Leonardo Romero

Sensitivity of the Radionuclide Release from a Repository to the Variability of Materials and Other Properties

Radionuclides leaking from a damaged canister spread into the clay surrounding the canister and then migrate through various pathways into the mobile water in the rock fractures. To quantify uncertainties regarding the release to the far field, a sensitivity analysis is performed for variations in the parameters governing the release. Variations in the size of the damage in the canister, uranium solubility, sorption capacity of the clay, and hydraulic properties of the fractured rock nearest to the canister have to be accurately determined in order to minimize uncertainties in the release.

Movement of the redox front at the Osamu Utsumi uranium mine, Poços de Caldas, Brazil

Abstract In the open pit uranium mine at Pocos de Caldas the uranium oxide ore has accumulated at a redox front. The redox front forms and moves by the oxidation of pyrite in the rock by oxygen dissolved in the infiltrating rainwater. The phonolitic host-rock was formed about 90 million years ago. It is possible that between 3 and 12 km of rock has been eroded since then. The redox front lies between tens and several hundred meters below the ground surface. The front is very irregular and locally there are oxidation “fingers” which extend well ahead of the main front. The redox fingers are often associated with fractures or fracture zones. An attempt has been made to model the formation and evolution of the redox front at Pocos de Caldas by applying chemical transport models of different complexity. One class of models used is based on advective flow in a porous medium; the continuum models. Another class of models used is based on fracture flow with transport of oxygen into the rock matrix by molecular diffusion. In the latter class of models, channels with different flowrates are modelled and the flowrate is allowed to vary over time stochastically to account for the changing properties of the hydraulic network. The erosion of the surface is as fast as the redox front movement which stabilizes at a distance below the ground surface because the water flux decreases with depth. Another possible cause for stabilization of the tips of the redox fingers is that the flow channels form a network with variable flow properties. The main chemical reactions, i.e. pyrite oxidation, clay formation and the accumulation of uranium oxide at the redox front, are well in accordance with expectations and with what a chemical transport model predicts. The formation of redox fingers may partly be caused by variable hydraulic conductivity of the rock and partly by the presence of highly conductive fractures and channels in fractures. Several mechanisms contribute to the formation of the complex shape of the redox front. Some of the processes have been conceptualized and some have been quantified in the form of mathematical models which were used to make quantitative calculations. The calculations have given considerable support to the determination of which mechanisms may be important. The overall oxidized zone can be explained by advective transport in a porous medium. The fingered zone around fractures would be explained by flow through fractures (channelling) and indicates that matrix diffusion is clearly present as a mechanism to access the interior of the rock.

The fast multiple-path NUCTRAN model -- Calculating the radionuclide release from a repository

The NUCTRAN model has been applied to the Swedish KBS-3 nuclear waste repository concept, where the migration of radionuclides is through various barriers and pathways. The escape of the nuclides from the canister occurs through a small hole. This hole controls the release of nuclides from the repository. NUCTRAN is a useful tool to calculate the nonstationary transport in a repository for high-level nuclear waste. The advantage of this model is the use of a coarse compartmentalization of the repository, which makes it flexible and easy to adapt to different geometries. The several radionuclide release calculations made with NUCTRAN have shown the capability of this to handle different situations rapidly and easily. The particularity of these calculations is the high accuracy obtained by using a coarse compartmentalization of the Swedish KBS-3 repository and the small requirements of computing time. At short times for short-lived nuclides, the calculated releases are exaggerated. The error can be considerably reduced by an additional subdivision of large compartments into a few compartments.

Movement of a Redox Front Around a Repository for High-Level Nuclear Waste

The evolution and formation of a redox front in the near field of a repository for high-level nuclear waste is modeled, considering the effect of a time-dependent source term of the oxidizing species produced by radiolysis of the water entering the canister. The transport of oxidants in the clay surrounding the canister occurs by diffusion. In the rock, the transport of oxidants is modeled as occurring through fractures, with diffusion into the rock matrix. The results show that it is not probable that the redox front will ever move past the bentonite. If it does, the tips of the redox front may move < 100 m over a million year period in a channel with a high flow rate

Transport from the Canister to the Biosphere: Using an Integrated Near-and Far-Field Model

A coupled model concept which may be used for performance assessment of a nuclear repository is presented. The tool is developed by integration of two models, one near field and one far field model. A compartment model, NUCTRAN, is used to calculate the near field release from a damaged canister. The far field transport through fractured rock is simulated by using CHAN3D, based on a three-dimensional stochastic channel network concept. The near field release depends on the local hydraulic properties of the far field. The transport in the far field in turn depends on where the damaged canister(s) is located. The very large heterogeneities in the rock mass makes it necessary to study both the near field release properties and the location of release at the same time. In order to demonstrate the capabilities of the coupled model concept it is applied on a hypothetical repository located at the Hard Rock Laboratory in Aspo, Sweden. Two main items were studied; the location of a damaged canister in relation to fracture zones and the barrier function of the host rock. In the study of the near field rock as a transport barrier the effect of different tunnel excavation methods which may influence the damage level of the rock around the tunnel was addressed.

Fast Model for Calculating Steady State Release of Radionuclides from the Near Field

A model has been developed which describes the steady state transport of dissolving species of radionuclides from a single canister in a repository for spent nuclear fuel into the passing water in fractures in the surrounding rock. The steady state transport of nuclides is described by a network of transport resistances, coupled together in the same way as an electrical circuit network. With the model a number of calculations are done for various sets of fracture geometry data. The calculations indicate that the resistance network model gives results comparable to those of a complex 3-dimensional numerical model. The present model is very simple and fast.