Predicting steroid hormone removal in a thin activated carbon layer coupled with ultrafiltration

[Display omitted] •Axial dispersion determines the breakthrough in millimetric activated carbon layer.•Prediction of required activated carbon layer thickness is possible.•High (99%) removal from a 50 ng L−1 estradiol feed is achieved with a 1 mm layer.•With a 6 mm layer, estradiol concentrations up to 200 ng L−1 decrease to 1 ng L−1. Ultrafiltration coupled with a thin polymer-based activated carbon (PBSAC) layer promises to tackle the challenge of estrogenic micropollutants in water. The proposed drinking water limit of 1 ng L−1 for 17β-estradiol (E2) is achievable with a feed concentration of 100 ng L−1 using a 2 mm PBSAC layer. Since actual hormone concentrations in surface water and wastewater effluents range from a few to 200 ng L−1, the prediction of layer thickness for different feed concentrations would support process design. In this work, a breakthrough curve model was formulated for a millimetric adsorbent layer, which includes the transport of E2 in the layer (convection, axial dispersion, and film mass transfer) and adsorption by the PBSAC. This model was validated on the prior experimental breakthroughs with different PBSAC sizes. Axial dispersion was found to be an important transport mechanism in a millimetric layer and its coefficient must be determined empirically, whereas traditional correlations for packed beds give inconsistent results. The validated model was finally applied to predict the required layer thickness for varied feed concentrations. A 1 mm layer packed with 80 µm PBSAC is effective up to a maximum feed E2 concentration of 50 ng L−1, whereas above a feed concentration of 200 ng L−1 (which is uncommon), a millimetric layer (≤6 mm) cannot achieve effective removal. The novel application of the existing breakthrough model for millimetric layers can assist the selection of the PBSAC or activated carbon configurations based on the actual feed micropollutant concentration for treatment.