Noise is one of the most widespread problems in working environments. Besides the obvious risk of hearing damage and masking of warning signals and speech, the effects on concentration, performance, behaviour and general well-being are serious consequences of annoying noise in the working environment. Added to that are other effects such as headache, stress, fatigue, etc. The methods used for hygienic assessments of noise and measures against noise sources are mainly based on sound level measurements according to the A-weighting of the sound pressure level. Several studies and experiments have shown that these methods are not appropriate for predicting the risk for noise annoyance. Studies have also shown that the noise level is important but that several other factors critically affect the annoyance. The thesis discusses the importance of different factors in the relation between noise and annoyance. Besides different level concepts, the importance of frequency characteristics, tone content, exposure time, low frequency content, level variations and the work tasks is discussed. Both existing and new assessment methods are tested. Exposure to high and low frequency noises under laboratory conditions is evaluated. According to the results of these studies, sound levels for noise above 15 kHz will be perceived as less annoying and uncomfortable than the same noise levels from noise at frequencies around 1 kHz. This means that the A-weighting may result in an overestimation of the annoyance and discomfort when evaluating high frequency sounds. The frequency character as a critical factor involved in the evaluation of noise annoyance is evaluated for ventilation noise of differing frequency characters. According to the results of this study, differences in annoyance responses will appear in a rather restricted frequency range due to differences in the frequency character of the sounds despite equal sound levels. The results also show that a 5 dB lowering of the low frequency ventilation noise level is enough to obtain a significant reduction of the annoyance responses. Possible effects of exposure time on noise annoyance during work are examined in the study including control rooms, offices and laboratories. The results reveal no changes over the working day in the control room and an indication of lower annoyance responses in the office and laboratory environments. Attempts to calibrate the individual ratings of annoyance are made. The relation between the sound level and annoyance is considerably strengthened when the individual ratings are calibrated. However no striking differences are obtained between the different tested methods for calibration. No clearly defined differences in the correlations between sound level and annoyance are found when comparing calibration based on recorded sounds and calibration based on imagined sounds. An evaluation of the use of different weighted noise levels for predicting the annoyance from a low frequency noise exposure is made. None of the tested methods, A, B, C, D weightings, correlates better to annoyance than any of the other. Neither do the use of Zwicker's loudness method, the LFN-R method or the dB(C) - dB(A) difference. Deeper analyses of the weighting methods are carried out. The correlations between the different weighting noise levels are very high. dB(A) as a method for assessing noise annoyance was not particularly misleading in comparisons with the dB(B), dB(C) and dB(D) values. The correlations to annoyance, however, are low. An evaluation of the use of the dB(C) - dB(A) difference for prediction of noise annoyance is also included. The use of the dB(C) - dB(A) difference makes a significant but very small contribution to the explanation of the differences in annoyance ratings. Analyses of the effect of level variations as a critical noise parameter involved in the annoyance response are carried out. The tested quantifications of the level variations significantly contribute to the explained annoyance variance.
Luleå: Luleå tekniska universitet, 1997. , 41 p.