Investigation of flow and solute transport through unsaturated chalk

• Main Author: van den Daele, Gerd Frans Agnes
• Published: University College London (University of London) 2005
• Subjects: 552.58
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.417027

The Chalk is a fractured rock with a fine-grained porous matrix, where the matrix provides most of the porosity and storage capacity and the fractures greatly enhance the permeability. For the unsaturated zone, the role of the fractures and the matrix in flow and solute transport is not well understood. Therefore a tracer test was carried out in the unsaturated chalk at the Fleam Dyke research site in Cambridgeshire. The experiment, under natural rainfall conditions, involved distributing deuterium and bromide on a grass-covered lysimeter (a cube of volume 125 m3) and on an adjacent 4x4 m plot. Up until 15 months after the tracer application, none of the tracer was detected in the lysimeter drainage at 5 m depth. Repeated core profiles on the adjacent plot revealed significant differences in vertical tracer distribution, indicating lateral heterogeneity of transport pathways. The results of the tracer test were modelled with MACRO 5.0, a numerical dual permeability model that was initially developed for macroporous soils. The model suggested that fracture flow is important at the site, but that it is only initiated below a depth of about 1 m. The extent of fracture flow appeared to be highly variable in different layers of the profile, mainly depending on the saturated hydraulic conductivity of the matrix. Regarding exchange of solutes between the fractures and the matrix, the model indicated that advective exchange could be important and might even outweigh diffusive exchange. The results suggested that the Chalk aquifer at the Fleam Dyke site was only moderately vulnerable to pollution, but it was concluded that even for moderate rainfall conditions some bypass flow was possible. When modelling recharge without solute transport, it was shown that a simple water balance model called DFIDGWR could perform similar or better than MACRO 5.0.