Influence of frozen conditions on the oxygen diffusion coefficient in unsaturated porous materials

The efficiency of soil covers used as oxygen barriers to control the generation of acid drainage from sulfidic mine wastes can be evaluated in terms of the diffusive oxygen flux reaching the underlying wastes. Oxygen diffusion has been extensively investigated over the last few decades for unsaturated porous materials that are not frozen. However, little attention has been paid to materials that are fully or partially frozen, and thus, the diffusion of oxygen through soil covers during the winter freezing period has been generally neglected. This paper presents a laboratory method developed to evaluate the effective diffusion coefficient of oxygen (D_e) in frozen, inert materials. The method is a modified version of the conventional double-chamber cell in which the temperature and unfrozen volumetric water content of the sample are measured in addition to the more commonly monitored change in oxygen concentration. Several tests were conducted on non-reactive materials: that is, a sand at multiple degrees of saturation (S_r?=?20, 30, 39, and 42%), a silt (S_r?=?47%), and a mixture of the two (S_r?=?90%). Experimental data were interpreted using the POLLUTE code. Values of D_e for frozen materials were slightly lower than those obtained at ambient laboratory temperatures. In addition to the development of an empirical method for determining D_e, a preliminary model based on the model proposed by Aachib et al. (Water Air Soil Pollut 156:163–193, 2004) was created for the prediction of D_e in frozen materials by defining the involved parameters as temperature-dependent. The results indicate that predicated values of D_e are slightly higher than experimental values, suggesting that there remains room for improvement in the model.