Multi-dimensional Modelling of Groundwater Flow and Contaminant Transport in Fractured Crystalline Rock of the Chalk River Laboratories Site

Abstract

Two- and three-dimensional (2-D and 3-D) hydrogeological simulations were performed to assist in the assessment of the suitability of the CRL site for hosting a deep geologic repository to provide safe long-term management of CRL's low- and intermediate-level wastes. The modelled flow domain includes the bedrock of the CRL site and its immediate vicinity with a number of highly permeable major faults/fracture zones explicitly represented. The 2-D modelling indicates that i) a 10-km length scale is sufficiently large, ii) the predicted groundwater flow field is only weakly influenced by in situ depth-dependent natural temperature and salinity at the CRL site, and iii) advection and dispersion, rather than diffusion, are the predominant contaminant transport mechanisms. A 3-D conceptual hydrogeological model, which encompasses a 165-km² area, extends to 3 km below surface and includes eight faults/fracture zones, was developed based on a preliminary 3-D geological framework assembled from data available in 2009. Particle tracking analysis for nearly 500 particles released from a 20-km² area at 500-m below surface into the predicted groundwater flow field showed that particles from the northeastern two-thirds of the release area would discharge to the Ottawa River while the remaining particles discharge to Maskinonge Lake, with substantially shorter travel times for the former discharge area. Using this and other information the conceptual design team proposed potential footprints for a repository with an approximate area of 1.6 km² at either 500 m or 1000 m below surface. Advective transport through the Geosphere (bedrock) was investigated by tracking more than 500 particles from each hypothetical repository. Predicted travel times suggest a moderately strong natural barrier. A network of 10-20 linear segments was constructed to approximate the flowpaths from each hypothetical repository to surface discharge. The network geometries, head values, travel times and corresponding repository sector information were provided to the Postclosure Performance and Safety Assessment team. Sensitivity analyses were performed to investigate the influence of various model input parameters. The head profiles predicted by the Base Case simulations using two codes and the simulation with the best-fit rock-mass permeability available in 2010 were compared to the measured head profiles in four boreholes. Agreement was only poor to fair, indicating that the subsurface head distribution has not yet equilibrated and that uncertainties remain in the conceptual and numerical models, and the sitespecific hydrogeological parameters. The simulation using rock-mass permeability fitted to data from three boreholes predicted significantly slower groundwater flow.