Abstract
According to the KBS-3H concept, each copper canister containing spent nuclear fuel will be surrounded by a bentonite buffer and a perforated steel cylinder. Since steel is unstable in wet bentonite, it will corrode and the corrosion products will interact with the surrounding bentonite in ways that are not fully understood. Such interaction may seriously impair the bentonite’s functioning as a buffer material, e.g. by lowering its CEC or decreasing its swelling capacity. This report presents results from two ironbentonite experiments carried out under quite different conditions at VTT (Finland) and JAEA (Japan). Both studies focused on long-term iron-bentonite interactions under
anaerobic conditions.
The study at VTT comprised eight years long experiments focused on diffusive based interactions between solid cast-iron and compacted MX-80 bentonite (dry density 1.5-1.6 g/cm3) in contact with an aqueous 0.5 M NaCl solution. The study at JAEA comprised ten years long batch experiments, each involving a mixture of metallic iron powder (25 g), an industrially refined Na bentonite, Kunipia F, which contains more than 99% montmorillonite (25 g), and an aqueous solution (250 mL). Samples were sent to B+Tech in airtight steel vessels filled with N2 and subsequently analyzed at various laboratories in Finland and Sweden. The JAEA samples differed with regard to the initial solution chemistry, which was either distilled water, 0.3 M NaCl, 0.6 M NaCl, 0.1 M NaHCO3, or 0.05 M Na2SO4.
The analyses of the MX-80 bentonite samples were carried out on samples containing a cast iron cylinder and also on corresponding background samples with no cast iron. In addition, the external solution and gas phase in contact with the bentonite were analyzed. Briefly, the gas contained H2, most possibly caused by corrosion of the cast iron, and CO2, mainly as a result of carbonate dissolution. The eight years old external solution exhibited, inter alia, reducing conditions, a pH of around eight, and measurable amounts of Mg2+, Ca2+, and SO4 2-. The bentonite was carefully divided into subsamples, which were studied with XRD, FTIR, SEM, ICP-AES, TEM-EDS, XANES, Mössbauer spectroscopy, and wet-chemical methods. Briefly, bentonite samples containing cast iron cylinders contained higher amounts of iron than the reference samples. The corroded iron was predominantly in the divalent form, and its concentration was highest close to the cylinder and decreased strongly with increasing distance from its surface. The average corrosion rate estimated from Fe profiles in the Fe-reacted samples is about 1.7 µm/a. The results from the Mössbauer spectroscopy analyses suggest that no reduction of the octahedral Fe3+ in the montmorillonite layers had occurred. The swelling pressure and the hydraulic conductivity were measured in undisturbed subsamples of the MX-80. The iron-bentonite interaction seemed to slightly decrease the swelling pressure, while the hydraulic conductivity was unchanged.
The corrosion rate of the Cu vessel surface was estimated from the Cu analysis in the clay to be about 0.035 µm/a.
The JAEA samples were analyzed with regard to the conditions in the water and in the bentonite. The water exhibited pH values in the approximate range of 11 to 13, and clearly reducing conditions with Eh values between -260 and -580 mV. XRD and FTIR analyses of the bentonite material, showed that montmorillonite was completely transformed to a non-swelling 7 Å clay mineral, most likely to the serpentine mineral berthierine, in samples containing 0.3-0.6 M NaCl solutions, with the highest pH values. The transformation was incomplete in samples containing 0.1 M NaHCO3 solution, and did not occur at all when the solution was either 0.05 M Na2SO4 or distilled water.
This report is a result of a joint project between Posiva and SKB. The report will also be printed as SKB Report R-10-52.
anaerobic conditions.
The study at VTT comprised eight years long experiments focused on diffusive based interactions between solid cast-iron and compacted MX-80 bentonite (dry density 1.5-1.6 g/cm3) in contact with an aqueous 0.5 M NaCl solution. The study at JAEA comprised ten years long batch experiments, each involving a mixture of metallic iron powder (25 g), an industrially refined Na bentonite, Kunipia F, which contains more than 99% montmorillonite (25 g), and an aqueous solution (250 mL). Samples were sent to B+Tech in airtight steel vessels filled with N2 and subsequently analyzed at various laboratories in Finland and Sweden. The JAEA samples differed with regard to the initial solution chemistry, which was either distilled water, 0.3 M NaCl, 0.6 M NaCl, 0.1 M NaHCO3, or 0.05 M Na2SO4.
The analyses of the MX-80 bentonite samples were carried out on samples containing a cast iron cylinder and also on corresponding background samples with no cast iron. In addition, the external solution and gas phase in contact with the bentonite were analyzed. Briefly, the gas contained H2, most possibly caused by corrosion of the cast iron, and CO2, mainly as a result of carbonate dissolution. The eight years old external solution exhibited, inter alia, reducing conditions, a pH of around eight, and measurable amounts of Mg2+, Ca2+, and SO4 2-. The bentonite was carefully divided into subsamples, which were studied with XRD, FTIR, SEM, ICP-AES, TEM-EDS, XANES, Mössbauer spectroscopy, and wet-chemical methods. Briefly, bentonite samples containing cast iron cylinders contained higher amounts of iron than the reference samples. The corroded iron was predominantly in the divalent form, and its concentration was highest close to the cylinder and decreased strongly with increasing distance from its surface. The average corrosion rate estimated from Fe profiles in the Fe-reacted samples is about 1.7 µm/a. The results from the Mössbauer spectroscopy analyses suggest that no reduction of the octahedral Fe3+ in the montmorillonite layers had occurred. The swelling pressure and the hydraulic conductivity were measured in undisturbed subsamples of the MX-80. The iron-bentonite interaction seemed to slightly decrease the swelling pressure, while the hydraulic conductivity was unchanged.
The corrosion rate of the Cu vessel surface was estimated from the Cu analysis in the clay to be about 0.035 µm/a.
The JAEA samples were analyzed with regard to the conditions in the water and in the bentonite. The water exhibited pH values in the approximate range of 11 to 13, and clearly reducing conditions with Eh values between -260 and -580 mV. XRD and FTIR analyses of the bentonite material, showed that montmorillonite was completely transformed to a non-swelling 7 Å clay mineral, most likely to the serpentine mineral berthierine, in samples containing 0.3-0.6 M NaCl solutions, with the highest pH values. The transformation was incomplete in samples containing 0.1 M NaHCO3 solution, and did not occur at all when the solution was either 0.05 M Na2SO4 or distilled water.
This report is a result of a joint project between Posiva and SKB. The report will also be printed as SKB Report R-10-52.
Original language | English |
---|---|
Publisher | Posiva |
Number of pages | 92 |
Publication status | Published - 2010 |
MoE publication type | D4 Published development or research report or study |
Publication series
Series | Posiva Working Report |
---|---|
Number | 2010-71 |
Series | SKB rapport |
---|---|
Number | R-10-52 |
ISSN | 1402-3091 |
Keywords
- Rauta-bentoniittivuorovaikutus
- raudan korroosio
- kuparin korroosio