AWWA WQTC58964

Surface water supplies the majority of drinking water in the Northeast. Experience with Riverbank Filtration (RBF) in Europe and more recently in the United States has demonstrated significant improvements in raw water quality, including removal of natural organic matter (NOM), biodegradable compounds, pesticides, microbes, and other water quality contaminants and compensation for shock loads of chemical contaminants (Kuehn & Mueller, 2000; Ray et al., 2002a; Ray et al., 2002b; Tufenkji et al., 2002; Weiss et al., 2003a; Weiss et al., 2003b). Because of these potential improvements, regulators and utilities in the United States have recently looked more strongly at RBF as a means for providing high-quality sources for drinking water. However, little data are available to compare the performance of RBF with that of conventional drinking water treatment processes more commonly used in the United States (e.g. coagulation, flocculation, sedimentation) from identical river water sources, especially with regard to the removal of organic disinfection byproduct (DBP) precursor material. In addition, little is known about the extent to which RBF may serve to reliably remove Giardia, Cryptosporidium, and other pathogens (e.g., bacteria, viruses) from river water. In particular, data are needed on the transport of microbial pathogens through riverbank systems relative to that of more easily measured indicator parameters, such as particles and coliform bacteria. In the above context, research was conducted to document the water quality benefits during RBF at three major river sources in the mid-western United States (the Ohio River at Jeffersonville, Indiana; the Wabash River at Terre Haute, Indiana; and the Missouri River at Parkville, Missouri), specifically with regard to reduction in DBP precursor organic matter and microbial pathogens. Specific objectives were to: evaluate the merits of RBF for removing organic DBP precursor material; evaluate whether RBF can improve finished drinking water quality by removing and/or altering NOM in a manner that is not otherwise accomplished through conventional processes of drinking water treatment (e.g. coagulation, flocculation, sedimentation); evaluate changes in the character of NOM upon ground passage from the river to the wells; evaluate the merits of RBF for removing pathogenic microorganisms; and, compare the transport of microbial pathogens with that of some potential surrogate or indicator parameters (e.g. particles, turbidity, coliforms, aerobic and anaerobic spores, diatoms, bacteriophage). To address objectives 1, 2, and 3, samples of the river source waters and the bank-filtered well waters from the three study sites were analyzed for a range of water quality parameters including TOC, DOC, UV-absorbance at 254-nm (UV-254), biodegradable dissolved organic carbon (BDOC), biologically assimilable organic carbon (AOC), inorganic species, and DBP formation potential. In the second year of the project, river waters were subjected to a benchscale conventional treatment train consisting of coagulation, flocculation, sedimentation, glass-fiber filtration, and ozonation. The treated river waters were compared with the bank-filtered waters in terms of TOC, DOC, UV-254, and DBP formation potential. In the third and fourth years of the project, NOM from the river and well waters was characterized using the XAD-8 resin adsorption fractionation method (Leenheer, 1981; Thurman & Malcolm, 1981). XAD-8 adsorbing (hydrophobic) and non-adsorbing (hydrophilic) fractions of the river and well waters were compared with respect to DOC, UV-254, and DBP formation potential to determine whether RBF alters the character of the source water NOM upon ground passage and if so, which fractions are preferentially removed. The ongoing research to address objectives 4 and 5 consists of: field studies at the three study sites to document actual changes in microorganism concentrations upon subsurface travel between the r