Effects of laboratory‐scale variability upon batch and column determinations of nonlinearly sorptive behavior in porous media

Porous media are considered to be composed of spatial distributions of particles, which individually follow Langmuir sorption isotherms (with their own distinct, randomly sampled sorption capacities and selectivity coefficients). Analyses of batch and column methods for determining the sorptive parameters of these porous media reveal that batch‐derived estimates of retardation based upon Langmuir and Freundlich isotherms systematically underestimate and overestimate that observed at low concentrations, respectively, when local equilibrium conditions are met in the columns. These results, derived by stochastically analyzing the sorptive behavior expected during batch and column experiments, follow from the failure of the homogeneous‐based Langmuir and Freundlich models to properly describe sorptive behavior in the ensembles of particles composing the media. As the variability in the underlying sorptive properties increases, the Langmuir isotherm ceases to describe well the behavior of the aggregate of (individually Langmuir) particles both under static (batch) and dynamic (column) conditions. In such composite porous media there tends to be an inherent uncertainty involved with characterizing sorptive behavior at lower concentrations.