Field Evaluation of a Fluctuating Plume Model for Odours with Sniffing Teams

Summary

In the Netherlands, up till now a preliminary air quality standard for odours is established. It is based on an hourly averaged concentration level, which is calculated with a generally accepted gaussian dispersion model in the Netherlands. For odours, hourly averaged concentration levels are of little use, however. Because of the turbulent character of the dispersion process, high concentrations may occur within an hour in which the average concentration is well below the detection limit. For this reason a dispersion model was developed, based on the Högström odour dispersion model. This is a fluctuating plume model and it yields estimates of the perceived odour frequency downwind from the source. The dispersion parameters in the model were adapted to the generally accepted ones in the Netherlands.

The model was evaluated around three different source types; a high point source, a low source and an area source. Measured odour frequencies were compared with model calculations with a bigaussian model and with the modified Högström model. The relation between observed odour frequencies and calculated hourly averaged concentrations was poor, especially for the high source, and different for each source type. The relation between the observed and calculated odour frequencies was much better and the same for all sources.

The conclusion is that the modified Högström model is a better tool for estimating odour exposures than gaussian models based on hourly averages, specially for high sources. This model has been tested for relatively short travel times. From a theoretical point of view it is likely, that for large distances the applicability is less. Since most odour problems are on a local scale, this is no serious limitation in practice.