Logo of the Physikalisch-Technische Bundesanstalt

Uncertainty of predictions in building acoustics


A new approach for calculating the measurement uncertainty made it possible for the first time to realistically evaluate the uncertainties of building acoustic predictions according to EN 12354-1 from the uncertainty contributions of the acoustic input quantities. It is expected that this leads to a significantly improved acceptance of such demonstrations being important for the noise protection in buildings.

Predictions in building acoustics establish the relation between the building acoustic performances of buildings and the performance of separate elements such as walls. The existing legal or other requirements for the noise protection within buildings require a demonstration in the planning process. Thus, the prediction and the proper consideration of uncertainties are of high importance. To eliminate the deficiencies of the existing practice, PTB has initiated a project with the aim to demonstrate the possibility of taking uncertainties into account. This project was funded by the Deutsches Institut für Bautechnik (DIBt).
The investigations were carried out using the example of the airborne sound insulation between two neighbouring rooms. The starting point was the fact that the sound insulation results from a total of 31 acoustic quantities in this case. An individual uncertainty is associated with each of these quantities. Then the uncertainty of the predicted sound insulation is yielded from the uncorrelated superposition of the single uncertainties weighted by the appropriate sensitivity coefficients. The whole calculation scheme was implemented into an Excel spreadsheet which can be obtained from the homepage of the Working Group "Building Acoustics". The results of the spreadsheet turned out to be consistent with other independent prediction results. Furthermore, 24 real building situations have been considered. It could be shown, that the deviations between measurement and prediction results can be explained essentially by the uncertainties (Figure 1). This has created a high level of transparency in the prediction and has thus considerably increased the acceptance by the users in practice.

Comparison between measurement und prediction results and corresponding 95%-tolerance ranges

Figure 1: Comparison between measurement und prediction results and corresponding 95%-tolerance ranges

Contact person:

Volker Wittstock, FB 1.7, WG 1.71, E-Mail: Volker.Wittstock@ptb.de