| Popis: |
The safe disposal of radioactive waste is one of the key aspects in the risk assessment of nuclear energy. A safe disposal requires that the long-term dispersion of the radionuclides (RN) from the waste to the environment is minimal. Disposal facilities are designed to contain the radioactive waste by engineered and natural barriers without the intention of retrieval. The assessment of RN dispersal requires a thorough understanding of the processes and mechanisms controlling the radionuclide transport around these barriers. This PhD is dedicated to a specific aspect of such assessments, i.e. the sorption of two medium-lived radionuclides, i.e. caesium-137 (137Cs) and strontium-90 (90Sr) on glauconite sands of the Paleogene and Neogene formations (Fm) in Belgium. Glauconite sands are not considered as a primary host for radioactive waste disposal in geological formations. The sands are highly permeable and facilitate high pore water flow rates. Glauconite sand is, however, one of the options for the complementary sorption sink below the surface disposal at the Dessel site in Belgium for the short-lived and low-level waste (Cat-A). For the high-level and/or long-lived radioactive waste (Cat-B & C), a geological disposal is considered. Host rocks that might qualify for long-term disposal are poorly indurated clay (Boom Clay or Ypresian clays) 200 to 600 m below the surface. Both Boom Clay and Ypresian clays are enclosed in glauconite rich sands or silts. Glauconite is an iron rich clay mineral formed in marine sediments and occurs mainly as coarse pellets, sizing up to 1 mm, and, to a smaller extent, as minerals in the clay size fraction (< 2 µm). The term glauconite refers to a mineral composition (end-member) similar in structure as illite. Glauconite sands are the dominant lithology in the Neogene succession with often high pelletal glauconite content (20- 30 %). The glauconite is characterised by a dark green well rounded pellet, resembling sand rather than clay. In this thesis the term glauconite is used for the description of the green grains. First, the sorption of 137Cs+ on glauconite was characterised. Glauconite sands were collected from several drilling cores in the Campine subsurface and from outcrops in the Hageland region, the latter representing weathered and oxidised samples. The radiocaesium interception potential (RIP) on glauconite sands was measured in batch in a background solution of 0.1 M calcium (Ca2+) and 0.5 mM potassium (K+). The log transformed 137Cs distribution coefficient (KD, L kg 1) after 30 days reaction ranged 3.4-4.3, surprisingly high for a sand formation and close to the KD of 137Cs+on the Boom Clay (3.5). Variations in the RIP among the sands can be explained by the glauconite content and the cation exchange capacity (CEC). Isolated glauconite fractions have 137Cs sorption potentials that are a factor of about 2 smaller than the reference Illite du Puy. A comparison of intact and milled pellets shows that the pellets reduce the sorption rate and sorption potential, suggesting that not all 137Cs+ binding sites inside the pellets are accessible. In strongly weathered pellets, the cracks in the grain surface increase the accessibility of the inner sorption sites which increases the rate and extent of 137Cs sorption. Sorption equilibrium on intact glauconite sands is reached after 1 month, though slow reactions lead to an 1.6-1.8 fold increase in the KD between equilibrium and 8.5 months. A desorption experiment showed that 137Cs sorption is not fully reversible. The similarity of Cs+ sorption on isolated glauconite fractions to illite suggests similar sorption sites characteristics. Caesium sorption on illite can be described by a three-site exchange model: frayed edge sites (FES) type I, FES type II sites and planar sites. The ammonium (NH4) - K selectivity on the FES was determined as a fingerprint for the highly selective FES. The selectivity ranged between 3.8-5.0 for three glauconite sands, in the range reported for illite and clay formations. This suggests that the glauconite sands have highly selective sites that bind 137Cs+ with similar properties as in illite. The Cs+ sorption isotherms for glauconite sands were determined between 10-8 and 10-4 M Cs+ in the equilibrium solution. The experimental isotherms were fitted with a three-site model. The FES type I capacities of the sands ranged from 0.04 to 0.06 % of the CEC (versus 0.25 % for illite) and the FES type II site from 1.7 to 2.2 % (versus 20 % for illite). The difference in the FES fractions between illite and glauconite is attributed to the smectite content on glauconite. Smectite interlayers increases the CEC and decreases the fractions of FES type I and type II sites. Glauconite has a RIP that is comparable to illite, suggesting that glauconite containing sands may be an effective geological barrier for radiocaesium. However, the combination of the presence of glauconite as coarse pellets (~ 0.25 mm diameter) and slow sorption indicates considerable sorption non-equilibrium may occur during reactive transport in these permeable sands. A breakthrough (BT) experiment was set up with 13 saturated columns at variable flow rates. The columns, all filled with the same Diest Fm glauconite sand, were leached with 10-6 M Cs+ for 154 days. The BT was observed in the columns with the higher flow rates after 890-1170 pore volumes. Breakthrough at smaller cumulative pore volume was only observed in the column with the highest flow rate (2.4 m d-1), indicating chemical non-equilibrium, but these flow rates are unrealistically large for the Neogene and Paleogene Fm. The BT curves were modelled with HP1 (Hydrus-PhreeqC), based on batch sorption data obtained at 90 days and assuming local reaction equilibrium. The BT curves were well predicted with this model, corroborating local equilibrium unless at the highest flow rates. Caesium sorption on highly permeable glauconite sands is sufficiently fast to delay breakthrough in realistic flow rates for the Neogene and Paleogene Fm (Diest, Berchem and Voort Fm). The adsorption of radiostrontium (85Sr2+) was determined for 45 subsurface and naturally weathered (surface) glauconite sands. The KD for 85Sr2+ in a background solution of 1 mM Ca2+ ranged 23-65 L kg- 1 for the intact sands and 50-144 L kg-1 for the glauconite fractions. Sorption equilibrium was reached within 48 hours. The KD values for intact and milled pellets are nearly identical, i.e. the sorption of 85Sr is not hindered by the glauconite pellet in contrast to observations for 137Cs. The KD values corresponded well with predictions of the KD by two existing models calibrated to soils based on the CEC. It is concluded that glauconite sands have a suitably high retention of radiostrontium and the sorption strength is in line with that of other geological barriers when judged from the CEC. An accelerate weathering study was performed to evaluate how weathering of glauconite can affect the radiocaesium (137Cs) retention given that redox or pH dependent transformation may change the Fe speciation, leaching of K and, hence, affect clay mineralogy. Three different glauconite sands were artificially weathered at ambient temperature during 27 months in four different scenarios: continuously purged with oxygen (oxidation), under cement water at pH = 13 in the absence of oxygen, in an acid solution of pH = 4 and in an anoxic setup at pH = 7 under N2 atmosphere. The CEC increased by factors 1.1-1.2 under alkaline conditions whereas it decreased by similar factors in the oxic and acid samples. The Fe2+ to Fetot ratio doubled under anoxic conditions and remained similar in the other treatments. The RIP enhanced by factor 1.1-1.3 under alkaline conditions consistent with the changes in CEC, however no decreases in RIP were detected in the oxic samples. These analysis suggests that glauconite sand is not highly sensitive to weathering under these conditions and that the effects on the radiocaesium sorption are marginal within the conditions used here. To summarise, this study shows that the glauconite sands exhibit strong sorption potential for radiocaesium and radiostrontium. The 137Cs sorption potential of a sand with 25-30% glauconite appeared surprisingly as high as that of Boom Clay with 30-60 % clay, which is one of the candidate hosts for category B and C waste. From the research work performed in this PhD, it is concluded that glauconite can act as an additional sorption sink for Cs and Sr. status: published |
