Noise Health Home 

[Download PDF]
Year : 2003  |  Volume : 5  |  Issue : 20  |  Page : 63--73

Impulse noise and risk criteria

J Starck1, E Toppila1, I Pyykko2,  
1 Department of Physics, Finnish Institute of Occupational Health, Helsinki, Finland
2 Tampere University Hospital, Dept. of Otolaryngology, Tampere, Finland

Correspondence Address:
J Starck
Department of Physics, Finnish Institute of Occupational Health, 00250 Helsinki


Impulse noise causes evidently more severe hearing loss than steady state noise. The additional effect of occupational impulse noise on hearing has been shown to be from 5 to 12 dB at 4 kHz audiometric frequency. Reported cases for compensated for hearing loss are prevalent in occupations where noise is impulsive. For impulse noise two measurement methods have been proposed: the peak level method and energy evaluation method. The applicability of the peak level method is difficult as even the recurrent impulses have different time and frequency characteristics. Various national risk criteria differ from international risk criteria. In France the maximum A-weighted peak level is 135 dB, and in the United Kingdom the C-weighted peak sound pressure is limited to 200 Pa (140 dB). This criterion of unweighted 200 Pa (140 dB) is used in European Union (EU) directive 86/188 and ISO 1999-1990 regardless of the number of impulses. The American Conference of Governmental Industrial Hygienists (ACGIH) has recommended that no exposure in excess of a C-weighted peak sound pressure level of 140 dB should be permitted. At work places these norms do not cause any practical consequences since the impulses seldom exceed 140 dB peak level. In several occupations the impulses are so rapid that they contribute only a minimal amount to the energy content of noise. These impulses can damage the inner ear even though they cause reduced awareness of the hazard of noise. Based to the present knowledge it is evident that there is the inadequacy of the equal energy principle in modelling the risk for hearing loss. The hearing protectors attenuate industrial impulse noise effectively due to the high frequency contents of impulses. Directive regarding the exposure of workers to the risks arising from noise requires that in risk assessment attention should be paid also to impulsive noise. So far there is no valid method to combine steady state and impulse noise. A statistical method for the measurements of industrial impulse noise is needed to get a preferably single number for risk assessment. There is an urgent task to develop risk assessment method and risk criteria for impulsive noise to meet the requirements of the upcoming European Union noise directive.

How to cite this article:
Starck J, Toppila E, Pyykko I. Impulse noise and risk criteria.Noise Health 2003;5:63-73

How to cite this URL:
Starck J, Toppila E, Pyykko I. Impulse noise and risk criteria. Noise Health [serial online] 2003 [cited 2023 Jan 30 ];5:63-73
Available from:

Full Text


Impulse noise is often defined as noise consisting of single bursts with a duration of less than one second with peak levels 15 dB higher than background noise. Impulse noise differs from steady state noise by the properties in the time domain. Impulse noise contains rapid sound pressure transients. The physical properties of impulse noise are characterised by peak level, rise and decay time, duration and number of impulses, spectral content, and level distribution (Starck et al., 1988; Pekkarinen et al., 1993; Hamernik and Hsueh, 1991; Patterson, 1991; Lahti and Starck, 1980). The criteria presented for impulse noise are based on the repetition rate of peaks of sound pressures. The procedures for hygienic evaluation of impulse noise are not uniform, due to the uncertainty of the existing knowledge.

Risk Criteria

At present there are limits for impulse noise to protect workers hearing that are based ondifferent parameters in time and frequency domain. European union directive (86/188/EEC, 1986) and ISO 1999 standard (ISO 1999, 1990) limits the unweighted peak level under 140 dB. The same limit value but C-weighed is used in UK and ACGIH (HMSO 1987; ACGIH 2001). The French guidance document limits the A­weighted peak level under 135 dB (INRS 1989). ISO 9612 specifies impulse adjustment that is based to the difference between A-weight impulse equivalent level (IEC 179A- 1973 impulse level) and A-weighted equivalent level (ISO 9612- 1997). For the differences more than 2 dB, 3-6 dB adjustment can be used. In practice 6dB adjustment means 5 - 8 dB additional median hearing loss at 4 kHz after 30 years of exposure to 85-95 dB noise (ISO 1999-1990).

Audibility of impulse noise

The human auditory system will provide full audibility when duration of a sound exceeds 200 ms at constant level. For shorter duration of sound less loudness is perceived. After reviewing several studies Bruel (1975) suggested that "fast" time constant (IEC 179A­1973) is the best choice to correspond to the perceived loudness of the sound transients shorter than 200 ms. [Figure 1] displays the reduction of the reading when different time constants are used for the various duration of impulses (IEC 651- 1979). The [Figure 1] also compares the standardised time constants to the responses of hair cells in the inner ear. As shown in [Figure 1] the inner ear is about 1000 times faster than the fastest time constant "impulse" (Bruel, 1975).

Correct equivalent levels will be achieved when using slow or fast time constant in the measurements. On the contrary impulse time constants do not provide any energy dependent result as the rise time and decay time are not symmetric. In "impulse" mode the rise time is 35 ms and decay time 3 s that is 100 times longer than for the increasing level (IEC 651, 1979). The meter indication will be higher than the corresponding energy because the long decay time constant will increase the surface area of impulse time function curve that is integrated in energy calculations.

Definition of duration of single impulse

The risk of hearing loss from various types of impulses is proposed to be evaluated according to the criteria based on the peak level and duration of impulses (Pfander 1975; Ward, 1968). Idealised impulse duration are defined for gunshot and hammer blow [Figure 2].

The A-duration is defined as the time from the onset of the impulse to the first zero level crossing. The B-duration is defined as the time during which the envelope of pressure fluctuation stays within 20 dB of the peak pressure level, also including the recurrent impulses [Figure 2].

Exact characterisation of industrial impulse noise is complicated due to variations in spatial and temporal properties [Figure 3],[Figure 4]. The noise dose does not give any information on the rapid variations of the signal. Impulse noise analyses are conventionally based on observations of the time history of instantaneous sound pressure. This method is impractical for industrial noise measurements because even recurrent impulses are dissimilar and none of them resemble ideal impulses shown in [Figure 2].

The impulse noise samples in [Figure 3] and [Figure 4] demonstrate the need for statistical analyses of the industrial impulse noise. A method that has been developed is based to the continuous calculation of crest factor. The crest factor is defined as the difference between peak and rms levels of the noise [Figure 5]. The crest factor is also in relationship to the effective length and to the energy of a single impulse (Bruel, 1975). The calculation can be done by dividing the sample to short intervals to calculate further their cumulative distribution function. The definition of impulse noise based to the effective length (Starck &Pekkarinen, 1987) will get the form:

I = L Ap - L AS >/= 15 dB

where L Ap is the A-weighted peak level and LAS is A-weighted rms level measured with slow time constant and I is the impulsiveness. The numerical value of 15 dB corresponds to an effective length of 31.5 ms and is close to the audibility of impulses.

The cumulative distribution function can be calculated for the difference between peak and rms level of the recurrent intervals [Figure 6]. This function describes the percentage of time F(I) in the sample during which a certain impulsiveness value of I is exceeded. According to the criteria impulse percentage of F 15 will be selected from the cumulative distribution function of I. Thus the impulse percentage F 15 shows the time in the sample when the impulsiveness have a value greater than 15 dB (Starck and Pekkarinen, 1987).

Effect of industrial impulse noise to hearing

To evaluate the effect the effect of impulsiveness of noise on hearing a cross-sectional study on three different groups of workers in Finland: forest workers (N = 199), shipyard workers (N = 179) (Starck and Pekkarinen, 1987). In a selected group of workers exposure to noise was evaluated both outside and inside the hearing protectors (HPD) to define the lifetime equivalent noise exposure level and impulsiveness by the crest factor method. The exposure data was used to predict the noise­induced hearing loss at 4 kHz by applying ISO 1999-1990 model that takes into the consideration exposure to noise using A­weighted equivalent level, age and gender [Table 1]. To evaluate the role of confounding factors the medical histories of the workers, use of drugs, serum cholesterol levels, blood pressure and tobacco smoking of the past three years were retrieved from the charts or queried. Special attention was paid to hearing protection - the types of hearing protectors used and the usage rates was questioned. All data were input into the database of NoiseScan that has routines for individual exposure evaluation (Pyykko et al., 2000).

Measured and predicted hearing levels were consistent among forest workers. Among shipyard workers measured hearing loss was 10 dB higher than the predicted one. The shipyard workers exposure to noise was impulsive as expressed by the impulse percentage in [Table 1]. Outside the HPD the impulse percentage was 4.2 % indicating duration in the noise sample when impulsiveness (I) is above 15 dB.

Totally 36 matched pairs were established with regard to confounders to compare the effect of impulse noise between forest workers and shipyard workers.

For the matched pairs the difference between the measured and predicted hearing loss at 4 kHz was 8 dB for the shipyard workers and 2 dB for the forest workers. The shipyard workers had somewhat more aggravated hearing loss than the forest workers but the difference was not statistically significant (p Shooting noise

Shooting impulses are recognised cause of noise­induced hearing loss (Salmivalli, 1967; Coles, 1968; Johnson, 1982). The prevalence of hearing loss is as high as 40-60 % among men who have exposed to gunfire noise, (Salmivalli, 1967; Walden, 1975; Ylikoski, 1987).

Exposure to shooting noise concerns male citizens in the countries where military service is mandatory. In Finland about 25 000 young men annually serve as conscript. Annually observed acoustic traumas have decreased from 709 to 231 in Finland during the last 10 years [Figure 8]. In about 80% of the cases the acoustic trauma is due to the shots with blanks. In target shooting usage of HPDs is controlled. More over safety regulations will specify the type of HPD and if needed also the combined use of earmuff and earplug for large calibre firearms.

Simultaneous usage of earmuff and earplug will increase the protection efficiency that is less than the sum of separate HPDs. An earmuff with a large cup volume will provide higher protection efficiency as small cup earmuff [Table 3].

[Figure 8],[Figure 9] show the time functions for the cannon shot inside and outside for two HPD combinations. In both cases it can be seen that high frequency components had attenuated most effectively. Also rise time has increased and the duration of the shot impulse have expanded inside the HPD.


The present paper aims to summarise the current knowledge on impulse noise and its effect to hearing. Industrial impulse noise causes in general about 5-12 dB more severe hearing loss than steady state noise. The exposure to impulsive noise is often composed of very rapid sound bursts that have short duration and low energy content. As a consequence the audibility is lower than the actual level of the impulses. So far there is no valid method to combine steady state and impulse noise. A statistical method for the measurements of industrial impulse noise is needed to get a preferably single number for risk assessment. The HPDs attenuate industrial impulse noise effectively but do not prevent from advanced hearing loss among workers.

The equal energy principle provides a good approximation for the vulnerability of hearing in steady state noise. However, the time domain characteristics of noise have been shown to affect the harmfulness of noise; the risk of NIHL is higher in the occupations where workers are exposed to impulse noise. In several occupations the impulses are so rapid that they contribute only a minimal amount to the energy content of noise. For example in impulsive noise among shipyard workers caused about 10 dB greater hearing loss than could be predicted by the ISO 1999-1990 model. The observed hearing levels were consistent with the model for forest workers, where the noise was not impulsive (Starck et al., 1988). The impulsive noise seems to be more harmful for hearing at high noise exposure levels (Toppila et al., 2000). The threshold level for industrial impulse noise causing increased risk of hearing loss is not known.

A recent study among force hammering workers showed that the hearing loss of workers exposed to very high impulsive noise correlated significantly with peak levels and number of impulses in combination (Suvorov et al., 2001). Technically most advanced method is based to the crest factor definition. It is physically well documented and allows statistical analysis for industrial impulse noise. Epidemiological field studies show that the method correlates with hearing loss risk of impulse noise (Starck et al., 1987; Starck et al., 1988; Toppila et al., 2000).

The HPDs attenuate industrial impulse noise more effectively than steady state continuous noise. This is due to the high frequency contents of impulses, which are attenuated effectively in earmuffs. Even though the earmuffs reduce the impulse noise rate, workers in the metal industry are still exposed to more impulsive noise than workers in paper mills and forestry (Starck et al., 1988). Hearing protection has proved to be less effective against impulses from large calibre fire-arms due to the non-linearity of the attenuation and the low frequency components (Ylikoski et al., 1987; Starck et al., 1987). Shooting with large calibre weapons produces impulses greater than 140 dB. In these instances a combination of ear plugs and muffs are recommended (Starck et al., 1987).

Shooting and hunting produce additional hearing loss to workers exposed to steady state noise (Pekkarinen et al., 1993). Forest workers who were exposed to gunfire noise had an additional 10 dB hearing loss compared to those who had only occupational exposure to chain saw noise (Pekkarinen et al., 1993). NIHL occurs at a younger age in the military than in other groups of workers exposed to excessive noise (Ylikoski et al., 1995). Exposure to gunfire noise is difficult to assess, since there is no standard method available to evaluate its effect on the inner ear.

The existing impulse measurement methods can be divided into two categories; the peak level methods and energy attenuation methods. With the peak level methods (Pfander, 1975; CHABA, 1968; ACGIH, 2001) the risk for hearing loss is related to the peak level and duration. These methods do not provide a way of combining different gunfire exposures or gunfire exposure with work noise exposure to a single exposure index. The latest approach is to apply the energy attenuation of the impulse in risk assessment (Dancer et al., 1996; Patterson and Johnson, 1996; ANSI 2000).

The new amendment proposal for a EU Directive regarding the exposure of workers to the risks arising from noise states that the employer shall give particular attention, when carrying out the risk assessment to the level, type and duration of exposure, including any exposure to impulsive noise (Council Amending Directive Proposal 2002/xx/EEC). The European Council and Parliament have accepted the directive proposal which mean that member states have to apply at least the minimum requirements to their national legislation within three years from official publication. At present there are no practical guidances how to access impulse noise risk criteria. There is an urgent task to develop risk assessment method and risk criteria for impulsive noise to meet the requirements of the newcoming European Union noise directive.[33]


1Acoustics - Determination of occupational noise exposure and estimation of noise-induced hearing impairment. (1990) Geneva: ISO. International Organization for Standarization, ISO 1999-1990.
2American Conference of Governmental Industrial Hygienists (ACGIH). (2001) TLV's and BEI's: threshold limit values for chemical substances and physical agents, biological exposures indices. Cincinnati, OH.
3ANSI S12.9 -1999x - Part 4 Draft 2 (2000) Determination of the effects of impulse noise, ANSI, New York.
4Bruel P.V. (1975) Noise: Do we measure it correctly. Bruel&Kjaer A/B Naerum, 40 p.
5CHABA. (1968) Proposed damage risk criterion for impulse noise (Gunfire). Report of working group 57, National Academy of Sciences, National Research Council. Washington DC. Committee of Hearing, Bioacoustics and Biomechanics.
6Coles R.R.A., Carinther G.R., Hodge D.C. and Rice, C.G. (1968) Hazardous exposure to impulse noise: J. Acoust. Soc. AM. 43:336-346.
7Council Directive on the protection of workers from risks related to exposure to noise at work (86/188/EEC) Off. J. Eur. Communities 24.5.86 n L 137/28, 1986.
8Council Amending Directive Proposal (2002/xx/EEC) of the European Parliament and of the Council on the minimum health and safety requirements regarding the exposure of workers to the risks arising from physical agents (noise) (individual Directive within the meaning of Article 16(1) of Directive 89/391/EEC), Brussles, 14th Nov 2002.
9Dancer A.L., Franke R., Parmentier G. and Buck K. Hearing protector performance and NIHL in extreme Environments: Actual Performance of Hearing Protectors in Impulse Noise/Nonlinear behavior. In: Scientific basis of Noise-induced hearing loss. Axelsson A., Borchgrevink H., Hamernik R., Hellstrom P.A., Henderson D. and Salvi R., eds. Thieme, New York, 1996.
10Hamernik R.P. and Hsueh, K.D. (1991) Impulse noise: some definitions, physical acoustics and other considerations. J. Acoust. Soc. Am. 90(1):189-196.
11Health and Safety Commission. (1987) Prevention of damage to hearing from noise at work. Draft proposals for regulations and guidance no 198. London: Her Majesty's Stationery Office (HMSO).
12IEC 179A. (1973) Recommendation for Impulse Sound Level Meters. Commission of Electrotechnique International. Geneva, 69 p.
13IEC 651. (1979) Sound level meters, Commission of Electrotechnique International,.Geneva, 73 p.
14INRS 134. (1989) La protection des travailleurs contre le bruit: Institut National de Recherche et Securite. Cah notes docum 134:17-28.
15ISO 1999. (1990) Acoustics - Determination of occupational noise exposure and estimation of noise induced hearing impairment. International Organization for Standardization, Geneva, 17 p.
16ISO 9612. (1997) Acoustics-Guidelines for the measurement and assessment of exposure to noise in a working environment. International Organization for Standarization (ISO). Geneva, 25 p.
17Johnson D.L. and Riffle C. (1982) Effects of gunfire on hearing level for selected individuals of the inter- industry noise study. J. Acoust. Soc. Am. 72:1311-1314.
18Lahti T. and Starck J. (1980) Industrial impulse noise measurements, Scand. Audiol. Suppl. 12, 61-69.
19Patterson J.H. (1991) Effects of peak pressure and energy of impulses. J. Acoust Soc. Am. 90 (1), 205-208.
20Patterson J.H. and Johnson D.L. (1996) Temporary threshold shifts produced by high intensity free-field impulse noise in humans wearing hearing protection in Scientific basis of Noise-induced hearing loss. Axelsson A., Borchgrevink H., Hamernik R., Hellstrom P.A., Henderson D. and Salvi R., eds. Thieme, New York, pp 313-320.
21Pekkarinen J., Iki M., Starck J. and Pyykko I. (1993) Hearing loss risk from exposure to shooting impulses in workers exposed to occupational noise. Br. J. Audiol. 27:175-182.
22Pfander F. (1975) Das knalltrauma. Springer-Verlag, Berlin. 134 p.
23Pyykko I., Toppila E., Starck J., Juhola M. and Auramo Y. (2000) Database for a hearing conservation program. Scand. Audiol. 29:52-58.
24Salmivalli A. (1967) Acoustic trauma in regular army personnel: clinical and audiological study. Acta Otolaryngol. Suppl. (Stockh) 222:1-85.
25Starck J., Pekkarinen J. and Aatola S. (1987) Attenuation of earmuffs against low frequency noise. J. Low. Freq. Noise and Vibr. 6:167-74.
26Starck J. and Pekkarinen J. (1987) Industrial impulse noise: Crest factor as an additional parameter in exposure evaluation. Appl. Acoust. 20:263-74.
27Starck J., Pekkarinen J. and Pyykko I. (1988) Impulse noise and hand-arm vibration in relation to sensory neural hearing loss. Scand. J. Work. Environ. Health 14:265-271.
28Suvorov G, Denisov E, Antipin V, Kharitonov V, Starck J, Pyykko I, Toppila E. (2001) Effects of peak levels and number of impulses to hearing among forge hammering workers. Appl. Occup. and Environ. Hygiene 16(8):816­822.
29Toppila E., Pyykko I., Starck J., Kaksonen R. and Ishizaki H. (2000) Individual risk factors in the development of noise- induced hearing loss. NoiseandHealth 8:59-70.
30Walden B., Prosek R. and Worthington D. (1975) The prevalence of hearing loss within selected US Army branches. Washington, DC: Walter Reed Army Medical Center. Technical Report; IAO 4745.
31Ward W.D. (1968) ed. Proposed damage risk criterion for impulse noise (gunfire). Washington, D.C: Committee of Hearing. Bioacoustics and Biomechanics, National Academy of Science, National Research Council. (Report on a working group; no 57).
32Ylikoski J., Pekkarinen J. and Starck J. (1987) The efficiency of earmuffs against impulse noise from firearms. Scand. Audiol. 6:361-364.
33 Ylikoski M., Pekkarinen J., Starck J., Paakkonen R. and Ylikoski J. (1995) Physical characteristics of gunfire impulse noise and its attenuation by hearing protectors. Scand. Audiol. 24:3-11.