ARDISC

To quantify this treatment of migration as influenced by kinetics, a model has been developed in which instantaneous or local equilibrium is not assumed. The model is called the Argonne Dispersion Code (ARDISC) ( ). In the model, adsorption and desorption are treated independently and the rates for adsorption and desorption are taken into account. The model treats one dimensional flow and assumes a constant velocity of solution through a uniform homogeneous media. 

This paper describes an experimental study of the applicability of the ARDISC model to laboratory studies of nuclide migration in geologic media. 

If the rate of adsorption, the rate of desorption, and the equilibrium partitioning of a nuclide between a solid medium and solution are known, then the rate of migration of a nuclide through the medium can be predicted with the ARDISC model. 

Model Predictions. The rate for desorption of americium from the fissure surfaces into solution was assumed to equal the rate for the adsorption of americium from solution by the fissure surfaces. The sorption rate and the equilibrium fractionation of americium that were determined in the static experiments were used to determine input parameters to the ARDISC model. The ARDISC model predictions for the distributions of americium on the fissure surfaces in both sets of experiments are presented in Figures 5 through 10 along with the autoradiographs and the experimental histograms representing the various distributions of americium on the fissure surfaces. 

To relate the six curves obtained experimentally with the ARDISC model, the rate for adsorption, the rate for desorption and the equilibrium distribution of the nuclide between glauconite and solution had to be determined. 

The rate for adsorption and the rate for desorption were determined by curve fitting with ARDISC. It was assumed that the rate for desorption equaled the rate for adsorption. The a and 3 parameters were held constant at 0.958 and 0.042 respectively. 

The F, G, and zone length input parameters to the ARDISC model were varied to fit the strontium migration data at a flow rate of 2 mL per minute. 

The ARDISC model predictions for 0.67 fissure volume additions of stock solution into fissures at the three flow rates also show that the peak concentrations of americium will be sorbed at the top of each fissure. The leading edges of the nuclide are predicted to extend into the fissures. The ARDISC model also predicts that at faster flow rates, the relative amount of americium sorbed in the peak concentration at the top of each fissure should decrease. Also, the length and relative amount of the leading edge extending into each fissure is shown to increase. 

Figures 6, 8, and 10 show that the distributions of americium on the fissure surfaces after the addition of 0.67 fissure volumes of stock solution followed by the elution of 20 fissure volumes of "schist -equilibrated water through the fissures at flow rates of 1.13, 2.29, and 4.77 cm/hr. A comparison was made between the americium distributions found on the fissure surfaces after the addition of americium stock solution and that found after elution of the americium by 20 fissure volumes of "schist"-equilibrated water. It was found that after the initial loading of americium into the fissures in the first 0.67 fissure volumes of solution, the peak concentrations of americium that were sorbed at the top of the fissures decreased in their relative concentration. The leading edges of the detectable nuclide concentration extending into the fissures had increased in length and relative concentration with subsequent elution by rock equilibrated water through the fissures. The ARDISC model predicted the same relationships.

In the column infiltration experiments with strontium, the model predictions closely resemble the experimental curves for the four flow rates compared. The input parameters to the ARDISC model were derived from experimental data obtained in infiltration experiments. The model predictions were based on the assumptions that the rate for adsorption and the rate for desorption were equal and that the sorption reactions were both first order. 

The results of the experiments presented in this paper demonstrate that the migration of nuclides in geologic media can be studied experimentally and treated in at least a semiquantitative fashion using kinetic and partitioning data. The ARDISC model is a useful aid in analyzing nuclide migration data obtained in laboratory experiments. But the ARDISC model is limited to first-order sorption kinetics. 

---