Abstract:

The normal method of predicting community noise levels involves first computing sound power levels from near-field sound-pressure levels and then assuming that the noise radiates from one or several point sources. This approach makes several major leaps of faith. Some of these leaps of faith or issues in predicting community noise levels for power plants are identified. The issues include validity of near-field sound-pressure levels, accuracy of sound power levels computed, radiation from large sources, transmission through building walls, and screening and reflection of noise by power plant equipment and buildings. The validity of representing large sources as one or a series of point sources is reviewed. The role of atmospheric effects and the dangers inherent in typical noise prediction software is briefly reviewed. Several examples and suggestions for making more accurate predictions are given. [See NOISE-CON Proceedings for full paper.]