Effectiveness of Predictive Computer Programs in the Design of Noise Barriers: A Before and After Approach

About the report:

The early work of the Virginia Department of Highways and Transportation on highway traffic noise was focused on a model and a computer program for predicting the noise levels likely to be generated by traffic on proposed highways. The computer program was also used to design barriers to shield neighborhoods from such noise. Because of the cost of noise barriers and the millions of dollars worth of barriers being planned, there was a need to evaluate the efficacy of the predictive computer programs used to design the barriers. For the evaluation, noise measurements were taken before and after construction of an earth berm and plywood wall alongside I-495 in Northern Virginia. The effect of the barrier was taken as the difference between the drop-off in noise level from the roadside to the location under study, as measured before and after construction of the barrier. The measured effect, obtained in this manner, was compared to the effect predicted by hand calculator using the current FHWA model. in a straightforward comparison,, all but a few of the measured effects were less than the predicted effects and less than half of them were within 3 dB of the predicted values. However, when the effects of the time of day and human activity within the neighborhood were taken into account, 65% of the noise values obtained during periods of low human activity were within 3 dB of the predicted values. Much of the discrepancy can be attributed to local noise, not accounted for in the prediction, thus it is thought that the computer program can be effectively used in designing noise barriers. By monitoring the neighborhood activity it was determined that while noise such as that from jet lanes and lawn mowers sometimes overrode the traffic noise from I-495, the latter clearly dominated the noise environment of the neighborhood. direct correlations made between variations in L10 levels and truck traffic and between L50 and L90 levels and automobile traffic seem to support the generalization that the L10 levels are controlled by trucks and the L90 levels are controlled by automobiles.

Authors