Adsorption columns are widely used in drinking water treatment to improve a number of water quality parameters. Granular activated carbon filtration (GAC) can be used to decrease the concentration of DOC, colour, taste, odour and pesticides. Columns filled with ion-exchange resins are used to remove colour, nitrate and DOC. The regeneration frequency of these types of filters depends strongly on the natural water quality. Especially the DOC-concentration and DOC-composition determines the efficiency of the processes. Because pilot-plant experiments with realistic contact-times will last for more than a year (for GAC) there is a need for a prediction of breakthrough based on shorter and less expensive laboratory experiments. The available models are not accurate enough because the exact parameters are not available. In batch experiments with natural water, with realistic (low) pesticide concentrations and the full grain size of the adsorbent the adsorption parameters are measured in an independent experiment. With the parameters obtained with these experiments an accurate prediction of the breakthrough curve is possible.
With the same parameters predictions of breakthrough curves are calculated under different process conditions. The possibilities of process optimisation can reduce the investment costs for new full-scale plants.