Water transport in unsaturated nonisothermal salty soil. II. Theoretical development.

A theory describing both vapor and liquid phases of moisture transfers in porous materials is developed. The two equations included extend previous work, which only considered moisture and temperature gradients, by also including solute concentration effects. The new equations include six diffusivities to be considered in describing vapor and liquid movement in soil. Three liquid diffusivities are for liquid-phase flow associated with water content, temperature, and solute concentration, and three vapor diffusivities are for vapor-phase flow. Steady-state heat and mass transfer laboratory experiments were performed by using moist salinized and solute-free closed soil columns. From observations of water content, temperature, and solute distributions, five of the diffusivities (vapor diffusivities due to temperature, water, and solute concentration and liquid diffusivities due to temperature and solute concentration) are determined. The isothermal liquid diffusivity as a function of soil water content is determined in a separate experiment. In general, the liquid diffusivities are larger than the vapor diffusivities for the experimental conditions considered. For the soil conditions investigated, the water fluxes due to the solute gradients are nearly equal to the water fluxes due to the temperature gradients within salinized soil columns, but the water fluxes resulting from gradients in soil water content are small in comparison. To test the validity of the theory, two different approaches are used. The first approach involves using water diffusivities of moist solute-free soil along with the new theory to predict steady-state water distributions within moist salinized soil for comparison with observations. The predicted values of soil water content are in good agreement with the observed values. The second approach is to predict steady-state soil water distributions within the moist salinized soil column while neglecting the solute effects on soil water movement. The predictions using this approach are markedly different from the experimental observations.