Experimental and modelling investigations of the biogeochemistry of gas production from low and intermediate level radioactive waste

The degradation of organic wastes and the corrosion of metallic wastes and steel containers in low and intermediate level radioactive waste (LLW/ILW) repositories are important processes that affect repository geochemistry and the speciation and transport of radionuclides. Gas is generated in association with these degradation processes and this has the potential to overpressure the repository, which can promote transport of groundwater and gas, and consequently radionuclide transport. Microbial activity plays an important role in organic degradation, corrosion and gas generation through the mediation of reduction–oxidation reactions. A large-scale gas generation experiment has been established at the LLW/ILW repository, Olkiluoto, Finland to examine gas generation from LLW in waste drums disposed of in the operational VLJ Repository (VLJ is a Finnish acronym which translates to “reactor operating waste”). The experiment has monitored, for a period of 9 a, the rate and composition of gas generated, and the aqueous geochemistry and microbe populations present at various locations within the experiment. There is considerable heterogeneity within the experiment, such that pH is observed to vary from pH 5.5 to pH 10 between organic-rich waste and water associated with concrete. The heterogeneity results in competing anaerobic processes occurring together in the experiment but within different niches. Microbial activity initially dominant in organic waste has after 7 a reduced the alkalinity of the concrete influenced regions. The experiment has been modelled using a biogeochemical reaction-transport code (GRM) using a blind testing approach. Using independent data, the model was able to reproduce, within a factor of two, the rate of gas production. In addition, the model represented the main anaerobic microbial processes leading to methanogenesis and the observed spatial and temporal variations in aqueous and gaseous species. In order to model the experiment, its heterogeneity was considered such that individual waste containers were represented and assumptions were made concerning transport rates of chemical species. Cellulose waste and H2 produced by corrosion provide microbial substrates for reduction processes and CH4 generation. However, gas generation is a complex interaction of waste degradation processes. Simple repository gas generation models that consider corrosion and cellulose degradation in isolation will tend to overestimate H2 content and gas generation. The GRM model is more realistic and utilises information concerning image and other oxidised species present in LLW/ILW to consider the competition between microbial groups for electron donors that reduce the extent of H2 and CH4 generation. Models such as GRM could be applied to other repository systems, such as for high level waste and spent nuclear fuel, to evaluate how H2 gas generation from corrosion and radiolysis may be affected by microbial activity. However, this will require estimation of appropriate microbial kinetic parameters for these more extreme environments.