Prediction of carbon BTEX adsorption capacity using field monitoring data

The need for predicting adsorption capacity for benzene, toluene, ethylbenzene, and xylenes (BTEX) onto granular-activated carbon (GAC) is a problem commonly associated with petroleum-spill remediation. In this study, monitoring data are compiled from operational records of ground-water pump and treat remediation sites where GAC adsorption is utilized as a primary treatment mechanism for BTEX. The monitoring data are reduced to adsorbed and equilibrium concentrations from which Freundlich isotherms and various linear and multivariate models are calibrated for prediction of BTEX capacity on GAG. The models are employed by themselves and with Ideal Adsorbed Solution Theory to predict capacity for total BTEX and benzene. Several models are selected based on prediction ability and are tested with independent data. Two simple models, a multivariate model and a Freundlich isotherm, are recommended. Complex empirical models and Ideal Adsorbed Solution Theory did not perform as well as the selected models and were rejected. From the Freundlich isotherm, new Freundlich constants are reported that describe adsorption of total BTEX on GAC from gasoline-contaminated ground water.