Analytical modelling of fringe and core biodegradation in groundwater plumes.

Biodegradation can be divided into two categories depending on the location at which it occurs within the plume: degradation at the plume fringes, and degradation in the interior (core). Available analytical solutions are limited to the consideration of either fringe or core degradation, which in turn limits the applicability of these solutions. Here, a new analytical approach to modelling plumes with both fringe and core degradation is presented. The approach relies on the use of readily available analytical solutions for solute transport. Using a well-known solution for three-dimensional solute transport from a planar source, an approximate solution is derived for the maximum plume length at steady-state conditions. This is verified through the use of a numerical solution. The solution suggests that the parameters controlling the plume length are: (i) the size of the contaminant source, (ii) electron acceptor to electron donor ratio, (iii) transverse dispersivities and (iv) the ratio between degradation rate constant and velocity (lambda/v). The latter term provides a simple check on the relative weights of transport to core degradation and can be used to estimate the importance of core degradation in the overall plume attenuation. The well-documented Bemidji field site has both fringe and core degradation. The new combined degradation model estimates the length of the plume with 10 m of the observed length; core only and fringe only solutions overestimate the length by more than a factor of 2.

Predictive modelling of dispersion controlled reactive plumes at the laboratory-scale.

A model-based interpretation of laboratory-scale experimental data is presented. Hydrolysis experiments carried out using thin glass tanks filled with glass beads to construct a hypothetical and inert, homogeneous porous medium were analysed using a 2D numerical model. A new empirical formula, based upon results for non-reactive (tracer) experiments is used to calculate transversal dispersivity values for a range of grain sizes and any flow velocities. Combined with effective diffusion coefficients calculated from Stokes-Einstein type equations, plume lengths arising from mixing between two solutes can be predicted accurately using numerical modelling techniques. Moreover, pH and ion concentration profiles lateral to the direction of flow of the mixing species can be determined at any given point downstream, without the need for result fitting. In our case, this approach does not lead to overpredictions of lateral mixing, as previously reported when using parameters derived from non-reactive tracer experiments to describe reactive solute transport. The theory is based on the assumption of medium homogeneity.