Fluorescence spectroscopy and parallel factor analysis as a dissolved organic monitoring tool to assess treatment performance in drinking water trains

Fluorescence excitation emission matrix (FEEM) spectroscopy was used to evaluate its applicability as a tool to track dissolved organic matter (DOM) in a drinking water treatment plant (DWTP) that incorporates a conventional line (consisting in ozonation and GAC filtration) and a membrane-based line (consisting in ultrafiltration, reverse osmosis and mineralization) working in parallel. Seven sampling points within the different process stages were characterized monthly during 2014. A global Parallel Factor Analysis (PARAFAC) was used to pull out underlying organic fractions from the fluorescence spectra. Accordingly a five components model was selected to describe the system and the pros and cons of the model were discussed by analysis of the residuals. Among the five fluorescent components, those associated to humic-like matter (C1, C3 and C4) showed a similar season variability in the river water feeding the DWTP (which resembled that of UV254 and TOC), whereas the two components associated to protein-like matter (C2 and C5) exhibited a different behavior. The maximum fluorescence intensity values (Fmax) were used to quantify DOM removals across the plant. Compared to the conventional line, water from the UF/RO membrane-based line showed between 6 and 14 times lower fluorescence intensity signal for the humic-like components and between 1 and 3 for the protein-like components as compared to the conventional line. The differences in DOM composition due to seasonal variations and along the treatment trains point out the suitability of using fluorescence measurements over other parameters such as UV254 as a monitoring tool to help optimize operation conditions of each treatment stage and improve produced water quality in a DWTP.Postprint (author's final draft