Decoupling wastewater impacts from hydrogeochemical trends in impacted groundwater resources

Abstract In the urban environment, anthropogenic activities provide numerous potential sources of contamination, which can often lead to difficulties in identifying the processes impacting groundwater quality (natural and anthropogenic). This is particularly relevant at Wastewater Treatment Plants (WWTPs) that are often subject to changes in land use and composition of contaminant sources over time and space, as well as multiple potential hydrogeochemical interactions. To help address this issue, we demonstrate how long-term time-series analysis of major ions and key contaminants of concern, which are routinely collected by WWTP operators, can be analysed using hydrogeochemical plotting tools, multivariate statistics and targeted isotopic analysis, to provide a means of better characterising key hydrogeochemical influences and anthropogenic inputs. Application of this approach to a WWTP in south-eastern Australia indicated that anthropogenic impacts were the primary driver influencing the local hydrogeochemical environment and groundwater quality. However, secondary processes, including mineral (particularly calcite) dissolution, ion exchange and possible dedolomitisation, as well as natural degradation/transformation of contaminants were also important. Long-term, time-series analysis of trends in NO3-N, NH4-N, Ca2+, SO42−, HCO3− and K+ in conjunction with the other lines of evidence, allowed for enhanced separation between individual contaminant sources, particularly when paired with a detailed site history and Conceptual Site Model (CSM). This indicated that off-site agricultural impacts post-date most site derived impacts, and to date, have not significantly added to the identified contaminant plume. The outcomes of this work have significant global application in the identification, assessment, and control of environmental and health risks at complex sites and show how significant value (rarely obtained) can be derived from the analysis of routine monitoring datasets, particularly when analysed using a multiple lines of evidence approach.