GROUNDWATER PROTECTION IN COMPLEX GLACIATED TERRAINS

Aquifer systems in complex glaciated terrains often pose a challenge to the hydrogeologist engaged in developing wellhead protection measures. Glacial systems can consist of many distinct but often discontinuous hydrogeologic units of varying permeability, some acting as aquifers and some acting as aquitards, making the system inherently three-dimensional. It can be problematic to represent this complexity within the context of conventional methodologies based on the delineation of well capture zones assuming a two-dimensional aquifer. A second problem arises in that a capture zone outlines the area within which a particle is captured by the well, but does not necessarily express the impact of contamination on the well. A third problem is the uncertainty in the distribution of the hydrogeologic parameters in three dimensions.

We discuss various approaches for meeting these challenges. All depend on a realistic and properly validated 3D model. The variability of the hydrogeologic parameters throughout the system can be represented by means of 3D kriging. The standard methodology for capture zone delineation is backward particle tracking, but this approach can be problematic in complex 3D systems. Capture probability modeling using a standard advective-dispersive transport model, which can be set up in backward mode to generate a 3D probability plume, is often more reliable. The plume can be interpreted as capture probability at ground surface, or as maximum capture probability over the aquifer depth. A third approach, the well vulnerability approach, uses the same type of model and gives the actual impact of a spill on the well. The impact can be expressed in terms of the time taken for the arrival of maximum concentrations, the maximum concentrations to be expected, or the first arrival time of concentrations above some drinking water standard. The information generated can be more useful than that obtained from conventional capture zone analysis. The results can also be interpreted in terms of the economic cost of contamination. The latter two approaches can take complexities into account in a realistic way, and can represent uncertainties at the local scale through the dispersion term. The methodology is demonstrated using the Greenbrook well field located within the complex Waterloo Moraine aquifer system in Ontario.