Abstract
Granular activated carbon treatment with postchlorination (GAC/Cl(2)) and chlorination followed by chloramination (Cl(2)/NH(2)Cl) represent two options for utilities to reduce DBP formation in drinking water. To compare the total cytotoxicity of waters treated by a pilot-scale GAC treatment system with postchlorination (and in some instances with prechlorination upstream of GAC (i.e., (Cl(2))/GAC/Cl(2))) and chlorination/chloramination (Cl(2)/NH(2)Cl) at ambient and elevated Br(-) and I(-) levels and at three different GAC ages, we applied the Chinese hamster ovary (CHO) cell cytotoxicity assay to whole-water extracts in conjunction with calculations of the cytotoxicity contributed by the 33 (semi)volatile DBPs lost during extractions. At both ambient and elevated Br(-) and I(-) levels, GAC/Cl(2) and Cl(2)/NH(2)Cl achieved comparable reductions in the formation of regulated trihalomethanes (THMs) and haloacetic acids (HAAs). Nonetheless, GAC/Cl(2) always resulted in lower total cytotoxicity than Cl(2)/NH(2)Cl, even at up to 65% total organic carbon breakthrough. Prechlorination formed (semi)volatile DBPs that were removed by the GAC, yet there was no substantial difference in total cytotoxicity between Cl(2)/GAC/Cl(2) and GAC/Cl(2). The poorly characterized fraction of DBPs captured by the bioassay dominated the total cytotoxicity when the source water contained ambient levels of Br(-) and I(-). When the water was spiked with Br(-) and I(-), the known, unregulated (semi)volatile DBPs and the uncharacterized fraction of DBPs were comparable contributors to total cytotoxicity; the contributions of regulated THMs and HAAs were comparatively minor.