Sound Level Test :

The purpose of a sound measurement whether pressure or intensity is to be enabled to estimate the
sound power emitted by the transformer. The information about noise level helps determine
appropriate noise control measures that need to be put in place. The noise pressure generated by
vibration of core and windings is transmitted to tank surfaces though the oil medium. Since the oil is
relatively incompressible, the noise is transmitted without appreciable damping. The tank responds to
these noise waves depending on its natural frequencies and complete shapes of vibrations. So to reduce
the vibration of tank sound level information is necessary.

Sound Measurement Methods:

The audible sound is generated by magneto-strictive deformation of the core and electromagnetic
forces in the windings, tank wall and magnetic shield. Sound generated by the magnetic field inducing
longitudinal vibration in the core laminations has been dominant. The amplitude of these vibration
depends on the flux density in the laminations and the magnetic properties of the core steel hence is
independent of the load current. Aside from flux density stray magnetic fields may induce vibration in
structural components. Force (and so vibration amplitude) is proportional to current squared, and
emitted sound power is proportional to vibration amplitude squared. This means that the load current
has a significant impact on the total sound power emitted.

General Instructions According to IEC 60076-10 & IEC 60076-10-1:

The sound power is the sound intensity integrated over the measurement surface that encompasses the
test object. Therefore, the measurement of sound intensity directly results in the assessment of sound
power. A second method for estimating sound power is to measure sound pressure using more
straightforward procedures, assuming that the sound pressure and particle velocity are in phase and
proportionate. Measuring sound pressure provides limited information and this method requires
corrections where the assumption is not fulfilled.

When sound pressure and particle velocity are in phase(free field condition) a unique relationship exists
between sound pressure and sound intensity, I according to the following equation,
2
|I |= P --------------------------- (1)
ρc
Where ρc is the acoustic impedance of the medium through which the sound is propagating. Under
standard conditions of pressure and temperature in air, ρc = 412kg/m2 s

Using Equation(1) in the normal sound intensity( LI) level equation, the relationship between normal
sound intensity level, LI, and sound pressure level, Lp, is shown in Equation(2) as follows

|I|
( ) ( ) ( )
2 2 2
P P0 P
LI = 10 lg = 10lg 2 + 10lg = LP + 10lg 0 ------------(2)
I0 P0 ρc I 0 ρc I 0

For the given reference values of I0 and P0(10-12Wm-2 and 20x10-6 Pa respectively) the term 10

( )
2
P0
lg is small when compared with Lp for most measurements in air at ambient conditions(0, 13dB
ρc I 0
at 22°C and 1.0101*10-5 Pa).
Sound pressure level measurements and sound intensity level measurements performed outside the
near-field i.e. in an ideal free field environment will therefore have the same numerical value.

Sound pressure method:

Sound pressure is a scalar quantity without directivity information; it is measured with a single
measurement microphone. Because the measurement device can locally disturb the sound field
significantly it is good practice to point the measurement microphone towards the test object.

Practical test environments usually differ significantly from the ideal free field condition. Transformer
sound pressure measurements are generally adversely affected as described below. Test measurements
will therefore require a correction based on an understanding of the acoustical properties of the test
environment as described below,

Calculation of the spatially averaged sound pressure level

For the walk-around procedure the instrument will automatically provide the spatially averaged
measurement data. In case of the point-by-point procedure the spatially averaged measurement data
may also be derived automatically by the instrument via post-processing or it has to be calculated as
described below. When required to report point-by-point measurements for each microphone position
it may be necessary to derive the total spatially averaged sound pressure level by calculation. The total
spatially averaged A-weighted sound pressure level for the test measurement L p A 0 shall then be
calculated from the total A- weighted sound pressure levels. L p A i . measured at the individual
microphone positions using Equation (3):

N
L p A 0 = 10 x lg( 1 ∑ 100.1 L ) ……….(3)
p Ai

N j=1

where

N is the number of microphone positions;

L pA i. is the measured total A-weighted sound pressure level of the test measurement at the
microphone position.

The total spatially averaged A-weighted background noise level, LbgA .. shall be calculated before and
after the test sequence using Equation (4):
N
1
LbgA = 10 x lg( N ∑ 100.1 L ) ……….(4)
bgA i

j=1

Where,

M is the number of microphone positions:

LbgA is the measured total A-weighted background noise level at the microphone position.
The same calculation procedure applies for each individual band v of the frequency spectrum resulting
v
in spatially averaged A-weighted sound pressure levels L p A o and spatially averaged background noise
v
levels Lb gA .

Validation of test measurements with respect to background noise

For practical purposes the validation process described below is based on the total spatially averaged A-
weighted sound pressure level L p A 0 and the total background noise level Lbg A Examination of individual
bands of the frequency spectrum is not required.

When initial and final background noise levels Lbg A differ by more than 3 dB and when the higher value
is less than 8 dB below the A-weighted sound pressure level of the test measurement L p A o the test
measurement shall be declared invalid and the test repeated. However, in cases where the test
measurement meets the guarantee, correction for background noise is not required. In this case the test
is declared a pass.

If the greater of the two background noise levels Lbg A is less than 3 dB below the A-weighted sound
pressure level of the test measurement L p A o the test measurement shall be declared invalid and the
test repeated. However, in cases where the test measurement meets the guarantee, correction for
background noise is not required. In this case the test is declared a pass.

Whilst this standard permits small differences between background noise and test measurement sound
levels, every effort should be made to obtain a difference of about 6 dB. When the difference becomes
less than 3 dB, the use of alternative measurement methods may be considered. The above requirement
is summarized in the following table

Calculation of environmental correction factor K:

The environmental correction K, expressed in dB, accounts for the influence of undesired sound
reflections from room boundaries and/or reflecting objects within the test area. The magnitude of X
depends principally on the ratio of the sound absorption area of the test room, A, to the area of the
measurement surface, S. The magnitude of K is not strongly influenced by the location of the test object
in the test room and K does not correct for measurements influenced by standing waves.

K shall be derived from Equation (5) or Figure 7 by entering the appropriate value of A/S.
4
K=10 x lg (1+ A /S )…………..(5)
If K is determined based on reverberation time the value of A is given by Sabine’s Equation
A = 0.16(V/T)………………(6)
Where
V is the volume of the test room in cubic meters;
T is the reverberation time of the test room in seconds.
If K is determined based on absorption coefficient the value of A is given by

A = ∑ α i s vi ………..(7)
i

Where
αi is the absorption coefficient for a partial surface(shown in the following table)
s vi is the area of the partial surface of the test room(walls, ceiling and floor) characterized by
α i in square meters.

The environmental correction factor K shall preferably be determined by measurement techniques.

However, for the purpose of this standard estimation for K is allowed by use of absorption
coefficients.For a test room to be satisfactory, the ratio A/S shall be ≥ 1 with a corresponding value for
the environmental correction factor K≤ 7 dB. However, if the ratio A/S is ≤ 2.5 with a corresponding
value for the environmental correction factor K ≥ 4.1 dB, then the determination of K shall be based on
acoustical measurements. Where conditions are close to free-field, i.e. essentially undisturbed by
reflections from nearby objects and the environment boundaries, as sometimes achieved for outdoor
measurements. then the value for A would tend to zero and no environmental correction is necessary.

Figure : Distribution of disturbances to sound pressure in the test environment.

Sound intensity method:
Sound intensity is a vector quantity which indicates the direction of sound propagation and therefore
allows the sound power entering and the sound power leaving through the measurement surface
around the test object to be distinguished. This characteristic allows measurements in non-ideal test
environments without corrections. The magnitude of the sound intensity is the time-averaged product
of the pressure and particle velocity.

As already stated, a single microphone can measure pressure; however, measuring particle velocity is
more complex. Particle velocity is related to the pressure gradient, i.e. the rate at which the
instantaneous pressure changes with distance.

The principle of particle velocity measurement is based on Newton's second law applied to air. Newton's
second law relates the acceleration given to a mass to the force acting on it. If the force and the mass
are known, the acceleration can be found and then integrated with respect to time to find the velocity.
In a sound wave, the pressure gradient accelerates the air of density ρ .

With the knowledge of the pressure gradient and the density of air, the particle acceleration can be
calculated using following Equation8 :

−1 δp
a = - ρ δr ………………….(8)

where a is the particle acceleration due to a pressure change δp in air with density ρ across a distance
δ r. Integrating the above, Equation 9 gives the particle velocity as follows:
1 δp
u=−∫ ( ) ⅆt ……………….(9)
ρ δr
It is possible to measure the pressure gradient with two closely spaced microphones, A and B, separated
by a spacer of length Ar. With a finite difference approximation of Equation (4) the pressure gradient can
be obtained by taking the difference in their measured pressures P A and PB and dividing it by the
distance Δr between them.

Figure : Microphone Arrangement

The pressure gradient signal is then integrated to give the time-averaged particle velocity was shown in
Equation (10):

−1 p A − p B
u= ∫ ⅆ t ……………….(10)
ρ Δr
Sound intensity is subsequently calculated as the product of the sound pressure in the midst of the
microphone pair and the time-averaged particle velocity:

−p A + p B p A− p B
I= ∫ ⅆ t …………….(11)
ρ Δr
This is the basic principle of signal processing in sound intensity measuring equipment.
The principle of the sound intensity method theoretically accounts for the disturbances discussed above;
in practical test environments however, there are limitations for its application.
Where the difference (PA-PB) in pressure between the two microphones is small compared to the
absolute pressure magnitude (PA+PB)/2 the determination of the pressure gradient tends to become
inaccurate.
An indication of this effect can be derived from the difference between the uncorrected sound pressure
level containing the disturbances and the sound intensity level of the sound power propagating from the
test object alone. Experiments have shown that sound intensity measurement accuracy strongly
correlates to this difference ΔL is called the P-I index:
ΔL=LPA 0−LI A ………………………(12)

It has to be recognized that all types of disturbance contribute to the P-I index and the larger a
disturbance, the larger is the P-I index.
The larger the P-I Index, the higher is also the tendency to underestimate the sound intensity level. For
this reason, the maximum permissible P-I index has to be limited to assure the measurement quality
(For detail, see 11.3.5 of IEC 60076-10:2016). It is also noted, that in such situations the sound pressure
is overestimated.
Sound waves leaving the radiating surfaces of the test object will be reflected from the floor, walls and
other objects in the test room. Due to the sensitivity of the sound intensity probe to the sound
propagation direction, sound intensity measurements are usually not adversely affected by such
reflections. In the situation of a highly or fully diffuse sound field (for instance inside an acoustic
enclosure) the sound intensity is close to zero and the P-I index is high, indicating that the measurement
is rejected.
Standing waves should be avoided or minimized by careful placement of the test object in the test
room. At locations where standing waves occur, the measuring device will measure intensity close to
zero because sound pressure and particle velocity are out of phase. In consequence the intensity level
will be reduced and this is indicated by a high P-I index.
Practical test environments are often affected by noise from external sources such as manufacturing
processes, the test generator and other energized test transformers. Figure 12 indicates such an
environment with steady-state background noise propagation from left to right across the test object.

Microphone pair positions are indicated by the white microphone A and the black microphone B.

Figure: Illustration of background sound passing through test area and sound radiated from the test object
With the test object not energized there will be negative intensity measured on the left side, positive
intensity on the right side and no intensity measured on the long sides. The spatially averaged intensity
will therefore be zero. When the test object is energized there will be an additional amount of positive
intensity on all four sides emitted. As mentioned above the net effect on the intensity due to the
background noise is zero and therefore has no effect on the test objects intensity. It is noted that this
does not apply for the sound pressure.

The ideal situation above does not hold true if the background noise is high relative to the test object
noise. In such situations the intensity on the left (short) side is small or even negative and on the right
(short) side the intensity is increased. This is not a problem as long as the absolute pressure magnitude
(PA + PB)/2 does not swamp the pressure difference (PA - PB). On the top and bottom (long) sides of the
test object this effect is more severe because both microphones are exposed to the same absolute
pressure increase due to the background noise. The P-I index accounts for this and it is essential to
derive the P-I index only from the spatially averaged measurements along the entire (closed)
measurement path around the test object.

In summary, the sound intensity method accounts for steady-state background noise but only up to a
certain extent. With increasing levels of steady-state background noise, the measured sound intensity
level of the test object decreases which is obviously unacceptable. At the same time the P-I index value
increases. Working within the limits for the P-I index as stated in 11.3.5 of IEC 60076-10:2016 maintains
the acceptability of the measurement.

Sound Intensity measurements distinguish between active and reactive sound fields; therefore, near-
field effects will not adversely affect the measured sound intensity level.

It is inherent to the sound intensity method that the measurement surface and therefore the
measurement path shall completely encircle the test object. Whilst spot indication of sound intensity
and P-I index can be informative to identify locations with problematic sound emission, measurements
performed to estimate sound power and the corresponding P-I Index shall be derived from
measurements along the entire measurement path.

Test Methodology:

Microphone Positions:

For the walk-around procedure, the microphone shall be moved with a constant speed of maximum 1
m/s on the prescribed contour(s) around the test object. At the given walking speed, the sampling rate
of modern integrating sound level meters is always sufficient for accurate spatial averaging up to a
resolution of 1/3-octave. The spatially averaged sound level over the measurement duration shall be
recorded together with the active measurement duration in the test report.

The "START STOP" and "PAUSE" functions of such sound level meters can be used to simplify the
measuring procedure, i.e. to negotiate obstacles and/or to change between prescribed contours.

For the point-by-point procedure, the microphone positions shall be on the prescribed contour(s),
equally spaced and not more than 1 m apart (see dimension D in Figures 1 to 5). There shall be a
minimum of eight microphone positions along each contour. The measuring duration shall be a
minimum of three seconds and be practically the same duration at each position.
It can be necessary to modify some measuring positions for certain test objects for safety reasons, for
example, in the case of transformers with horizontal high voltage bushings where part of the prescribed
contour(s) may be confined to the safe zone.

Calculation of Measurement Surface Area:

Measurement surface area for measuring distances up to 30 m

The area S of the measurement surface, expressed in square metres, is given by Equation (13):
S=(h+x) Im …………….(13)
Where,
h is the height of the principal radiating surface in meters
Im is the length metres of the prescribed contour;
x is the measurement distance in meters from the principal radiating surface to the prescribed
contour.
NOTE 1 Equation (8) applies for the measuring distances of 0,3 m, 1 m, 2 m but also any other measuring
distance up to 30 m.
NOTE 2 Equation (8) is also applicable for the calculation of the sound pressure level from the sound
power level.
Measurement surface area for measuring distances larger than 30 m

The area S of the measurement surface (a hemisphere), expressed in m2 is given by Equation (14)

S=2 π R2…………..(14)
Where, R is the distance in meter(m) from the geometrical center of the transformer/cooling device to
the considered location in the far-field.

Selection of appropriate sound measurement method:

Both sound pressure method and sound intensity method can be used for the estimation of sound
power and they are expected to give comparable values. The sound intensity method inherently
accounts for the physics of sound and this is reflected in the fact that there is no requirement for
corrections. Measured results are acceptable provided the limits for the P-I index (11.3.5 of IEC 60076-
10:2016) are maintained. The value of 4 dB as the base limit is introduced after careful consideration
and evaluation of the effects imposed by each individual type of disturbance and is backed-up by
practical experience. Such experience also revealed that sound power estimations based on sound
intensity measurements within this limit represent the true value of the sound power more accurately
than sound power estimations based on corrected sound pressure measurements. Beyond the base limit
of 4 dB for the P-I index, sound intensity tends to underestimate and sound pressure tends to
overestimate sound power. In this situation the reported sound intensity level is assumed to be 4 dB
below the measured sound pressure level, see 11.3.5 of IEC 60076-10:2016. This approximation avoids
the risk of sound power underestimation.

The upper limit for the P-I index of 8 dB given in 11.3.5 of IEC 60076-10:2016 is intended to maintain a
minimum quality of the test environment. Where the above conditions cannot be met, ie. the P-I index
exceeds 8 dB and where the sound pressure level is believed to overestimate the sound power.
Example:

Noise Level Test of Transformer:

Parameters:

Height of the Transformer = 3.90 m

Height 1
(1/3 height of the transformer) = 1.30 m

Height 2
(2/3 height of the transformer) = 2.60 m

Room Height = 18.01 m

Room Length = 20.30 m

Room Width = 11.90 m

Contour Length = 18.10 m

Total Number of Test Positions = 20

Distance between Positions:

Positions Distance (m) Positions Distance (m)

1–2 0.92 11 – 12 0.92

2–3 0.91 12 – 13 0.88

3–4 0.95 13 – 14 0.93

4–5 0.88 14 – 15 0.90

5–6 0.83 15 – 16 0.92

6–7 0.90 16 – 17 0.93

7–8 0.96 17 – 18 0.85

8–9 0.88 18 – 19 0.87

9 – 10 0.84 19 – 20 0.89

10 – 11 0.93 20 – 1 0.96
Sound Level For Rated Voltage:
Position LpA-NL (transformer empowered) LbgA-NL (background sound)
No. Height 1 Height 2 Height 1 Height 2

dB(A) Pressure (Pa) dB(A) Pressure (Pa) dB(A) Pressure (Pa) dB(A) Pressure (Pa)

1 56.2 0.012913 56.5 0.013367 54.9 0.011118 54.6 0.010741

2 55.9 0.012475 55.9 0.012475 55.1 0.011377 55 0.011247

3 55.6 0.012051 55.6 0.012051 55.3 0.011642 55.2 0.011509

4 54.9 0.011118 54.8 0.010991 54.9 0.011118 54.8 0.010991

5 55.8 0.012332 55.6 0.012051 54.8 0.010991 54.6 0.010741

6 55.3 0.011642 55.6 0.012051 55.8 0.012332 55.8 0.012332

7 55.6 0.012051 55.9 0.012475 55.7 0.012191 55.6 0.012051

8 54.9 0.011118 55.1 0.011377 55.1 0.011377 55.2 0.011509

9 56.1 0.012765 56.2 0.012913 55.2 0.011509 55.4 0.011777

10 56.2 0.012913 56.1 0.012765 55.2 0.011509 55.1 0.011377

11 56.1 0.012765 56.4 0.013214 54.9 0.011118 55.1 0.011377

12 55.7 0.012191 55.8 0.012332 55.6 0.012051 55.5 0.011913

13 55.4 0.011777 55.6 0.012051 55.4 0.011777 55.6 0.012051

14 55.6 0.012051 55.7 0.012191 55.3 0.011642 55.4 0.011777

15 55.2 0.011509 55.4 0.011777 55.6 0.012051 55.7 0.012191

16 55.8 0.012332 55.9 0.012475 55.2 0.011509 55.3 0.011642

17 56.1 0.012765 56.2 0.012913 55.6 0.012051 55.8 0.012332

18 55.9 0.012475 56.4 0.013214 55.2 0.011509 55.4 0.011777

19 56.2 0.012913 56.3 0.013063 54.8 0.010991 54.7 0.010865

20 56.3 0.013063 56.5 0.013367 54.8 0.010991 54.9 0.011118

Average Sound Pressure = 0.012358 Avg. Sound Pressure = 0.011554

Transformer Noise + Background Noise = 55.82 Background Noise = 55.23

Noise Level of Transformer for Rated Voltage without Environmental Correction

 LPO −NL=10 log ( 100.1 × 55.82−100.1 ×55.23 )=46.86 dB

Sound Level For Rated Current Test:

Position LpA-FL (transformer empowered) LbgA-FL (background sound)

No. Height 1 Height 2 Height 1 Height 2

Pressure
dB(A) Pressure (Pa) dB(A) Pressure (Pa) dB(A) Pressure (Pa) dB(A)
(Pa)

1 56.1 0.012765 56.4 0.013214 54.9 0.011118 54.6 0.010741

2 55.9 0.012475 55.8 0.012332 55.1 0.011377 55 0.011247

3 55.6 0.012051 55.7 0.012191 55.3 0.011642 55.2 0.011509

4 55.6 0.012051 55.8 0.012332 54.9 0.011118 54.8 0.010991

5 55.9 0.012475 55.6 0.012051 54.8 0.010991 54.6 0.010741

6 55.1 0.011377 55.6 0.012051 55.8 0.012332 55.8 0.012332

7 54.9 0.011118 54.8 0.010991 55.7 0.012191 55.6 0.012051

8 56.1 0.012765 56.3 0.013063 55.1 0.011377 55.2 0.011509

9 55.8 0.012332 55.7 0.012191 55.2 0.011509 55.4 0.011777

10 55.8 0.012332 55.9 0.012475 55.2 0.011509 55.1 0.011377

11 55.6 0.012051 55.4 0.011777 54.9 0.011118 55.1 0.011377

12 55.7 0.012191 55.3 0.011642 55.6 0.012051 55.5 0.011913

13 55.7 0.012191 55.6 0.012051 55.4 0.011777 55.6 0.012051

14 55.9 0.012475 55.8 0.012332 55.3 0.011642 55.4 0.011777

15 55.9 0.012475 55.7 0.012191 55.6 0.012051 55.7 0.012191

16 56.2 0.012913 56.1 0.012765 55.2 0.011509 55.3 0.011642

17 55.7 0.012191 55.9 0.012475 55.6 0.012051 55.8 0.012332

18 55.9 0.012475 55.8 0.012332 55.2 0.011509 55.4 0.011777

19 55.9 0.012475 56.1 0.012765 54.8 0.010991 54.7 0.010865

20 56.1 0.012765 56.5 0.013367 54.8 0.010991 54.9 0.011118

Average Sound Pressure = 0.012313 Avg. Sound Pressure = 0.011554

 Transformer Noise + Background Noise = 55.79 Background Noise = 55.23

Noise Level of Transformer for Rated Current without Environmental Correction

LPO −FL =10 log ( 100.1 ×55.79−10 0.1× 55.23 )=46.62dB

Calculations for Environmental Correction:

( ) ( )
N M
1 1
L pA 0=10 log
N
∑ 10 0.1 L pAi
, LbgA =10 log ∑
M i =1
10 0.1 L bgAi
,
i=1

2
Principal radiating surface, S=1.25× h l m=1.25 ×3.90 m ×18.1 m=88.24 m

Where, lm is the length of the contour.

The average acoustic absorption coefficient, α = 0.15 (Room with furniture, rectangular machinery
room, rectangular industrial room)

The total area of the surface of the test room,

2
SV =2×(18.01 ×20.30+ 18.01× 11.90+20.30 ×11.90)=1642.984 m
2 2
The sound absorption area of the test room, A=α SV =0.15 ×1642.984 m =246.448 m

Environmental Correction Factor, K=10 log 1+ ( 4

A /S )
=3.86

The corrected average A-weighted sound pressure level for rated voltage,

LPA −NL=LPO −NL−K =46.86−3.86=43.00 dB

The A-weighted sound power level of the test object for rated voltage,

LWA−NL=L PA−NL +10 log ( SS )=62.46 dB

O

The corrected average A-weighted sound pressure level for rated current,

LPA −FL=L PO−FL −K=46.62−3.86=42.76 dB

The A-weighted sound power level of the test object for rated current,

LWA−FL =LPA− FL +10 log

( SS )=62.22 dB
O

2
Where, Reference Area, SO =1 m

Total Noise Level:

LWA =10 log ( 100.1 × L WA−NL

+100.1 × L WA −FL
) =65.35 dB