Technical Report NTB 97-03

Complexation of radionuclides by humic substances and the concomitant remobilisation of sorbed radionuclides is a source of uncertainty in performance assessment of radioactive waste repositories. Despite more than 30 years of world-wide intensive research in this field, only very few datasets are available which are relevant for performance assessment. The reason for this unusual situation is found in the poor understanding of chemical structures of humic substances, which is a label for a most complex, heterogeneous mixture of organic molecules. Because of the ignorance of the "real" structures of the complexes formed between radionuclides and humic substances, no binding model was developed up to now, which is firmly based on thermodynamics and which would allow extrapolations outside the chemical parameter space at which it has been calibrated.

It is therefore the main purpose of the present work to develop means for performance assessment to conservatively bound the effect of humic substances on radionuclide speciation in solution at relevant chemical conditions. The work is divided in two parts:

• Experimental work: Using the equilibrium dialysis-ligand exchange method developed at the Paul Scherrer Institute, conditional stability constants for the formation of Ca²⁺-, NpO₂⁺-, Co²⁺-, Ni²⁺-, UO₂²⁺-, and Eu³⁺- humic acid complexes and Co²⁺-, UO^₂2+-, and Eu³⁺-fulvic acid complexes were measured in a pH region between 5 and 10 and at ionic strengths between 0.02 and 0.2 M. The data are presented as equilibrium coefficients that are free from any model assumptions. The quality of the data is assessed in an extended discussion of statistical and systematical errors, and by a critical "re-review" of the auxiliary stability constants used for the calculation of the equilibrium coefficients.
• Modelling work: Relying both on the data presented here and on supporting literature data, an empirical model was developed. The model uses a minimum set of adjustable parameters. It consistently describes the effect of metal loading, pH, ionic strength, competing anions and ternary complex formation on metal-humate formation. However, the validity of the model is restricted to the chemical parameter space covered by this work and most of the studies published in the literature. The simplicity of the model has to be paid by a certain degree of inaccuracy, which, however, is minimal for the chemical conditions of the farfield of a radioactive waste repository (i.e. slightly alkaline pH, relatively high concentrations of bicarbonate, trace metal concentration). The model parameters are chosen in such a way, that the resulting predictions are conservative with respect to remobilisation of sorbed radionuclides.

The model is applied to groundwaters relevant for host rocks for storage of radioactive waste in Switzerland. It is predicted that a strong effect on lanthanides and three-valent actinides has to be expected for humic substance concentrations exceeding 10^-6 g/l (lower ppb range), whereas the effect on transition metals (Ni and Co) and hexa-valent actinides becomes evident only at humic substance concentrations exceeding 10^-3 g/l, i.e. in the lower ppm range. Speciation of alkaline earth metals and of penta-valent actinides is influenced by humic substances only at humics concentrations exceeding 0.1 g/l. The effect on tetra-valent actinides is at the moment rather speculative, since almost no experimental data are available. First estimates of the complexation of tetra-valent actinides with humic substances predict that the effect is weaker than for three-valent actinides by at least one order of magnitude.