Modeling Np and Pu Transport with a Surface Complexation Model and Spatially Variant Sorption Capacities: Implications for Reactive Transport Modeling and Performance Assessments of Nuclear Waste Disposal Sites

One-dimensional (1D) geochemical transport modeling is used to demonstrate the effects of speciation and sorption reactions on the ground-water transport of Np and Pu, two redox-sensitive elements. Earlier 1D simulations (Reardon, 1981) considered the kinetically-limited dissolution of calcite and its effect on ion-exchange reactions (involving ⁹⁰Sr, Ca, Na, Mg and K), and documented the spatial variation of a ⁹⁰Sr partition coefficient under both transient and steady-state chemical conditions. In contrast, the simulations presented here assume local equilibrium for all reactions, and consider sorption on constant potential, rather than constant charge, surfaces. Reardon's (1981) findings documenting the spatial and temporal variability of ⁹⁰Sr partitioning are reexamined and found partially caused by his assumption of a kinetically-limited reaction. In the present simulations, sorption is assumed the only retardation process controlling Pu and Np transport, and is modeled using a diffuse-double-layer-surface-complexation model. Transport simulations consider the inflow of Np- and Pu-contaminated waters into an initially uncontaminated environment, followed by the cleanup of the resultant contamination with uncontaminated water. Simulations are conducted using different spatial distributions of sorption capacities (with the same total potential sorption capacity, i.e. the same total number of sorption sites, but with different variances and spatial correlation structures). A case with a spatially uniform distribution of sorption capacities was also simulated. Results obtained differ markedly from those that would be obtained in transport simulations using constant K_d, Langmuir, or Freundlich sorption models. When possible, simulation results (breakthrough curves) are fitted to a constant K_d advection-dispersion transport model and compared to each other. Functional differences are often great enough that they prevent a meaningful fit of the simulation results with a constant K_d (or even a Langmuir or Freundlich) model, even in the case of Np, a weakly sorbed radionuclide under the simulation conditions. Functional behaviors that cannot be fitted include concentration trend reversals and radionuclide desorption spikes. Other simulation results can be successfully fitted but the fitting parameters (K_d and dispersivity) vary significantly depending on simulation conditions (e.g. infiltration vs. cleanup conditions). Notably, an increase in the variance of the specified sorption capacities results in a marked increase in the dispersion of the radionuclides, and a decrease in the fitted K_d. These results have implications for the simulation of radionuclide migration in performance assessments of nuclear waste disposal sites, for the future monitoring of those sites, and more generally for modeling contaminant transport in ground-water environments.