|dc.description.abstract||The Swedish Nuclear Fuel and Waste Management Company (SKB)  is in charge of dealing safely with all the radioactive waste from nuclear power plants in Sweden. In 2009 Forsmark (Sweden) was selected to build the spent nuclear fuel repository of the country. The radioactive waste is stored deep in the soil and therefore, studying the properties of the area is essential.
The clay present in the selected region can be defined as fractured crystalline rock, meaning, rocks have fractures and negative charges in their surface. This negatively charged media has a relevant effect on the diffusion process for charged solutes, for instance, it affects the diffusion of spent nuclear fuel. This phenomenon, known as anion exclusion, causes a reduction in the effective porosity for anions in the clay soil (meaning the porosity of the soil through which anions can flow) and, consequently, it reduces the retention into the rock matrix.
Current existing software lack capabilities to model anion exclusion phenomenon. For this reason, SKB started a project called RetNet to provide more knowledge of the transport behaviour of spent nuclear fuel in the Swedish repository. As part of this project, Amphos 21 Consulting S.L. (a scientific and strategic environmental consulting) together with SKB, opted for improving an existing computer code for reactive transport modelling. In particular, PFLOTRAN is the open-source reactive transport software used for such project.
Amphos 21 has already developed a first improvement of PFLOTRAN that allows modelling anion exclusion in a single fracture-matrix system. To do so, the Nernst-Planck equation (an equation to describe the transport of charged chemical species) is included in the transport module of PFLOTRAN code. Nonetheless, for large-scale multicomponent reactive transport model the current improvements are still not enough. In this case, the dual continuum formulation already implemented in PFLOTRAN should be adapted, including in the model fracture and rock matrix processes.
Therefore, the main objectives of this thesis are to validate the first implementation done by the consultancy with a test case, and to start the development of the second improvement. For this second goal it is also essential to include the Nernst-Planck equation in the code.
The results obtained from the test case of the first implementation prove that the transport of ionic species in a single fracture-matrix system is as expected and it concurrently validates the Nernst-Planck approach. The results obtained from the second implementation were compared with a benchmark drawn from a report, concluding that the results fit the benchmark data. Nonetheless, after performing additional tests, some issues regarding the convergence capability of the code arise. For this reason, further studying and reviewing of the new implementation is required. Therefore, some hypotheses are suggested to allow continuing with the RetNet project and further improve the PFLOTRAN code.