An Experimental Method for Predicting the Adsorption of Trace Organic Contaminants in Partially Saturated Granular Activated Carbon

The pore and surface diffusion model of adsorption was used to develop an equation that indicates that the fraction of a trace contaminant removed across a granular activated carbon (GAC) column during water treatment is an exponentially decaying function of the empty bed contact time (EBCT), and a parameter phi that accounts for GAC preloading. An experimental method using partially saturated GAC harvested from a plant was proposed to allow the value for phi to be determined for GAC from that site. With this information, the ability of the GAC to remove a contaminant of interest could be predicted for different EBCTs, or for different GAC service times, potentially allowing utilities to predict how much service time is left before the GAC needs to be replaced to meet some treatment target. Two conditions of the current method are that organic loading must be relatively consistent throughout the GAC lifetime and the water temperature during model calibration should be the same as that at full scale; however, the model allows for a future modification to account for these variables. The prediction capability of the model was validated by comparing the breakthrough from independent pilot-scale tests with the predicted breakthrough calculated using the calibrated model and bench-scale tests.