Chemical profiling of industry wastewaters to identify industry sources of contaminants

Wastewater treatment plants are increasingly recognised as an important collector of diverse industrial discharges and for those chemicals that are not degraded in the treatment process, they become a potentially important source of contamination through release via the effluent and /or reuse of biosolids. For such chemicals, controlling and limiting release into the sewer is key to management, though knowledge of specific sources is unclear. This study aimed to profile industry wastewater sources of chemicals to sewer and identify characteristic industry markers. The largest industry wastewater dataset to date was generated, comprising 132 samples from 20 industrial sectors quantified for 117 analytes, including 88 emerging contaminants (pharmaceuticals, pesticides and PFAS). Concentrations spanned orders of magnitude and revealed sector-specific profiles, such as elevated acesulfame in beverage wastewaters, PFHpS only in primary metal, and outstandingly high contaminants in landfill and waste facilities. To support contaminant-driven source identification, a Random Forest method was developed and single-analyte thresholds for individual industries were determined. Model performance reached 50 % accuracy, with errors attributed to underrepresented classes and pairs with overlapping chemical signatures. These findings highlighted needs for broader industry wastewater sampling beyond this proof of concept, with increasing temporal coverage and an expanded analytical panel to strengthen source identification and support practical applications in targeted pollution control.