Nuclear waste repositories in salt mines: a new approach to safety assessment

Abstract

The long-term safety of radioactive waste repositories in rock-salt mines in the deep underground benefits significantly from the barrier effect of overlying rocks. The concentrations of radioactive substances released from the repository and migrating in the aquifer up to the biosphere are greatly reduced during passage through these rocks. In former safety analyses of waste repositories this transport has generally been modelled as a combination of the involved phenomena, e.g. convection, dispersion, adsorption, etc. The data required for a numerical evaluation of the overall effect are obtained either as (conservative) estimates based on experience or are empirical, based mainly on laboratory experiments. The approach presented here is much simpler and entirely empirical, and therefore more transparent. It makes use of the fact that the groundwater in the overlying rocks always contains dissolved salt from the salt formation and carries it continuously into the receiving channels or the drainage system. The relation between the total amount of dissolved solids present in a certain subsurface catchment area and their steady-state concentration in the receiving channels is assumed to be equivalent to the relation between the given amount of radionuclides released from the repository and their concentration in the receiving channels, the latter leading to a certain radiation exposure of the population. Two versions of this approach are discussed: version (a) assumes a continuous stream of radionuclides released from the repository, and version (b) assumes a pulse release of radionuclides from the repository. A simple calculation using data from the Gorleben exploration leads to the inequality C _max(W)≤7.85 · 10⁻¹⁵ W(Bq/l) (1) where C _max is the maximum radionuclide concentration (with respect to time) in the receiving channels and W(Bq) is the amount of radionuclides released from the repository in a very short time. C _max obtained from (1), is supposed to be an upper limit of the radionuclide concentration in the receiving channel and, therefore, a conservative estimate. In some catchment areas the salt concentration in a small region near the surface is higher by a factor of <2 to 3 than the concentration in the receiving channel. In order to be conservative, this empirical factor may be used to calculate the concentration of radioactive substances. Surprisingly, the values of radiation exposure resulting from both versions of the new approach are far below those calculated by applying elaborated models of the transport processes. The respective factors are 10–100 in case (a) and as much as 1000 in case (b). The reasons for these significant differences are discussed.