Trace organic contaminant removal from drinking waters by granular activated carbon: Adsorption, desorption, and the effect of background organic matter

Granular activated carbon (GAC) adsorption can be an effective barrier against a wide array of trace organic contaminants; however, assessing GAC performance in the presence of background dissolved organic matter (DOM) is difficult. The objectives of this thesis are to investigate overall performance of GAC adsorption of trace organic contaminants in the presence of background organic matter and to make the rapid small-scale column test (RSSCT) a more useful bench-scale tool for prediction full-scale performance of GAC adsorption of trace organic contaminants. The role of GAC particle size on the reduction of GAC adsorption capacity for trace organic contaminants by DOM was examined and applied to performance scale-up. Equilibrium adsorption capacity tests on GAC preloaded with DOM and RSSCT breakthrough curves at three different GAC particle sizes indicated that GAC adsorption capacity was dependent on GAC particle size when DOM was present. Thus, the RSSCT cannot be expected to match full-scale results unless a scaling factor is applied to the results. Kinetic tests indicated intraparticle diffusivity of five probe compounds at environmentally relevant concentrations was dependent on GAC particle size, i.e., nonconstant. A method to convert breakthrough curves from an RSSCT with a known diffusivity factor to another value of diffusivity factor was developed and validated. A series of RSSCTs were performed to assess the role of DOM in GAC adsorption of trace organic contaminants. More favorable GAC use rates were found with shorter empty bed contact times (EBCTs), because longer beds are subject to more fouling by the DOM. Adsorbers operated for taste and odor or disinfection byproduct control were shown to result in good removal of the 18 trace organic contaminants tested. The role of concentration gradient driven desorption on GAC adsorber performance was investigated. Experiments developed to simulate backwashing did not lead to an impact on the trace organic contaminants breakthrough curves. Desorption after a pulse loading was found to be less severe than expected because of hindered back diffusion.