Use of a Monte Carlo technique to complete a fragmented set of H2S emission rates from a wastewater treatment plant.

The impact of ambient concentrations in the vicinity of a plant can only be assessed if the emission rate is known. In this study, based on measurements of ambient H2S concentrations and meteorological parameters, the a priori unknown emission rates of a tannery wastewater treatment plant are calculated by an inverse dispersion technique. The calculations are determined using the Gaussian Austrian regulatory dispersion model. Following this method, emission data can be obtained, though only for a measurement station that is positioned such that the wind direction at the measurement station is leeward of the plant. Using the inverse transform sampling, which is a Monte Carlo technique, the dataset can also be completed for those wind directions for which no ambient concentration measurements are available. For the model validation, the measured ambient concentrations are compared with the calculated ambient concentrations obtained from the synthetic emission data of the Monte Carlo model. The cumulative frequency distribution of this new dataset agrees well with the empirical data. This inverse transform sampling method is thus a useful supplement for calculating emission rates using the inverse dispersion technique.

Comparison of two peak-to-mean approaches for use in odour dispersion models.

In this paper, two approaches to estimate odour concentrations in dispersion models are compared. The approaches differ in the estimation of the momentary (peak) odour concentration for the time interval of a single human breath (approximately 5 s). The Austrian Odour Dispersion Model (AODM) is a Gaussian model with peak-to-mean factors depending on wind speed and atmospheric stability. The German Lagrange code AUSTAL2000 uses a constant factor 4 in all meteorological conditions to derive the maximum odour concentration over a short integration time. As the Lagrange model, in contrast to the Gauss model, can be applied also in complex topography and with isolated buildings, the implementation of the Austrian peak-to-mean approach in AUSTAL2000 would enable for more realistic separation distances in these environments. In a current scientific project, this implementation will be carried out, and a comparison of separation distances with AODM and AUSTAL2000 will be undertaken.