Sensitivity Analysis of a Coupled Hydro-mechanical Paleo-Climate Model of Density-dependent Groundwater Flow in Discretely Fractured Crystalline Rock

Main Article Content

Stefano D. Normani
Jonathan F. Sykes

Abstract

A high resolution three-dimensional sub-regional scale (104km2) density-dependent, discretely fractured groundwater flow model with hydro-mechanical coupling and pseudo-permafrost was developed from a larger 5734km2 regional-scale groundwater flow model of a Canadian Shield setting. The objective of the work is to determine the sensitivity of modelled groundwater system evolution to the hydro-mechanical parameters.

The discrete fracture dual continuum numerical model FRAC3DVS-OPG was used for all simulations. A discrete fracture network model delineated from surface features was superimposed onto an approximate 790000 element domain mesh with approximately 850000 nodes. Orthogonal fracture faces (between adjacent finite element grid blocks) were used to best represent the irregular discrete fracture zone network. Interconnectivity of the permeable fracture zones is an important pathway for the possible migration and subsequent reduction in groundwater and contaminant residence times. The crystalline rock matrix between these structural discontinuities was assigned mechanical and flow properties characteristic of those reported for the Canadian Shield. The variation of total dissolved solids with depth was assigned using literature data for the Canadian Shield. Performance measures for the sensitivity analysis include equivalent freshwater heads, environmental heads, linear velocities, and depth of penetration by conservative non-decaying tracers released at the surface.

A 121000 year North American continental scale paleo-climate simulation was applied to the domain with ice-sheet histories estimated by the University of Toronto Glacial Systems Model (UofT GSM). Hydro-mechanical coupling between the rock matrix and the pore fluid, due to the ice sheet normal stress, was included in the simulations. The flow model included the influence of vertical strain and assumed that areal loads were homogeneous. Permafrost depth was applied as a permeability reduction to both three-dimensional grid blocks and fractures that lie within the time varying permafrost zone. Values of ice sheet normal stress and proglacial lake depth from the UofT GSM were applied to the sub-regional model as surface boundary conditions using a freshwater head equivalent to the normal stress imposed by the ice sheet at its base. The sensitivity of glacial meltwater penetration to different conceptualizations of hydro-mechanical properties were investigated.

Article Details

Section
Articles