4 Acoustic Noise and its Measurement
Definitions
Noise level meters, unless very cheap, have primarily two settings, giving readings in dBA and LIN (or dBC).
dBA or dB(A) – either seems currently acceptable in the UK. This is a measurement which gives a reasonable approximation to loudness as perceived by the average person. The ‘A’ weighting network gives the measuring instrument a response which allows for the ear’s lack of sensitivity to low frequencies. See Figure 10.
Figure 10 A, C and LIN frequency weightings
dB(C). Some noise level meters, particularly in the USA, have a setting marked ‘C’. (Older meters may have a ‘B’ setting. This represents a weighting network which is between ‘A’ and ‘C’ characteristics. This has fallen into disuse.)
LIN. Flat response.
Leq,T – equivalent noise level. This is the sound level which, integrated over a period of time, T, is equivalent to a steady SPL.
If, as is usual, measurements are made with the A weighting network then this is indicated by the subscript A: LAeq,T.
LAeq,T is used widely in assessing the likely effect of industrial, and other noise sources such as roads, on residential areas.
The British Standard BS 4142: 1990 (‘Rating industrial noise affecting mixed residential and industrial areas’) should be consulted for more information. See the note on page 27.)
Other terms found in various aspects of noise measurement are:
Ln where n is a number, often 10, 50 or 90, meaning that the SPL indicated by L is exceeded for n per cent of the time. Thus if L90 were 60 dBA the value of 60 dBA would be exceeded for 90 per cent of the time under consideration.
LeP,d is very similar to LAeq,T in that essentially the same type of equipment is used to measure it. It represents, however, the personal daily ‘dose’ of noise. In effect, then, LAeq,T applies to locations, LeP,d applies to people.
Hearing Damage
The sound levels which are a risk to a person’s hearing can be a subject of some controversy, especially as it seems that different individuals may have different susceptibilities. However it is generally accepted in the UK and much of Europe that there is little risk of hearing damage if the ears are exposed to an LA of 90 dB for 8 hours each day.
Above 90 dB the permissible exposure time is halved for each 3 dB increase. See table opposite.
LAeq |
Permissible exposure time |
90 |
8 hours |
93 |
4 hours |
96 |
2 hours |
99 |
1 hour |
102 |
30 mins |
105 |
15 mins |
108 |
8 mins |
111 |
4 mins |
Current health and safety regulations
In the UK these can be summarized by stating that there are two Action Levels:
1. If the LeP,d for the employee(s) exceeds 85 dBA but is less than 90 dBA then the employer must inform the employee(s) and ensure that suitable hearing protection is available to all who ask for them.
2. Above 90 dBA or when the peak sound pressure exceeds 200 Pa (140 dBA) the employer must reduce noise levels as far as is practicable by means other than ear protectors and mark all zones of noise levels higher than 90 dBA.
The above is the barest of outlines and HSE literature should be consulted for more information.
Sound level meters
There are four defined grades:
Grade 0: the very highest standard and mainly used for laboratory work;
Grade 1: for precision measurements;
Grade 2: for general purposes;
Grade 3: for sound surveys.
Most meters except perhaps Grade 3 will have the ability to measure the following:
‘Impulse’ or ‘Peak’ readings
‘Fast’, which averages over 1⁄8 s
‘Slow’, which averages over 1 s.
Noise measurements in the field should normally be made with at least 1.2 m of clear space (including the ground) all round the microphone. The noise being measured should be at least 10 dB above background noise. If not a correction should be applied as in the table below.
Noise level reading, LAeq,T minus background, LA90,T dB |
Correction: subtract from noise level reading dB |
6 to 9 |
1 |
4 to 5 |
2 |
<3 |
Difficult to correct |
Addition of Noise Levels
Noise levels in dB cannot meaningfully be added together, as, being logarithms the result will be a multiplication.
Antilogs of the readings must be used, as shown below. Note that the noise reading should be divided by 10 before finding the antilog as antilogs of 100 or more will result in an ERROR indication on normal calculators. To take an example:
Find the sound level resulting from 92 dBA + 98 dBA + 96 dBA + 68 dBA
Result = 10 × 10.075 = 100.75 dBA
(Note that the contribution of the reading of 68 dBA is insignificant, being some 30 dB below the other readings.)
The above method, adjusted to take time durations into account, may be used to find an LeP,d from a series of measurements of LAeq.
National and International Standards
In the list below British Standards relating to acoustic noise and its measurements are given with equivalent standards in bold.
The following abbreviations are used:
| ANSI | American National Standards Institute |
| BS | British Standards Institute |
| IEC | International Electrotechnical Commission |
| ISO | International Standardization Organization |
BS 3539: 1986
Sound level meters for the measurement of noise emitted by motor vehicles
*BS 4142: 1990
Method for rating industrial noise affecting mixed residential and industrial areas
BS 5969: 1981 (IEC 651: 1979)
Specification for sound level meters
BS 6402: 1994 (IEC 1252: 1993)
Personal sound exposure meters
BS 6698: 1986 (IEC 804: 1985)
Integrating–averaging sound meters
BS 7580: 1992 (IEC 645–1: 1992)
Specification for the verification of sound level meters
ANSI S1.4 – 1983
Specification for sound level meters
ANSI S1.25 – 1991
Specification for personal dosemeters
* Most of BS 4142: 1990 is relevant to any outdoor measurement. It can be regarded as a safe procedure to use in many applications additional to industrial noise, such as traffic, entertainment and animal sounds, such as kennels.