Experimental Characterisation of the Low Frequency Noise Annoyance Arising from Industrial Plants Alessandro Cocchi, Patrizio Fausti and Stefano Piva Instituto di Fisica Tecnica, Facolta' di Ingegneria, viale Risorgimento. 40100 Bologna. Italy (Received July 10th 1992) ABSTRACT Complaints about low frequency noise annoyance are frequently reponed although the sound levels. measuredaccording to usual procedures. do not seem tojustify such reactions. In the presentpaper some experimentaldata. gatheredby the authors, are illustrated. The results areanalyzedon the basisoftheavailablemethodologies withtheaim ofgivinga useful contribution to the discussion about the necessity of a special methodologyfor the correct evaluation of the lowfrequency noise annoyance. 1. Introduction Complaints about noise disturbance are frequently reported in working places or near discotheques and traffic ways,although the sound level, measured according to usual procedures, does not seem to justify these reactions. In some cases, the annoyance reported in these situations may be attributed to the prevalence oflow frequency com- ponents in the noise spectrum, included in the range 20-200Hz. Many sources of high level low frequency noise have been identified, like burners and boilers for heating plants, air conditioning and ventilation systems, fans and pumps, diesel engines and amplified music. A very useful characterisation of typical low frequency sources has been presented by Leventhal!'. Of particular interest is the examination of the annoyance situations that arise in the vicinity of industrial plants. In a survey of low frequency complaints (limited to the range up to 100Hz) carried out in the United Kingdom, Tempesr puts in evidence that 35% ofthe annoyance arises from factories, while only 13% is from discotheques or more generaUy from music and II % from vehi- cle noise. One of the most peculiar characteristics of low frequency noise is that high distur- bance is reported even when the sound level in dB(A), orin another of the many noise criteria and reference scales developed in recent years, is relatively low. All these noise evaluation techniques are intended to giveboth annoyance and loudness on the basis of only one parameter. But, for example, as has been explained by Persson et al. 3 , it is known that under 200 Hz the corresondence between annoyance and loudness, measured in dB(A), stops. The experience of the authors is similar. In the past the solution to this problem has been looked for in the definition of new parameters or in the modification of those already available, like PNdB proposed by Broner and LeventhaU 4 • However at the present time, a definitive methodology of analysis has not been developed. Recently attention has been turned to the evaluation of the characteristics of acoustic insulation oflow frequency noise. as proposed by Schomeret al.S. In fact this behaviour can have great significance. keeping in mind that one of the characteristics of low frequency noise is its facility of propagation, whether through solid structures or through air. Furthermore. in such cases. there is an increasing control of medium and high frequency noise due to the easier energy attenuation in those frequency ranges. This reduces the dB(A) levels but increases the sensation of loudness, as Ben- ton and Leventhall note", On this basis, another proposal suggests the reduction of the admissible dB(A) levels. if the noise spectrum shows a prevalence of low frequency components. distinguishing between day, evening and night-time. as proposed by Ver- cammen' for the Dutch Norms. Journal of Low Frequency Noise and Vibration Vol. JJ No.4 1992 124