Improving the assessment of the odour annoyance potential

Industrial and livestock activities are regarded as vital to everyday life, but both can produce bad smells. These offensive odours can lead to nuisances and potential adverse health effects on the population living nearby the emission sources. In the field of environmental health, odour is therefore a crucial pollutant of worldwide significance. Accordingly, the duty of protecting residents from odour-related air pollution is an essential responsibility for governments. A suitable approach to help regulatory authorities in this task is to calculate the necessary separation distance (in meters) between residential areas and sources of odour. This approach is flexible; it can be used to enforce a minimum distance for new sources or check compliance for existing ones. The calculation of the separation distance depends on different kinds of information, including the amount of odour released in the atmosphere, the meteorological conditions such as wind direction and speed, and how humans smell a particular odour. All these factors can be combined into a mathematical model that ultimately shows the no-go area. However, there are significant challenges and limitations in current ways of dealing with such factors. Using state-of-art mathematical models to simulate the dispersion of odorous gases in many locations across Austria, this research project aims to improve the assessment of the odour annoyance for a reliable and realistic determination of the separation distance. This project has three goals, as follows: The first goal explores the necessity to take into consideration the variation over time of the odour amount released in the atmosphere, contrary to the current utilization of a fixed emission value; The second goal deals with the challenge of (i) adapting the meteorological data in a way that the real odour dispersion will be better described, and (ii) discovering the smallest amount of meteorology data that is representative to assess the potential for odour impacts, in contrast to the usual practice of using weather measurements for a period of one year; The third goal investigates a more appropriate method to make a bridge between the odour exposure estimated by mathematical models to the perceived annoyance by humans. Community surveys will validate the results of this method. The expected findings are relevant in providing more accurate data for epidemiological studies, developing robust strategies of emission control, and increasing the cost- effectiveness of odour assessments. And the findings have another major implication. They can be used by policymakers internationally to develop and optimize environmental odour guidelines and regulations on a sound scientific basis.

Dispersion modelling is the most widely used method for assessing odour impacts of industrial and livestock activities. In several countries worldwide, model-predicted time series of ambient odour concentrations are evaluated against impact criteria to assess the annoyance potential of emission sources. However, the entire impact assessment chain has shortcomings. The overarching goal of this project was to improve the assessment of modelled odour impacts in terms of annoyance. Several components of the modelling chain were investigated, from the main model inputs to considering features related to the human perception of odours. We use different dispersion models in the investigations (Gaussian and Lagrangian). Efforts were devoted to identifying determinants of odour impacts predictions and determining the effect of different factors on assessment outcomes. We confirm that the wind distribution of a specific site is one of the main elements driving the shape of impact maps. These maps were also shown to be sensitive to the exit temperature of point sources. We pinpoint that impact predictions are more similar across models for calculations based on hourly-averaged odour concentrations than when sub-hourly peak concentrations are considered in the assessment. On top of this, we examine the performance of three schemes for calculating sub-hourly peak concentrations. We unveil the advantages and disadvantages of these schemes by evaluating them with observations from a field dispersion experiment. The scheme known as concentration-variance computation performs best, with a tendency to overestimate the results. In addition, we show that simple analytical equations, based on a power function, can be incorporated as screening tools in tiered assessment frameworks. We find that analytical equations can potentially determine separation distances representative of modelling, in particular in the direction of prevailing winds. However, the conditions for which they were developed have to be rigorously observed. In addition, our results show that more faithful representations of time-varying odour emission rates are needed to ameliorate exposure calculations. We establish that the common assumption of constant emissions over time introduces a risk of biased impact estimates. In the largest investigation to date on this issue, we reveal that releases affected by nearby obstacles such as buildings are less sensitive to the temporal variability of source emissions. Also, we consistently disclose that the importance (or lack thereof) of many model inputs depends strongly on the prescribed components of impact criteria, especially the exceedance probability or percentile. In this respect, we highlight the need for improved harmonisation by finding that impact criteria from different jurisdictions can return disparate impact estimates even for equivalent levels of protection. Results of this project can help to achieve more consistent assessment conclusions. They have the potential to influence policy development and practice, and future research in the field of odour pollution.