Interaction of fresh and saline waters with compacted bentonite

The interaction of compacted sodium bentonite with fresh groundwater simulant (Allard water) and saline groundwater was studied. The parameters varied in the experiments were water/bentonite ratio, the density of bentonite, composition of solution, temperature, and oxygen and carbon dioxide content in the surroundings. The main interest of the study were the chemical and mineralogical changes in solution and bentonite together with the microstructural properties of bentonite. The results presented in this interim report are preliminary and can change when the experiments are continued. The behaviour of individual cations presented a principal difference between the interactions with Allard water and saline water, while the differences caused by aerobic and anaerobic conditions were minor. The effect of the bentonite/water ratio was clear. The concentrations of the solutions changed more quickly when the b/w ratio was high and the obtained concentration levels, caused by dissolving accessory minerals, were higher, too. The major processes with Allard water were the diffusion of sodium, potassium, sulphate, bicarbonate and chloride from bentonite, and the diffusion of calcium and magnesium into bentonite. The density of bentonite affected the equilibrium state of the cations, which may be explained by changing cation exchange selectivity. The major processes in the experiments with saline water were the diffusion of sodium, magnesium, sulphate and bicarbonate from bentonite, and the diffusion of calcium into bentonite. The highest pH values were obtained in aerobic conditions both with saline and Allard water, intermediate values were obtained with Allard water in anaerobic conditions and the lowest values with saline anaerobic water. The copper tubes, which contained the bentonite during the experiments, also participated in the chemical reactions producing dissolved ions and precipitations in the solutions. The experimental equilibrium concentrations of the solution vs. b/w ratio and the values obtained by equilibrium model calculation were strikingly similar, although quantitative differences appeared. However, no definite conclusions regarding the validity of the bentonite model are to be made until more experimental data, especially at higher b/w ratios with special emphasis on pH, are available. The obtained surface areas by the BET-method and Hg porosimetry, about 20 m2/g, mean that each montmorillonite aggregate in bentonite would be formed on an average of 30 montmorillonite layers. In the freeze-dried sample about one third of the pores were smaller than 60 Å and not seen by the Hg porosimetry. The pore sizes seen in the SEM micrographs correspond qualitatively rather well with the results given by the Hg intrusion porosimetry.