PubMedDate not supplied
Personal monitoring of noise exposure was conducted in a small dry dock, where mainly tug-boats and high sea fishing-boats are refitted. Manual tools were the prevalent noise sources. Noise exposure could be affected and amplified by factors extraneous to workers' tasks. The workers might be employed in different jobs simultaneously also in confined and small areas, and were mainly involved in non-routine tasks. Six different tasks were investigated: dry-dock supervisor, welder/ship-wrights, painters, electricians, woodworkers and unskilled workers. The number of workers employed in the tasks, with the exclusion of the supervisor, might change from day to day according to the particular work phase carried out on the boat. Daily personal exposure was measured using Larson Davis mod. 700 integrating sound level meters. The all shift time history of short equivalent continuous sound level with 10 s integrating time was collected for each selected worker. A 1/2 inch microphone was attached to the helmet at a distance of 11 +/- 1 cm from the right outer ear. Measurements were repeated over 13 days and randomly distributed over about four months. The number of data-points collected was nine for the supervisor, twenty-seven for the welders, eighteen for the painters, twelve for the electricians, six for the woodworkers and eleven for the unskilled workers. Ranges of the daily noise exposure level were from 15 to 30 dBA for all tasks, with the exclusion of the woodworkers. The time histories relative to workers grouped in the same task were matched to verify the correlation. The daily exposure levels derived from correlated samples were substituted by their arithmetic mean. The analysis of variance applied to experimental distributions of daily exposure for four of the welders indicates that the difference of means (p > 0.3) is not significant. A similar conclusion is reached for painters when only this task is considered. If we take into account the exposures of painters employed also in different tasks, the probability is included in a range of lower significance, between 0.1 and 0.05. The experimental distributions of daily noise exposure for all tasks fit the normal distribution. The Kolmogorov-Smirnov test was used at a p > 0.1 confidence level. These results may have an important role in defining sampling strategies for collecting noise exposure data both for compliance and epidemiological purposes.