Abstract
This paper considers the resaturation process of a British ILW repository. Water flow into the vaults is modelled using a multiphase flow simulator TOUGH2. The time it takes for the vaults to fill with water is estimated paying particular attention to the effect of the repository design. The effects of the permeability of the backfill, fill material of the crown space and the properties of a low-permeability liner around the vaults are considered. The resaturation time of the vaults is found to be approximately 6.5 years. The presence of a low-permeability liner around the vaults is discovered to have a large impact on the resaturation time. The resaturation time is calculated using a detailed model of the vault. Another, computationally less expensive, way to model the vault is to lump the vault materials together: the vault containing the waste packages is represented by one set of parameters which combines the properties of the waste packages, backfill, crown space and liner. Previous studies have taken the permeability of the 'lumped material' to be the volume-weighted average permeability of the repository materials. This paper presents an alternative approach to the derivation of the effective permeability of the lumped material. The effective permeability across the vault is found to be approximately 3x10^<-18> m^2. The permeability along the vault is higher, 7x10^<-17> m^2, indicating that flow along the vault is faster than flow through it. The resaturation time of the repository, using the lumped permeabilities, is calculated to be 1.2 years; however it is recognised that resaturation may take considerably longer in lower permeability host rocks. It is also concluded that using volume-weighted averages may result in over-estimation of the effective permeability.
