Uncertainties in the Assessment of Volatile Hydrocarbon Pollution of Urban Stormwater

Urban stormwater transports toxic pollutants into receiving waters such as rivers, posing risks to human and ecosystem health. Improving urban liveability has been a challenging issue to many cities around the world as the fundamental processes (pollutant build-up and wash-off) that drive stormwater pollution are affected by increasing intensity of anthropogenic activities and the changing climate. The processes that stormwater pollutants undergo during dry weather periods and during storm events are inherently uncertain. This is because the temporal and spatial variations in the load and composition of pollutants built-up and washed-off from urban surfaces. The accurate and quantitative information about this uncertainty is essential for informed planning and management decision making in the context of mitigating stormwater pollution. However, the tools that are typically used to replicate stormwater pollutant processes (stormwater modelling tools) have significant limitations in being able to account for the uncertainty inherent to pollutant processes. Thus, urban water management personnel often face difficulties in interpreting modelling outcomes, resulting in designing less effective pollution mitigation strategies.

This study employed statistical modelling (Bayesian Networks) to investigate how anthropogenic and climatic factors contribute to creating variability in hydrocarbons build-up, which influence the variations in wash-off. This will contribute to understanding how the build-up process uncertainty could influence decision making in pollution mitigation. The hydrocarbons investigated included seven volatile compounds (benzene, toluene, ethylbenzene, p-xylene, m-xylene, o-xylene and styrene), 21 semi-volatile compounds (C8 – C28) and 12 non-volatile compounds (C29 – C40). These pollutants are primarily generated by traffic. It was found that hydrocarbons exhibit different patterns of build-up over the antecedent dry period during different seasons. These different behaviours could be attributed to season-specific conditions and physico-chemical characteristics of particulates. Compared to anthropogenic factors, most hydrocarbons were found to have stronger relationships with climatic factors. The research study also revealed that anthropogenic factors are more related to generating hydrocarbons and their prevalence can be largely determined by climatic factors. These new insights are expected to enhance the understanding of the mechanisms related to hydrocarbon build-up process, and thereby to contribute to the formulation of accurate process replication models that would enable the quantification of uncertainties. This will not only help in the design of effective stormwater pollution mitigation strategies, but also strengthen risk assessment in relation to stormwater reuse.