Simulating groundwater flow and contaminant migration in heterogeneous alluvial sediments

Sellafield is a large nuclear fuel reprocessing, nuclear waste storage, nuclear decommissioning and former nuclear power generating site on the coast of Cumbria. It is due to be fully decommissioned by 2120. Many of the UK nuclear sites, are located on coastal plains and are underlain by thick fluvial-glacial alluvium deposits. These sediments are highly heterogeneous in nature and consist of layers ranging from clay size right up to cobble sized materials. Lateral variability in sediment texture reflects changing depositional environments. Historically there has been several events where unauthorised releases of radionuclide and/ or hydrocarbon containing fluids to groundwater have occurred. These include radionuclides such as Tc-99 and H-3 that are highly mobile in groundwater, others such as Sr-90, C-137 and U which display sorption controlled transport, and hydrocarbons and their resulting dissolution products. At Sellafield, over 200 groundwater monitoring boreholes exist, where aquifer materials and groundwater data have been recorded. This shows extensive radioactive and chemical contamination plumes exist in this very complex unconsolidated aquifer structure.

The project will develop new modelling approaches which considers the three-dimensional complexity of unconsolidated fluvial aquifers, and the variable sediment and groundwater compositions present, to better understand and predict contaminant migration.

A numerical modelling study will be undertaken to establish the influence of sediment heterogeneity on the migration of radionuclides, chemical contaminants and saline interface behaviour in glacio-fluvial sediments. In the first phase of the work, fundamental approaches to modelling solute transport such as particle tracing, advection-dispersion, and Continuous Time Random Walk (CTRW) will be assessed for their suitability for tackling these problems. Once appropriate fundamental modelling approaches have been identified, it will be necessary to review and select from the available codes for implementation.
In the second phase, permeability and sorption properties (e.g. distribution co-efficient Kd) of the glacio-fluvial sediments will be obtained from pre-existing data on the site core samples [measured in lab experiments conducted in a currently-running EPSRC-DTP funded PhD project]. Permeability and Kd will be estimated empirically from lithology data on grainsize and sediment composition, and spatial statistical properties will be determined. Available approaches for the stochastic generation of simulated permeability/Kd fields will be reviewed. Selected approaches will be used to generate permeability/Kd fields representing the sediments.
In the third phase of the work solute transport modelling will be applied to the generated permeability/Kd fields in order to constrain the envelope of possible contaminant impacts from the site. Site groundwater monitoring data will constrain modelled hydraulic gradients and boundary conditions. The influence of heterogeneous sediment permeability and sorption properties on three specific solute transport scenarios relevant to contaminant impacts may be investigated i) migration of radionuclide plumes (i.e. non-sorbing contaminants U and H-3; sorbing contaminants Sr-90, Tc-99 and Cr-137) ii) transport, distribution and degradation of dissolved phase hydrocarbon contaminants (relevant to the longevity of such contaminants in terms of the proposed 100 year period to site closure) and iii) behaviour of the saline-freshwater interface at this coastal site, which may influence both sorption/migration and degradation of dissolved-phase contaminants.