Chemical Contaminants in Urban Rainwater Tanks

Abstract

This project examined the chemical water quality of urban rainwater tanks, with a focus on quantifying the contribution of contaminants from urban air pollution. It also assessed the health risk for people utilising rainwater tanks as their main drinking water source. Tank water was sampled from 26 tanks at 23 locations spread across Brisbane, on a monthly basis, for a whole year. Sampling of atmospheric deposition (bulk deposition) at 16 sites was conducted concurrently. 13 locations had both tank water and bulk deposition monitored at the same site. Tank water and bulk deposition was analysed for a suite of 30 metals, 8 anions, organic carbon and inorganic carbon. The physicochemical characteristics of pH, electrical conductivity, hardness, langelier index and temperature were also monitored in tank water whilst bulk deposition was also analysed for total solids. Selected bulk deposition and tank water samples were also analysed for 122 pesticides, up to 19 polycyclic aromatic hydrocarbons and 16 phenolic compounds. The variation of water quality was examined for tanks on daily and annual time scales, at different locations and for different catchment and tank materials. The deposition of chemical contaminants was also examined for variation with location and season. Some observations on roof materials input, tank sludge, outlet height and filtering are included. To quantify the sources of contaminants, various tools for source apportionment were applied to both tank water and bulk deposition data. These included dispersion modelling using TAPM (Hurley 2005a) and multivariate receptor modelling using Positive Matrix factorisation (US EPA 2008), as well as some lead (Pb) isotope and particle size analysis. Results show that water from rainwater tanks in Brisbane was generally of good chemical water quality and in most cases presented minimal health risk if used for drinking. The harvested rainwater was soft and generally slightly acidic, with the exception of new concrete tanks which had alkaline water. The Langelier index of corrosion potential indicated that harvested rainwater was moderately corrosive. The corrosion of plumbing materials and fittings has the potential to increase contaminants such as copper (Cu), nickel (Ni) and lead (Pb) with the use of rainwater. The alkaline water from concrete tanks had low Pb concentrations, as most Pb precipitates out at pH >7. The limited sampling of organic contaminants did not identify any as a health hazard in this project. The main identified health hazard was Pb, which exceeded the Australian drinking water guideline (2004) in 15% of samples and nearly 14% of tanks. The annual volume weighed average concentration of Pb in atmospheric deposition was only 2 μg/L, and thus was not the main source of Pb in tanks with concentrations above the ADWG. The majority of Pb in atmospheric deposition originated from crustal matter, probably due to historical contamination of urban soil from leaded fuel use. The second major contribution of Pb in atmospheric deposition was from anthropogenic activity due to a mixture of motor vehicle, industrial and secondary pollutant sources. The contribution from motor vehicles was generally the dominant source of current primary emissions. The anthropogenic sources of Pb in deposition were increased in inner city and industrial areas. However, where the concentrations were >5 μg/L the majority of Pb in tank water originated from Pb in paint, or Pb flashing on the roof. Plumbing materials also contributed Pb in tank water. Roof catchment and tank materials generally contributed much more to tank water concentrations of relevant contaminants (e.g zinc (Zn)) than atmospheric deposition.