In a first attempt, we examine the vertical transmission, afterwards we evaluate the differences we need
to take to obtain an assessment of the horizontal and diagonal transmissions.
A4-5.3.1

Terminology to be used

Impact sound pressure level

This is the sound pressure level L as a function of the frequency, measured in the reception room in
octave intervals or in third-octave intervals. In laboratories, theses are conducted at third-octave intervals. In actual buildings, these tests are generally carried out in octave intervals.
Normalised impact sound pressure level
Ln = L + 10 log (A/A0) in dB (40)
2
A is the equivalent absorption area in m found in the reception room during measurements. It is determined using the SABINE equation (T = 0,16 (V/A)) after having measured the reverberation time in
the reception area.
2
A0 is a reference equivalent absorption area fixed at 10 m , unless otherwise indicated in a regulation
or in a contractual agreement.
Standardised impact sound pressure level
LnT = L 10 log (T/T0) in dB (41)
T is the reverberation time in seconds established in the reception room during measurements.

41

T0 is a reference reverberation time fixed at 0,5 seconds, unless otherwise indicated in a contractual
agreement.
One can compare the standardised pressure level LnT and the standardised Ln. If T0 = 0,5 s and A0 =
2
10 m , then:
3
3
LnT = Ln 10 log (V/V0), where V0 is the reference room volume of 31,5 m , rounded to 30 m .
This is equivalent to: LnT = Ln 10 log (V) + 15
In laboratories, the impact sound is characterised by the normalised impact pressure level Ln.
In actual buildings, it could be characterised by the pressure level normalised or standardised. But
unlike in laboratories, where the configuration can be studied so that no lateral transmissions occur,
the walls linked to the floor to be tested contribute to the energy transmission towards the reception
area (Figure 27). This is why in actual measurements the normalised pressure level of structureborne sound is called Ln and the standardised is called LnT.

Laboratory

Tapping machine

Figure 28: Measurement of impact sound in a laboratory, without lateral transmissions

Measurements of impact sound are made according to the ISO series of standards 140. There are other
ISO standards (717) dealing with a single-number value to characterise the impact sound pressure level.
The standard ISO 717-2 sets the definition of the impact sound pressure level by a single-number value,
obtained from a comparison of the sound spectrum measured by frequency intervals to a reference spectrum. For the measurement in third-octaves, one only takes into account the intervals between those centred on 100 Hz and those centred on 3150 Hz. For the octaves, the intervals used are centred on 125,
250, 500, 1000 and 2000 Hz. The symbols to be used are repeated below:
Characterisation in a laboratory: starting from the values in third-octaves
Single-number value
Weighted normalised impact sound pressure level Lnw

Origin of calculation
Ln in dB by third-octaves

Characterisation in actual buildings: starting from the values in third-octaves or in octaves

Single-number value
Weighted normalised impact sound pressure level L'nw or Weighted
standardised impact sound pressure level LnTw

Origin of calculation
L'n or LnT in dB by octave or
third-octaves

42

A4-5.3.2

Efficiency of floor coverings

A floor covering put in place on the floor normally allows to lessen the impact sound transmission, be
carpeting being more or less soft, or in case of hard floors having a soft under-layer. The efficiency for
these coverings is characterised by the difference L between the normalised pressure level Ln0 measured on the naked floor and the level Ln measured on the covered floor (see ISO 140-8 Acoustics
Measurement of sound insulation in buildings and of building elements Part 8: Laboratory measurement
of the reduction of transmitted impact noise by floor coverings on a standard floor).
To obtain the single value of the performance of floor coverings to use, one applies the L obtained in
each frequency interval to a reference floor. The difference between the weighted impact pressure level
Lnw for the reference floor before the application of L and the value of the weighted impact sound pressure level after the application is the LW "weighted reduction in impact sound pressure level".
A4-5.4

The prediction of the impact sound pressure level

For floors made of homogenous concrete or of wide slab concrete see Figure 28 and, under certain conditions of timber-framed concrete or brick, the standard EN 12354-2 "Estimation of acoustic performance
of buildings from the performance of elements Part 2: Impact sound insulation between rooms" proposes the equation below:
L'nTw = 181 35 log (m0) - Lnw 10 log (V) + K

(42)

m0 is the surface mass of the floor in kg/m , Lnw is the efficiency of the floor covering in dB, V is the vol3
ume of the reception room in m and K is the level addition caused by lateral transmissions.
A4-5.4.1

Lateral transmissions
Floor
m0
(kg/m2)
100
150
200
250
300
350
400
450
500
600
700
800
900

Average surface mass of homogenous lateral elements without

thermal acoustic linings (kg/m2)
100
150
200
250
300
350
400
450
500
1
0
0
0
0
0
0
0
0
1
1
0
0
0
0
0
0
0
2
1
1
0
0
0
0
0
0
2
1
1
1
0
0
0
0
0
3
2
1
1
1
0
0
0
0
3
2
1
1
1
1
0
0
0
4
2
2
1
1
1
1
0
0
4
3
2
2
1
1
1
1
1
4
3
2
2
1
1
1
1
1
4
3
2
2
1
1
1
1
4
3
3
2
2
1
1
1
4
3
2
2
2
1
1
4
3
3
2
2
2
2
2

Table 12: Values of K, correction terms to cater for lateral transmissions

Since the lining of a wall has a negative acoustical effect, one adds 1 dB to the term K.
Generally, K is most frequently set at between 0 and 2 dB. This is sufficiently precise, an exact calculation
would be very long and does not render a significant improvement in the overall result.
A4-5.4.2

Effect of the room dimensions

We have seen that in case of the airborne sound level difference between two rooms the standardised
acoustic insulation depends on the relation V/S of the volume of the reception room to the surface of the
separating wall common to the two rooms. In case of impact sound, the standardised pressure level of
depends on the volume V of the reception room.
As a consequence, the performance of coverings necessary to obtain a level below the limits required by
a regulation or by a contractual agreement can be weaker in case of large rooms compared to rooms of a
smaller volume. This is why one can more readily use parquets and tiles on supporting layers in large
halls than in smaller rooms.

43

A4-5.4.3

Behaviour of naked floors regarding impact sound

The methods of prediction proposed in the standard EN 12354-2 use an estimation of the normalised
weighted sound pressure level of impact sound Lnw for a homogenous floor by the equation:
Lnw = 164 35 log (m0) in dB

(43)

Table 13 demonstrates that for both the airborne and the impact sound for massive floors the predicted
values are close to the values measured in a laboratory. On the other hand, it seems that for walls made
of hollow wide slabs concrete the predictions based on the surface mass of the wall are quite optimistic
for the impact sound and quite pessimistic for the airborne sound. For floors of steel beams complemented by concrete or ceramic, the cavities which they contain create parasite resonance frequencies
which make all predictions accidental.
Nature of floor
Concrete 14 cm
Concrete 16 cm
Concrete 18 cm
Concrete 20 cm
Concrete 22 cm
Concrete 25 cm
Concrete 25 cm
Alveolate concrete
265 W + 60
Alveolate concrete 200 W +
80
Alveolate concrete 160 U
Concrete hollow blocks
16 + 4
Concrete hollow blocks
8 + 13
Concrete hollow blocks
8 + 13 + 5
Ceramic hollow blocks
12 + 4
Ceramic hollow blocks
16 + 4

m0
kg/m2
329
376
423
470
517
564
588

measured
78
76
73
71
69
67
66

LnW in dB
predicted
76
74
72
70
69
68
67

difference
-2
-2
-1
-1
0
1
1

measured
54
57
59
61
62
64
65

RW in dB
predicted
54
57
58
60
62
63
64

difference
0
0
1
1
0
1
1

500

76

70

-6

63

61

430

73

72

-1

64

59

283

83

78

-5

56

52

310

87

77

-10

51

53

-2

420

80

72

-8

57

58

-1

530

76

69

-7

60

62

-2

260

94

79

-15

46

51

-5

320

91

76

-15

49

54

-5

Table 13: Impact versus airborne sound insulation measured and predicted for different floor
constructions
In conclusion, one can not sensibly use the methods of prediction (law of mass for the airborne and the
formula above for the impact noise) as in the case of solid floors made of concrete.

See A3-3.3.1.1

44

Figure 29: Types of basic floor constructions

A4-5.4.4

Floor coverings

The efficiency is measured in laboratories by the determination of their LW .

One discerns three great categories of coverings: thin floor coverings, hard floor coverings placed directly
on a soft under-layer and floating floors constituted by a layer of mortar, concrete, synthetic anhydride ...
placed on compressible mats.
A4-5.4.4.1

Thin floor coverings

They do not pose special problems in application and can have a very good acoustical efficiency. On the
other hand, possibilities of choice of a product type are poor: one does not normally appreciate rugs and
other textile carpetings (dirt, maintenance, durability) and plastic coverages are frequently considered
vulgar. Additionally, in case of a replacement of these ruined coverages one must not forget to select a
new product having the same acoustical characteristics as the old one.
Rugs and textile carpetings frequently have a LW possibly reaching from 17 dB to more than 30 dB.
Plastic covers with an integrated under-layer have LW from 5 dB to 19 dB.
A4-5.4.4.2

Hard coverings directly placed on top of a soft under-layer

These are most frequently tiles or parquets. The acoustic quality of these products on their under-layer is
not the only consideration to take into account. Actually, their mechanical characteristics must also be
satisfactory. This leads to a compromise for the rigidity of the under-layer which should be as soft as possible for acoustical reasons and relatively rigid to avoid the problems of concentrated loads.
Tiles glued on a thin under-layer (thickness below or equal 3 mm) can have a LW between 10 and 18 dB.
Parquets placed on thin under-layers can reach the same values.

45

In every case, it is prudent to claim from the manufacturer of the under-layer the complete record made of
the laboratory measurements.
On the other hand, it is indispensable to pay attention to the disconnection of the solid cover from the
vertical walls limiting the treated room.
A4-5.4.4.3

Floating floors

Floating floors on thin compressible mats of a thickness below or equal 3 mm: their efficiency can reach
from 17 to 21 dB. Taking the poor compressibility of the under-layer into account, it is not necessary to
reinforce the screed. The thin compressible mats can directly be placed under the mortar accommodating
the bedded floor tiles.
Floating floors on thick compressible mats based on mineral wool or elastic polystyrene: their efficiency
can go beyond 27 dB. The under-layers used have a thickness between 1 and 4 cm. The covering layers
must be reinforced.
Whether the under-layer be thin or thick, the floating floor needs to float in every direction, that is to say
both vertically and horizontally. This requires important precautions when installing the under-layer on the
supporting floor and peripheral elastic strips along the walls and any other infringements (canalisation,
door thresholds).
On the ground, the different parts of the under-layer must be fixed to one another, for example by selfadhesive tape. This is intended to avoid a movement of one opposite the other when applying the outer
layer.
In case canalisation is placed in the supporting floor, it is desirable to case it in a cladding or in stabilised
sand before putting in place the under-layer.
The peripheral compressible bands, along the vertical walls, must be sufficiently high to reach above the
level of the final layer of the floor coverage. The tightness of the under-layer or peripherical bands , opposite the mortar wetness must be assured. A foil of polyethylene, placed over the under-layer, allows for
the assurance of this tightness.
The peripheral bands must be cut only after the putting into position of the last elements of the floor coverage. Again, it must be taken care that these baseboards do not create a rigid connection between the
final layer of the floor coverage and the vertical walls. To avoid this one must arrange a gap which will be
filled by a flexible mastic in order to realise tightness (see the sketch of the principle in Figure 30).

Figure 30: Principle of floating floor

46

a
b
c
g
h

supporting floor
canalisation
filling material up to the top level of the canalisation
Adhesive element for tightness
floating layer

d
e
f
i
j
k

mats of under-layer with a water-tight cover

self-adhesive tape for the connection between mats
Peripherical elastic band
top layer
baseboard
soft mastic

The under-layers having been well protected, the base boards well decoupled, it must be taken care that
the furniture itself does not recreate the connection. Consequently, the kitchen furniture or the bathroom
furniture sitting on floating floors must be disconnected from the vertical walls.
Note: A floating floor covered with a rigid material or with a rigid coverage on an under-layer is more
noisy in the room where the impact noise is produced as if the rigid coverage would be glued directly on the supporting floor. The under-layer captures part of the energy which is radiated back
into the room of emission. This allows to develop a simple test not requiring any acoustical measurements to verify that the floating floor does not contain any fixed points endangering its efficiency. It suffices to "sound" the coverage with a hammer. In case the sound is clear and intensive,
the cover floats well. In case the sound is dampened, their must be a point of rigid connection in
the vicinity of the impact point.
The efficiency of floating floors for airborne sound
The floating floor is composed of a casing of cement or concrete of 4 or 5 cm thickness which corre2
sponds to a surface mass added to the supporting floor of 80 to 120 kg/m . However, experience shows
that one cannot always count on this weight addition to improve the sound reduction index of the supporting floor. This depends essentially on the stiffness of the under-layer.
As a first approximation, one can utilise the following rule:
Since the floating system has a LW below 20 dB, one considers that the floating floor does not improve
the airborne sound insulation. Anyway, this is a relatively pessimistic assessment and in case one needs
to improve the sound reduction index of the supporting floor by 1 dB, one could count on it that this decibel is added by the floating floor.
Where the floating floor has a LW between 20 and 23 dB, one can assume that the floor has an acoustic
reduction index in the vicinity of the index corresponding to the law of mass, supporting floor plus floating
floor.
Where the floating floor has a LW above 23 dB, one assumes the total law of mass to which one adds 1
or 2 dB. For supporting floors of a thickness above 18 cm of concrete, one adds 1 dB, for floors of lesser
thicknesses one adds 2 dB.
Finally, floating floors efficient against airborne sound allow for the attenuation of lateral transmissions
between two adjacent rooms as would thermo-acoustical linings of the lateral walls.
A4-5.4.5

Example for the application of the prediction method

16 cm of concrete + coverage (- 2 dB)

3m

5m

4m
Concrete 18 cm

Room

Living room

Lightweight partitions

Figure 31: Sketch for the example considered

47

Figure 31 shows a normal configuration which allows to obtain the regulated standardised acoustic insulation DnTA of 53 dB between storeys and between a bedroom in one flat and a living-room in a neighbouring flat. The supporting wall and the floors are of 18 cm concrete. The facade is of 16 cm concrete with a
thermal coverage based on expanded polystyrene and the interior walls in the flats are lightweight partitions.
Already now, one can calculate the sound pressure level of normalised impact sound between flats above
3
3
each other. The level is 2.2 dB higher in the room of 30 m when compared to the room of 50 m (10 log
(50/30) = 2.2).

Estimation of the sound pressure level L'nTw in a room when the floor is without any coverings:
(42)
L'nTw = 181 35 log (m0) - LnW 10 log (V) + K
2
Surface mass of floor: m0 = 423 kg/m (see the table on page 34)
LnW of floor coverings = 0
3
Volume of the room = 30 m
Increase of K of the level as a result of lateral transmissions (see Table 12):
Average surface mass of homogenous lateral elements: these elements are of concrete (facade and
partition wall) and have the same height (2.5 m).
2
Then: m = 2350 ((3 0.16 + 4 0.18 ) / 7) = 402 kg/m
and K = 1 dB which leads to an L'nTw = 75 dB. An even more precise evaluation using a software
which allows for the calculation of all transmission pathes gives a forecast in the vicinity of 74 dB.
For the standardised impact sound pressure level in the lower living-room, when the impact sound
generator works in the living room above, a simplified method allows to estimate the value at 73 dB.
Determination of LW for the floor coverage must be planned so that the weighted sound pressure
level of standardised impact sound does not exceed 58 dB (maximum value acceptable in France)
LW > 75 58 = 17 dB in the room and 73 58 = 15 dB in the living-room. One could envisage a carpet in the room and a parquet or a tiled floor on a flexible layer in the living-room.

A4-5.5

Improvement of the impact sound insulation

Where the regulated or contracted requirements are not satisfactory, a better solution to improve the impact sound insulation is obviously to replace the coverage which has been applied on the floor by a more
efficient product. However, there are cases when this is not possible.
If somebody complains that he hears to much of the impact noise from the room above, it will be rare that
the tenant of that room accepts to improve his floor coverage. In this case, one can envisage to have to
work with the complaining party. For this, one uses acoustical coverages (linings) identical to those used
for the improvement of airborne sound insulation.
The influence of linings, reinforcing the sound reduction index on the underside of ceilings, on the insulation of impact sound
In case these linings have an efficiency R for airborne sound, they have an efficiency Ld of the same
value on the direct transmission of impact sound. On the other hand, the lateral transmissions are not
changed and the Ld is not reflected in the overall result of the sound pressure level.
Example:

48

Emission room

Reception room

Figure 32: A complex coverage is realised at the ceiling of the reception room of impact noise
th

In this flat from the 60 , the walls were of 14 cm concrete. The tenant of the emission area had replaced
the original floor coverage by a parquet glued on a flexible layer. The tenant of the reception area complained about the walking noise of his neighbour. The measured weighted impact sound pressure level
was 68 dB. One decided to find a solution by treating the reception area to lower the level of impact
sound by 10 dB and one placed a lining complex on the ceiling of the reception area which allowed for an
increase of the acoustic reduction index R of 10 dB (10 cm mineral wool and a gypsum plate on frames).
The final result is deceiving.
Indeed, the situation before the reinforcement was the following: The measured impact sound level is the
combination of the direct transmission and the lateral transmissions. If one considers an increase of impact sound by 1 dB as a result of lateral transmissions (K = 1 dB), the direct transmission corresponds to
67 dB and the lateral transmissions to 61 dB (the addition of 67 dB and 61 dB gives 68 dB).
After the reinforcement of the ceiling, only the direct transmission had been lowered by 10 dB, the lateral
transmissions remained the same as before. Now, one needs to combine 67 10 = 57 dB for the new
direct transmission with the 61 dB of the lateral transmissions which gives a resulting level of 61.5 dB,
and an overall improvement of 6.5 dB instead of the 10 dB aimed for.
So it is required to complete the treatment of the ceiling in the reception area by a treatment of the vertical
walls. Starting with the lightest, if they are rigid. In certain cases, all the vertical walls need to be treated,
which indeed decreases the living space in the reception area substantially and is very costly.
A4-5.6

Horizontal or diagonal transmissions of impact noise

Figure 32 shows that in case the walls are identical in each and every level, a diagonal transmission is
the same as the horizontal transmission. On the other hand, the lateral transmissions are higher where
the distributing walls are light. The table below allows to assess the difference between the direct transmissions and the horizontal or diagonal transmissions as a function of the walls limiting the reception
room. This table gives only an estimation. To obtain more precise values, one uses the equations based
on the dispositions of the standard EN 12354 "Estimation of acoustic performance of buildings from the
performance of elements Part 2: Impact sound insulation between rooms".

49

Emission room

Figure 33: Vertical, diagonal and horizontal transmissions of impact noise

The arrows show the portions of impact energy on the different walls in the rooms neighbouring the emission room.
Type of wall
Light wall
Heavy wall

L'nTw diagonal L'nTw vertical

- 3 dB
- 6 dB

Table 14:
A4-6

Noise levels of building equipment

The building equipment is manifold and the corresponding sources of noise are very different. For all
sorts of equipment, the only common value is the noise level acceptable at the area to be protected.
Where this applies, one examines the points which constitute sources of noise in the different installations. These sources produce either airborne sound (acoustic energy communicated to the ambient air) or
impact sound (acoustic energy directly communicated to the walls of the building).
The general principles used are described in the future standards prEN 12354-5 "Acoustics in buildings
Estimation of acoustic performance of buildings from the performance of elements Part 5: Sound levels
caused by building equipment".
A4-6.1

Maximum sound pressure level at the reception areas

The sound pressure level can be given at octave intervals, between the interval centred on 63 Hz and the
interval centred on 4000 Hz. They are either normalised (the level which one obtains when the equivalent
2
absorption area of the place of reception is 10 m ) or standardised (the level which one obtains if the reverberation time in the reception area is 0,5 seconds).
On the other hand, the objective values, given in the form of acoustic pressure levels, which should not
be exceeded in the area of reception are frequently given in dB(A) or in dB(C). These values are calculated starting from the spectrum of octave intervals.
The values most frequently used are the following:
LA, max, Ln, A, max, LnT, A, max: maximum sound pressure level, expressed in dB(A), raw normalised or standardised.
LA, eq, Ln, A eq or LnT, A eq: equivalent continuous levels, raw normalised or standardised, expressed in dB(A).
LC, max, Ln, C max, LnT, C max: maximum sound pressure level, expressed in dB(C), raw normalised or standardised.
LC, eq, Ln, C eq or LnT, C eq: equivalent continuous levels, raw normalised or standardised, expressed in
dB(C).

50

The relations between the different levels used are the following:
Ln = L + 10 log (A/A0)
LnT = L 10 log (T/T0)
LnT = Ln 10 log (0,032 V)
L is the level of the unweighted sound pressure level in dB.
Ln is the level of the normalised sound pressure level in dB.
LnT is the level of the standardised sound pressure level in dB.
2
A is the equivalent absorption area in the reception room (m ), A0 is the reference equivalent absorption
2
area, generally taken as 10 m .
T is the reverberation time in the reception room (s), T0 is the reference reverberation time, generally
taken as 0,5 s, and occasionally similar to the reverberation time laid down in contracts for the type of
room under consideration.
3
V is the volume of the reception room (m ).
A4-6.2

The different types of energy transmission to be considered

The airborne noise transmitted by the walls of a building: the sources of airborne noise of installations
produce a sound pressure level in the room where they are installed. Part of the corresponding energy is
transmitted towards the reception area by the walls of the building.
The airborne noise transmitted by or across the ducts and canalisation pipes and transported to the reception area.
The structure-borne noise, corresponding to a vibration energy communicated directly to the walls of the
building, which conduct part of it to the reception area.
If one considers the normalised sound pressure level in the reception area, the level Ln is obtained by the
equation below:
n
0
m
/ 10
/ 10
L
/ 10
L
L
Ln = 10 log10 10 n,a,i + 10 n,d, j +
10 n,s,k
i =1

j =1
k =1

(44)

where Ln,a,i is the normalised sound pressure level caused by the transmission in the structures of the
building by the source i.
Ln,d,j is the normalised sound pressure level conducted by or across ducts, canalisations, pipes, for the
source j.
Ln,s,k is the normalised sound pressure level caused by the transmission of structure-borne sound produced by the source k.
A4-6.2.1

Transmission of airborne sound by the structure of the building

This transmission is similar to that studied in the general problem of acoustic insulation between two
places, with the direct transmission through the separating wall and the flanking transmissions by the
lateral walls (see chapter A4-2). Let us always remember that one cannot argue using the single values of
insulation (DW , Dn, A or DnT, A) since the spectrum of emitted noise is rarely pink noise or traffic noise. In
fact, one should not exceed a sound pressure level expressed in dB(A) in the reception area, taking account of the actual noise spectrum emitted by the source in the emission area. One is therefore required
to make a calculation by frequency intervals to determine, finally, the sound pressure level in dB(A) in the
reception area.
The only frequency intervals that could be used are octave intervals, since the source is practically always characterised by the acoustic power levels given by octave intervals.
The characteristics of the walls, measured in laboratories, are normally known by third-octave intervals
between those centred on 100 Hz and those centred on 3150 Hz. This allows to calculate and determine
the characteristics of the octaves from the octave interval centred on 125 Hz up to the one centred on
2000 Hz. However, numerous types of equipment generate a noise with important sound levels in the
octave centred on 63 Hz, and occasionally relatively high in the octave centred on 4000 Hz. These two

51

frequency intervals are frequently not present in the given characteristics of walls. One has therefore to
make hypothesis, generally by extrapolation.
A4-6.2.2

Transmission of airborne sound by or across the ducts

There are three different types of transmission.

Transmission of the initial source to the farthest end of the network in the reception area: If one considers a network of ducts or of canalisation ducts, a source of noise (ventilator, pump, ...) delivers a
certain acoustic power at the entrance of that network. This acoustic power is partly consumed by the
peculiarities of the network (elbows, long stretches, derivations, cross-section changes, diaphragms,
...). From this results a sound pressure level obtained at the end of the network, in the area of reception. This level of acoustic pressure allows to calculate the sound pressure level at the location, knowing the distance between the end of the network and the reception point, knowing also the absorption
characteristics of that place (see document A5).
Transmission of the noise regenerated by the different elements of the network: The elements of the
network normally attenuate the acoustic pressure taken along by the fluids transported in the network,
but they regenerate equally a noise more or less important, dependent upon the flowing speed of the
fluids. This is how a silencer placed in a ventilation circuit is an attenuator of the noise, but in the
same way a generator of noise. The aim is that the regenerated noise is below the noise which is
supposed to be silenced. Most frequently, one calculates the acoustic pressure at the exit of the element caused by attenuation and one determines just behind the regenerated acoustic energy, with
the aim to obtain the overall acoustic pressure behind that element.
The transmission of noise existing in a room into the interior of a network, by the walls of ducts or
canalisation ducts, and transmission by the walls of the ducts or the canalisation ducts of the noise
transported to the room which is transversed.

A4-6.2.3

Transmission of structure-borne sound

This is the least well controlled part in the problem area of equipment noise.
The sources of equipment noise do not only produce airborne sound, but also structure-borne sound,
which is transported by vibrations which proceed through the walls towards the reception area. One can
frequently determine the frequencies in which the vibration energy is the most important, but one has few
means for the evaluation of the power transmitted through the building structure. To treat this problem,
one takes recourse to anti-vibration materials placed under the machines, on the suspensions of canalisation ducts, on the decoupling of wall beams. In short, one takes precautions, judiciously selected as a
function of frequencies to be attenuated, without being able to precisely calculate the effect of these precautions.
The provisional European standard quoted before does not give any useful indication on how to master
these problems of structure-borne sound.
A4-6.3

The main equipment parts of a building and the types of transmission which they produce

A4-6.3.1

Ventilation systems

The sources, the origin of noise in a ventilation system, are frequently inside a casing (central air duct)
placed in a equipment room or below the roof of a building. In these cases, one finds a ventilator, an electrical motor, a mechanical transmission line, a compressor, several filters and so on.
Regarding the airborne noises produced by the casing, they are characterised by three sound power levels, given at octave intervals (between 63 Hz and 4000 Hz or 8000 Hz).

The sound power of admission: In the case of a blowing network, it is the acoustic energy which is
used for a calculation of the network to take the air in and to distribute it, or of the network used to introduce fresh air.

52

The sound power of the output: This is the acoustic energy considered at the entrance of the blowing
network. The calculation of the network consists principally of the determination how that acoustic
energy is partially consumed by the characteristics of the network, to arrive at the level of acoustic
energy to be distributed in the reception area. This calculation must take into account the regeneration of noise by the elements of the network, as we have indicated earlier.
The sound power radiated by the housing of the material: This energy serves mainly to determine the
sound pressure level in the reception area which is caused by the transmission by the walls of the
building.

Regarding the structure-borne sound, one should at least know the rotation speed of propellers in the
casings (motors, ventilators), the weight of the casings and the distribution of that weight on the respective foundations. This given would allow for a determination of the anti-vibration characteristics needed to
transmit the lowest possible amount of vibration to the walls supporting the casings.
Regarding the ventilation network, the elements to be considered are numerous:

Ducts: Attenuation and regeneration by the straight runs (the regeneration is the stronger, the higher
the speed of the air).
Elbows: Attenuation and regeneration a right-angle elbow attenuates and regenerates more than a
round bend.
Derivations: Frequently associated with elbows.
Noise attenuators: A silencer attenuates and regenerates.
Cross-section changes: It is required that the cross-section ratio should be high to be able to count on
a significant attenuation.
Fire-stopping valves: Normally not having great influence since they are open, which is their normal
configuration.
Valve regulators: Attention must be given to the regeneration.
High-pressure equipment: Problem of regeneration.
Terminals: At the end of the network, the grids and diffusers occasionally produce a regeneration
which is not negligible and have an indigenous attenuation which is added to the effect of the reflection across the air outlet in the room.

We are not going into more details on the chapter relating to ventilation installations since it is too a big
part treated in chapter A6-2.3 of document A6 "Sound propagation in ducts, pipes and ventilation systems".
A4-6.3.2

Heating installations

Heating can be realised by different types of terminals:

Radiators or convectors of hot water.

Electrical radiators or convectors (to remember, since they cause smaller acoustical problems when
they are not completed by air-blowing ventilators).
Hot-air openings.
Under-floor heating systems.
Heated ceilings.

Amongst the electrical systems, the heat is produced by a heater which is the principal source of noise
and of vibrations in the installation. The airborne noise is produced by the fire and by the ventilator supplying the air. Let us remark that equally an important resonance noise can be produced linked to the
burning chamber, to the connection to the smoke duct and to the smoke duct (chimney) itself.
In case of a collective heating system for a building, the heating installations are in a technical installation
area called "heatery". In this place, one finds other sources of airborne noise, but also always vibrations,
such as from the pumps (normally on a foundation on the ground) or the circulation pumps (frequently
directly placed on the pipes).

53

For a pressurised burner for the heating of a power P in kW, the sound power level could be estimated by
the relation: LWA = 57 + 12 log (P) in dB(A) (see prEN 12354-5). A well designed casing of the burner can
diminish this level of energy to 20 dB. Attention! The sound levels in the low frequencies are especially
important.
A French study of measurements of 145 heating systems with burners with blowing air has shown that
90% of the heating systems produce sound levels below the following values:
Lp 55 + 10 log (P) dB(A),
where P is the thermal power of the heating system in kW.
In the low frequencies, the logarithmic mean of the octaves centred on 63 Hz, 125 Hz and 250 Hz:
Lp 65 + 8 log (P) in dB.
In the mean frequencies, the logarithmic mean of the octaves centred on 500 Hz and 1000 Hz:
Lp 51 + 10 log (P) in dB.
In the high frequencies, the logarithmic mean of the octaves centred on 2000 Hz and 4000 Hz:
Lp 58 + 12 log (P) in dB.
At similar thermal power, the noise levels are very different dependent on the generators used.
For example, if one compares the noise of a heating system of a power of 1000 kW, with burners of blowing air or with burners using atmospheric air, the levels of sound pressure are much lower for the second
type of burners.
The table below shows a comparison of the sound pressure levels found in a boiler room with a burner
using atmospheric air, without a ventilator of extraction, and in a burner room with a burner using blowing
air, for a thermal power of 1000 kW.
Frequencies
Type of burner
Burner with atmospheric air
Burner with blowing air

Frequencies
Low
(63, 125 and 250 Hz)

Medium
(500 and 1000 Hz)

High
(2000 and 4000 Hz)

58
89

48
81

42
74

LW in dB(A)
52
85

Table 15: Sound pressure levels in two similar boiler rooms with different types of burners
In a boiler room with an atmospheric air burner, the combustion is not very noisy, and the noise of the
pumps can be revealed. Thus, in the boiler room used in the example above, the level of sound pressure
passes from 52 dB(A) to 65 dB(A), since the pumps have been brought into action.
In case of individual heating systems, the burners are frequently using atmospheric air and are walled
(situated in a wall of a flat room, frequently the kitchen). For the heating systems of type B, for which the
removal of the burned gases takes place in a duct and the combustion air is taken directly in the room,
the sound power is frequently found between 45 dB(A) and 50 dB(A), since the heater works fully
charged.
For the walled heating systems of type C, the combustion circuit has no communication with the place
where they are installed. Their sound power frequently varies between 44 dB(A) and 48 dB(A).
A4-6.3.3

The installations for water supply

Sources of noise are principally the taps, the pumps and the water gates. But one must see also other
sources of noise, such as the fall of the water into a bathtub when it is filled.

54

The European standard prEN 12354-5 quoted earlier gives a table listing the possible sources of noise in
a water-supply installation with the types of noises generated (see Table 16).
The acoustic performance of a tap is characterised by the sound pressure level Lap defined by the relation
Lap = Lsr + L Ls
where Ls is the sound pressure level obtained in the room of measurement when s standard tap (generator of hydraulic noise) functions with a supply pressure of 3 bar.
Lsr is the reference sound pressure level when a standardised generator works.
L is the sound pressure level measured when the tap to be tested works.
The level Lap allows to classify the taps according to their respective more or less silent characteristic.
The weaker the Lap, the less noisy the tap.
Regarding the standardised generator of hydraulic noise, it is designed to make much noise (nobody
would like it in his flat).
The noise of a tap increases as the supply pressure of the water increases.
The noise of a tap increases as the amount of water dispensed increases.
Table 16 gives the noise sources of a water-supply installation together with the type of noise generated.
Generation of
structure-borne
sound
X
X
X

fluid-borne
sound
X
X
X

gate valves, inline fittings:

stop taps

corner stop valves

back flow prevention devices

(X)

(X)

throttle valves
pressure reducing valves

X
X

X
X

X
(X)

X
(X)

X
X

regulating elements at the outlet of

taps:

jet regulators

flow control devices

anti-vacuum valves

back flow prevention devices

X
(X)

X
X
X

X
X
X

Component
valves and taps of any kind:
draw-off taps
hose taps
flush taps

pipes
flushing boxes
(circulating) water heaters, boilers
basins:
bathtubs
shower tubs
lavatory basins
sinks
toilet-basins
booster pumps

airborne
sound

Transmission to the building by

direct
mounting
pipes
mounting
elements
X
X
X
X
X
X
X
X
X

X
X
X

X
X
X

X
X
(X)

Table 16: Compilation of sources and relevant types of transmission in water-supply systems
A4-6.4

Some principal precautions to take to limit the noise of certain installations

This chapter treats cases frequently encountered in practice and gives some advice regarding the first
precautionary measures to take. Taking into account the number of different equipment and the variety of
materials in each category, this is only an overview of the acoustical problems, however, which allows to
explain a certain number of useful principles.

55

A4-6.4.1

Equipment installed at the outside

Examples: air conditioners, cooling towers, ventilators, air extractors, cooling assemblies ...
These machines pose two types of acoustical problems: the first is linked to the protection of the environment, the second is related to the protection of the persons living in the equipped building.
For the protection of the environment, one has to:

Put the largest possible distance between the equipment and possible complainants, even if this
leads to very long channels.
Avoid placing the equipment in front of a wall which might reflect the acoustic energy towards a zone
to be protected.
Foresee the possibility to place screens between the machine and the possible complainants, these
screens must allow for the assured maintenance of the machine, its air intake ... and must also respect the urban regulations regarding the height of buildings.
Direct the least noisy face towards the zone to be protected, in case the equipment has a marked
directivity in a certain direction.

For the protection of the occupants of the equipped building, one has to foresee anti-vibration measurements under the equipment without forgetting the decoupling from the pipe work serving them.
In all these cases, the criterion of acoustic emission must be considered when choosing the equipment.
For the same working characteristics, there are frequently some machines less noisy than others.
Finally, it is frequently possible to limit the working periods of the equipment, especially by stopping them
or have them work at a reduced rate during the night.
A4-6.4.2

Heating systems in occupied places

Choosing preferably atmospheric heaters.

Not placing the pipes under separating floors.
Put the oven on anti-vibration foundation and foresee cuffs between the serving pipes and the oven.
Pumps with anti-vibration and soft cuffs.
Chimneys independent of the building.
A4-6.4.3

Basement heating stations

In basement heating stations, the noise sources are the pumps, the pressure reducers for the water or
the steam (most noisy), the heat exchangers, the regulation valves, the movement of the water or the
steam in the pipes.
To limit the irritation of occupants in adjacent rooms, it is prudent to choose heavy walls or double walls to
isolate the pipes and to limit the speed of the steam in the pipe work.
The table below gives the speed limits of steam as a function of the pipe diameter.
Diameter in mm
Speed in m/s

40
22

50
28

65
34

80
42

100
49

125
57

150
65

175
74

200
80

These limitations apply to the straight portions of the network. In case of irregularities, such as elbows,
changes in diameter, it is necessary to use lower speeds.
A4-6.4.4

Electrical transformation installations

The table below has a collection of precautionary measured to take.

56

Reduction of the transmission of noise and of vibrations

Project: indispensable precautions 1)
Actual situation: noise reduction
Walls and floors
Choice of a more silent transformer 4)
Surface mass above 200 kg/m2, composed
of homogenous, not porous material with
15 mm spaces between the elements which
are covered by an apron
Ventilation openings
Openings
Avoid to have these outlet at distances
- Acoustically treated grid
below 30 m from buildings or in a passage
Airborne
- Acoustic obstacles
onto which residential buildings give
Transformer
Doors
Doors
Foresee frames that are able to accommoWhere a noise reduction is necessary, the
date doors with an enhanced acoustic
door might pose a problem, its acoustic
insulation
insulation might need to be reinforced.
Prefabricated items
Screens (prefabricated items)
Choice of a less noisy outlet 3).
Such a height that the screen prevents direct
transmission to the neighbourhood 5).
Avoid the placement below a room. Install
Anti-vibration foundations
Structureanti-vibration foundations under the legs 2). Same
borne
Decouple at the connection
StructureDisconnect from the ground and walls with
Same
Table BT
borne
the help of anti-vibration foundations 2).
Disconnect from walls and the ceiling with
Same
the help of anti-vibration foundations and
StructureCables MT and BT
borne
connections of BT to the transformer by soft
connections (ropes, metal bands ...).
1) These precautions allow to minimise the risks and to avoid later modifications which are difficult to achieve, if not impossible and
always costly.
2) Measures to take in cases where the installation presents a possible trouble potential.
3) This is not a special make, but a choice between accepted materials ...
4) This concerns the choice of a less noisy machine, of special make, allowing to obtain a reduction of at least 15 dB(A).
5) A solution difficult to execute, not to attempt unless the minimum distance could not be respected and unless the solutions 3
and 4 would not be sufficient.
Source of noise
and/or of vibration

Propagation

Table 17:
Reduction of noise transmissions and vibrations caused by transformers (source
EDF)
BT = standard tension
MT = medium tension
A4-7
Symbol
A
A0
Atot
C; Ctr
Cd
D
Dc
Dd
DI
Dlat
Dn
Dn,i,j
DnA
DnAtr
Dne
Dnew
DnT
DnTA
DnTAe,tr
DnTAtr
DnTe
DnTew

Symbols, quantities, units

Quantities
equivalent absorption area in a room, of a surface or of an object
2
reference equivalent absorption area (generally 10 m )
attenuation on the progress of a wave
adaptive terms for the measured acoustic attenuation indices or the measured
acoustic insulation (C for a pink noise emission and Ctr for a traffic noise emission)
correction for diffusion
acoustic insulation
directivity correction for a point source
direct acoustic insulation (transmission through the separating wall)
index of directivity (10 log(Q))
acoustic insulation on the lateral transmission pathes
normalised acoustic insulation (reference A0)
normalised insulation on the transmission by separation i versus the separation j
normalised acoustic insulation = Dnw + C
normalised acoustic insulation = Dnw + Ctr
normalised acoustic insulation of a small element
measured normalised acoustic insulation of a small element
standardised acoustic insulation (reference T0)
standardised acoustic insulation = DnTw + C
DnTew + Ctr
standardised acoustic insulation = DnTw + Ctr
standardised acoustic insulation of a small element
measured standardised acoustic insulation of a small element

Unit
2
m
2
m
dB
dB

dB
dB
dB
dB
dB
dB
dB
dB
dB
dB
dB
dB
dB
dB
dB
dB
dB
dB

57

Symbol
DnTw
Dnw
Dv,i,j
D
K0
Ki,j
L, Lp L1 L2
L'nTw
L'nw
LA,max
LAeq
Lap
Leq
Ln
LnA,max
LnAeq
LnT
LnTA,max
LnTAeq
LnTw
Lnw
Ls
Lsr
LW
m0
Q
R
R'
RA, RA,tr
Rg
Rlim
RW
S
S0
Sg
Sr
T
T0
Tl
V
V0
W
W1
W2
W3
Lnw
R

A4-8
ISO 717

Quantities
measured standardised acoustic insulation
measured normalised acoustic insulation
vibration suppression in a connection between two walls
index for a solid angle
constant for a position of a certain value in a room
index of vibration attenuation in a connection between two walls
naked sound pressure level (L1 in the emission room, L2 in the reception room)
measured standardised sound pressure level in a place
measured normalised sound pressure level in a place
maximum sound pressure level in dB(A)
continuous equivalent level in dB(A)
sound pressure level of a tap
equivalent continuous level
normalised sound pressure level
normalised maximum sound pressure level in dB(A)
continuous normalised equivalent level in dB(A)
standardised sound pressure level
standardised maximum sound pressure level in dB(A)
continuous standardised equivalent level in dB(A)
measured standardised sound pressure level
normalised measured sound pressure level
sound pressure level of a standard tap
reference sound pressure level
sound power level
surface mass of a separating wall
directivity of a sound source
acoustic attenuation index of a wall
apparent acoustic attenuation index of a wall, measured in place
RA = RW + C and RA,tr = RW + Ctr
acoustic attenuation index of a wall in a duct
limit index of acoustic attenuation of a wall composed of two elements
measured acoustic attenuation index (unique value)
surface, normally of a wall
2
reference surface (10 m )
surface of a duct in a room passed through
surface of a stiff and light wall linked to a separating wall between two rooms
reverberation time
reference reverberation time (normally 0.5 s)
insulation loss caused by lateral transmissions between two rooms
volume of a room
3
reference volume (normally 30 m )
sound power
sound power hitting a separating wall
sound power transmitted through a separating wall
sound power transmitted by other elements than the separating wall
improvement caused by a floor coverage
variation of indices of sound attenuation of a wall caused by a complex coverage
transmission factor of a wall
Standards and bibliography
Acoustics Rating of sound insulation in buildings and of building elements
Part 1: Airborne sound insulation
Part 2: Impact sound insulation

Unit
dB
dB
dB
dB
dB
dB
dB
dB
dB
dB(A)
dB(A)
dB
dB
dB
dB(A)
dB(A)
dB
dB(A)
dB(A)
dB
dB
dB
dB
dB
dB
dB
dB
dB
dB
dB
dB
2
m
2
m
2
m
2
m
s
s
dV
3
m
3
m
W
W
W
W
dB
dB

58

EN 12354

EN ISO 140

EN ISO 16032
EN ISO 3740
EN ISO 3741
EN ISO 3743

EN ISO 3744

EN ISO 3746

EN ISO 3747

DIN 4109

Building acoustics - Estimation of acoustic performance of buildings from the performance of elements
Part 1: Airborne sound insulation between rooms
Part 2: Impact sound insulation between rooms
Part 3: Airborne sound insulation against outdoor sound
Part 4: Transmission of indoor sound to the outside
Part 5: Installation Noise
Part 6: Sound absorption in enclosed spaces
Acoustics Measurement of sound insulation in buildings and of building elements
Part 1: Requirements for laboratory test facilities with suppressed flanking
transmission
Part 3: Laboratory measurements of airborne sound insulation of
building elements
Part 4: Field measurement of airborne sound insulation between rooms
Part 5: Field measurement of airborne sound insulation of facade elements
and facades
Part 6: Laboratory measurement of impact sound insulation of floors
Part 7: Field measurements of impact sound insulation of floors
Part 8: Laboratory measurements of the reduction of transmitted noise
by floor coverings on a heavyweight standard floor
Part 11: Measurement of impact sound reduction through ceiling covers
on light reference ceilings in testfacilities
Part 14: Directives for special building conditions
Part 16: Measurement of the improvement of sound-reduction indexes
through additional insulation-linings
Part 18: Measurement of noise caused by rain on building components
in testfacilities
Acoustics Measurement of sound pressure level from service equipement
in buildings - Engineering method
Acoustics - Determination of sound power levels of noise sources
- Guidelines for the use of basic standards
Acoustics - Determination of sound power levels of noise sources using
sound pressure Precision methods for reverberation rooms
Acoustics - Determination of sound power levels of noise sources
- Engineering methods for small, movable sources in reverberant fields
Part 1: Comparison method for hard-walled test rooms
Part 2: Methods for special reverberation test rooms
Acoustics - Determination of sound power levels of noise sources using
sound pressure
- Engineering method in an essentially free field over a reflecting plane
Acoustics - Determination of sound power levels of noise sources using
sound pressure
- Survey method using an enveloping measurement surface over a
reflecting plane
Acoustics - Determination of sound power levels of noise sources using
sound pressure
- Comparison method for use in situ
Sound insulation in buildings

La pratique de l'acoustique dans le btiment

Mathias Meisser

L'acoustique du btiment par L'exemple

Mathias Meisser

L'acoustique des locaux scolaires

Mathias Meisser

L'isolation acoustique et le gros uvre

Mathias Meisser

59

Encyclopdie des mtiers maonnerie et pierre de taille: l'isolation

acoustique

Mathias Meisser

Guide AICVF Bruits des quipments

Les sciences du btiment acoustique

REFF volume II C.S.T.B.

DISCLAIMER
The information contained in this document is considered by us to be good practice and
industry guidance (Guidance). The Guidance is provided by a commission of FESI
(www.fesi.eu), and is considered to be correct at the date of publication. Whilst we are
confident the information contained within it is up to date and accurate, it is a reference
document only. It is your responsibility to ensure your knowledge of the matters discussed in
the Guidance is kept up to date.

The Guidance is intended to be used for general purposes only and is not intended to take
precedence over appropriate national and international standards, guidelines or laws (where
applicable). The Guidance is not intended to replace detailed calculations and assessments
of prevailing physical conditions in complicated building assignments.

The Guidance does not constitute professional advice and specific queries should be
referred to qualified professionals. Any reliance placed on the Guidance without seeking
such advice is strictly at your own risk. We make no representations or warranties of any
kind, express or implied, about the completeness, accuracy, reliability or suitability of the
Guidance.

The Guidance in its original form is written in English. We accept no responsibility for any
inaccuracies contained in any translation of the Guidance in languages other than English.

The Guidance is provided free of charge and as such in no event will we be liable for any
loss or damage including, without limitation, indirect or consequential loss or damage, or any
loss or damage whatsoever arising from any reliance placed on the Guidance. None of these
exclusions or limitations, however, are intended to limit any rights which may not be
excluded, nor to exclude or limit liability to you for death or personal injury resulting from our
negligence or for other liability which we cannot exclude or limit by law.