Technical Report NTB 94-08

Kristallin-I Estimates of Solubility Limits for Safety Relevant Radionuclides

The safety concept for the Swiss high level radioactive waste repository is based on a multiple barrier system. Within the concept of the safety analysis KRISTALLIN-I, the waste glass starts corroding after failure of the massive steel canister and nuclides are released to the bentonite backfill. This release is limited by restricted solubility of (secondary) solid phases ("solubility limits").

The present work quantifies the maximum expected concentrations of the elements Th, Pa, U, Np, Pu, Am, Cm, Tc, Ni, Pd, Se, Ra, Zr, Nb, Sn, Pb, Sb, Bi and Sm within the reference bentonite porewater with pH = 9, Eh = -400 mV and I = 0.08 M at 50°C. In a first step, maximum expected concentrations were calculated with a geochemical speciation code (MINEQL) based on a documented thermodynamic database. In a second step, the values obtained in this way were carefully reviewed and modified, based on extended geochemical considerations and system-dependent parameters. Thereby, the relevance of potentially limiting solids, chemical analogies, absolute and relative inventories and recent experimental findings from laboratory and natural systems were particularly considered. The expected groundwater composition in the crystalline host rock (modified by the barrier material bentonite) covers a rather narrow pH range from 8.5 to 9. Within this narrow pH range, solubility limits may be termed as pH independent since computable pH effects are never significant compared to the general uncertainty of the solubility limits.

The chemical model defining the reference groundwater predicts a system-wide Eh ranging from -400 mV up to +100 mV. A slightly oxidising near-field will stabilise the generally more soluble higher oxidation states of redox sensitive radionuclides. Based on the available thermodynamic data the elements U, Tc, Se and Pa are predicted not to be solubility limited at +75 mV. Therefore, a more detailed investigation of the redox behaviour of critical elements and, much more importantly, a very careful review of the near-field redox model is strongly recommended.

The reference solutions defined for the near- and far-field slightly differ in pH, Eh, ionic strength and carbonate content. The near-field solubility limits were therefore checked for consistency with the far-field reference water. In particular, a possible precipitation of secondary minerals (including the formation of rapidly transportable colloidal matter) was tested by means of speciation calculations. Selenium exhibits a solubility minimum in the Eh range of the far-field reference water (-180 mV), therefore the precipitation of secondary Se minerals cannot be ruled out. For americium and plutonium, the geochemical calculation predicts a moderate oversaturation but, in these cases, a clear judgement would require the review of corresponding carbonate complex data. All other predicted solubility limits are compatible with the far-field water.

The results of the geochemical speciation calculations clearly indicated inconsistencies and shortcomings in the available thermodynamic data of the actinides. The predicted behaviour is not as consistent as expected from the known chemical similarities among the actinides. Based on recently published data, a proposal for a consistent thermodynamic data set for the tetravalent actinides is made. However, a chemically consistent prediction of an overall actinide behaviour must additionally include consistent data sets for the tri-, penta- and hexavalent elements, as well as the corresponding redox equilibria.