Variations in inorganic & organic composition of roof-harvested rainwater: studies at the regional & individual site level in Eastern & Southern Australia

Higher Doctorate - Doctor of Philosophy (PhD)Research in the field of rainwater harvesting and harvested rainwater quality has increased substantially over the past decade, as countries explore alternative water sources to supplement mains water supplies. Despite the increased interest in rainwater harvesting, there remain several gaps in the literature with regard to both external sources of contamination of rainwater tanks (e.g. atmospheric deposition), and components of the rainwater harvesting system (RWHS) itself. Previous studies have focussed either on relatively few study sites, low sample numbers, or have investigated only a handful of organic or inorganic contaminants. This research project comprised a relatively large number (41) of RWHSs located in eastern and southern Australia. Samples were collected from the available sampling points of RWHSs, including: a downpipe for the collection of roof runoff, a tap located close to the base of the rainwater tank (‘tank tap’), an indoor cold tap supplied by the RWHS, and an indoor hot tap supplied by the RWHS. Samples were then analysed for a range of organic and inorganic contaminants. A method was developed that allowed harvested rainwater samples to be screened for a large number of organic compounds from different classes, using gas chromatography – mass spectrometry (GC-MS). These compounds included fatty acids, sterols, phthalate esters, polycyclic aromatic hydrocarbons (PAHs) and aliphatic hydrocarbons. Inorganic contaminants of interest included a suite of 26 metals and non-metals, analysed by high resolution inductively coupled plasma – mass spectrometry (HR ICP-MS). A number of statistical comparisons were undertaken to determine whether organic and inorganic composition of roof-harvested rainwater differed significantly at the regional and individual site level. Organic and inorganic composition of roof-harvested rainwater was found to be highly variable, resulting in large standard error values. Roof-harvested rainwater samples contained organic compounds consistent with or indicative of biological debris and decay, faecal contamination from birds and small mammals, pollutants from atmospheric deposition and compounds leached from catchment surfaces. High variances meant that there were no significant differences in organic composition of roof-harvested rainwater at the regional level. Variations did, however, occur between urban, rural and industrial locations, and between RWHSs located in the same regions. This was assumed to be a result of differences in the micro-climates in which individual RWHSs were located, in addition to differences in individual system components (e.g. roofing / tank materials). Roof-harvested rainwater collected from sites located in industrial areas contained significantly higher levels of barium, copper and lead than samples collected from urban and rural areas (P<0.05). When inorganic composition of roof-harvested rainwater was compared between sampling points of all 41 RWHSs included in the large-scale study, only cobalt was found to differ significantly, with hot tap samples containing significantly higher levels of this element than tank tap and cold tap samples (P<0.05). However, a number of case studies were also carried out, in order to characterise potential variations in composition between sampling points at the individual site level. These case studies indicated that inorganic composition could indeed differ consistently and significantly between sampling points. Levels of zinc and manganese were elevated in roof runoff samples compared with samples from other available sampling points, from 1 case study site (Tank 7). This may have been the result of atmospheric deposition and/or leaching from roofing materials. In-tank processes such as sedimentation, flocculation, dilution and actions of biofilm bacteria may then act to reduce levels of these elements in tank tap samples and other outlets downstream of the storage tank. Many elements were lower in concentration in surface runoff samples compared with various sampling points for the stored rainwater at another case study site (Site E), suggesting that atmospheric deposition may not always represent the major contribution to the total contaminant load of stored rainwater. Copper piping was shown to contribute to levels of copper, and old PVC piping was shown to contribute to levels of lead, nickel and zinc in roof-harvested rainwater. Tank construction materials did not appear to significantly affect elemental composition of roof-harvested rainwater, when Aquaplate™, concrete and plastic tanks were compared. Levels of copper, zinc, lead and nickel were also shown to differ significantly between the tank tap outlets of 4 plastic RWHSs located on the same site and fed by the same roof catchment surface (P