Probabilistic analysis of fluorescence signals for monitoring dual reticulation water recycling schemes.

Improved techniques are required for the detection of inadvertent cross-connections between recycled water and potable water systems in dual reticulation schemes. The aim of this research was to assess the potential for fluorescence spectroscopy to be developed as a tool to distinguish recycled water from potable water. Weekly grab samples of recycled and potable water were obtained over 12 weeks from within an Australian dual reticulation site and analysed for fluorescence excitation-emission matrix (EEM), dissolved organic carbon (DOC), electrical conductivity (EC), and pH. Probabilistic techniques including distribution function fitting and Monte Carlo simulation were used to assess the ability to distinguish between recycled water and potable water sample pairs and the reliability of doing so. Fluorescence EEM spectroscopy was determined to be the most effective for the reliable differentiation by monitoring the protein-like fluorescence at peak T(1)--an excitation-emission wavelength pair of λ(ex/em)=300/350 nm. While EC could distinguish between recycled and potable water, it was shown to be less sensitive and less reliable than peak T(1) fluorescence.