Impact of the Variability in Hydraulic Parameters of the Vadose Zone on Pollutant Plume Development

Spatial variations in textural properties of surficial formations at field scale result in heterogeneities in hydraulic parameter fields that govern vadose zone flow. The analysis of a global soil database has shown that these spatial variations are site-dependent, the variances of the main hydraulic parameters varying in about one order of magnitude depending on the area. This study aims at quantifying the influence of spatial variability in hydraulic parameters on pollutant plume resulting from a localized source of pollution (tritium) at ground level.Series of random fields representing three of the Mualem-van Genuchten parameters, namely saturated hydraulic conductivity (Ks), inverse of air-entry value (α) and pore-size distribution (n), were generated with various variabilities and scale properties and set as input to an unsaturated flow and transport model. The simulated pollutant plumes within the vadose zone were then analyzed with regards to their size, location of center of mass and shape.Simulations show that the variance of the hydraulic parameters under study variously impacts the pollutant plume features. By comparison with homogeneous parameter fields, Ks-, α- and n- random fields generated with the mean variances observed in natural environment respectively result in (i) 25, 20 and 65% increase in plume size; (ii) 0.8, 1 and 1.8 m horizontal offsets of the plume center; and (iii) 20, 30 and 50% decrease in plume circularity. In addition, changes in the variance values within one order of magnitude appear to have critical consequences only for the n parameter. Besides, when the three parameters are varying in a correlated way, the effects of their variability on the pollutant plume are adding up together.The issue of spatial variability of hydraulic parameters is thus crucial for characterizing the evolution of pollutant plumes within a vadose zone. The understanding of that variability could help to better constraint estimations of the soil volume that needs to be decontaminated, especially within the framework of remediation strategies for industrial sites.

This work is part of Kri-Terres project, funded by the French National Radioactive Waste Management Agency (Andra) through the program "Investing for the Future" sponsored by the French Ministry of Higher Education, Research and Innovation.