Mineralogical investigations of the alternative buffer material test (Aespoe, Sweden)

Document available in extended abstract form only. For the investigation of the performance of HLRW multi-barrier systems ideally large scale tests are conducted. Concerning the performance of bentonite as geotechnical barrier several important tests are conducted at the Aespoe hard rock laboratory managed by SKB in Sweden. One of the still unanswered questions concerns the selection of the optimum type of bentonite for this application. Accordingly SKB has setup the so-called 'Alternative Buffer Material Test' (ABM) in which eleven different clays were assembled around a heated Fe tube. Altogether 30 blocks were used so that each material at least occurred twice in the sequence (Fig. 1). Of the three identical packages (each 30 blocks) the first package was retrieved after 28 months. Samples of each bentonite type were investigated with respect to mineralogical changes. The most striking change concerned the exchangeable cation population which apparently equilibrated with the surrounding rock water. Some of the bentonites were dominated by exchangeable Na and some were dominated by Ca. After the experiment, the cation population of all materials was similar. The remaining small differences need further investigation. In contrast to the type of exchangeable cation, the CEC did not change significantly in the 1 cm to 7 cm parts of the blocks (only a few % decrease was observed, however this can also be discussed as methodological effects during CEC determination) indicating that the cation exchange expectedly was a reversible process. Irreversible changes were restricted to the very contact, i.e. sample at 0.1 cm which was gained by scratching off the surface with a sharp knife. In some instances a significant Fe accumulation was observed. The most significant could be explained by siderite precipitation. The typical corrosion products occurring at the contact of bentonite and iron (7 or 14 Angstrom phases) were not observed. Some samples showed S accumulation at the contact, e.g. anhydrite in the case of sample FEB. The comparison of XRD traces suggests that cristobalite and clinoptilolite, if present as accessory, was dissolved at the contact. One of the most relevant questions concerns the stability of the smectite. Hence possible smectite alteration processes are of pronounced interest. In the present study XRD and IR indicates the formation of tri-octahedral domains in the case of three bentonites. They were identified both by changes of the d{sub 060} reflection and by an increased intensity of the 680 cm{sup -1} vibration. This band, however, coincides with an anhydrite band which means that the accumulation of anhydrite in some cases (samples FRI + FEB) explains the increased intensity at 680 cm{sup -1}. At least for sample LOT for which no S accumulation was found the increase of tri-octahedral domains is strongly indicated by both XRD and IR data. A Mg increase at the contact was already found in other experiments. Data of the present study indicate that this Mg increase may be related to the formation of tri-octahedral domains. However, for a more detailed understanding of this process further investigations with ideally larger concentrations of the tri-octahedral domains are required. The results of the present study prove that different bentonites may differently interact with the fluid and the tube and hence point to the yet unanswered question about the optimum bentonite for HLRW disposals. (authors)