Microbiology of bentonite clay relevant to a deep geological repository for used nuclear fuel

Abstract

Nuclear energy is an important source of energy globally, but results in the production of used nuclear fuel. When removed from a nuclear reactor after its useful lifetime, used nuclear fuel is still highly radioactive and must be stored safely for approximately one million years until it returns to the radioactivity level of naturally occurring uranium ore. Canada, along with other countries that have an inventory of used nuclear fuel, is in the process of designing a deep geological repository (DGR) for long-term storage of used nuclear fuel. In this system, used nuclear fuel, itself a stable solid, will be stored in copper-coated steel used fuel containers, and placed in bentonite clay buffer boxes made of highly compacted Wyoming MX-80 bentonite. Buffer boxes would then be stacked in placement rooms approximately 500 m below ground in a suitable host rock with spaces between buffer boxes and host rock filled in with a granulated form of bentonite referred to as gapfill material. To ensure the longevity of a DGR, it is important to consider the role that microorganisms could play, particularly through contributing to corrosion of used fuel containers through a process termed microbiologically influenced corrosion. Expected to dominate occurrences of microbiologically influenced corrosion under anoxic conditions, sulfate-reducing bacteria (SRB) are a primary target of research, but rarely live in isolation, thus necessitating the study of microbial communities that could live in the bentonite clay surrounding used fuel containers on a broader scale.

Due to the potentially detrimental role that some microorganisms could play in a DGR, a goal in designing a DGR is to compact bentonite, a type of swelling clay, to a sufficiently high dry density that microbial growth is suppressed upon saturation. The first goal of this thesis was to investigate the bentonite dry density necessary to suppress microbial growth in Wyoming MX-80 bentonite under oxic conditions and gapfill material under both oxic and anoxic conditions. A set of pressure vessel experiments demonstrated suppression of microbial growth under oxic conditions in bentonite compacted to a minimum dry density of 1.4 g/cm3. Under anoxic conditions, growth of heterotrophs was suppressed in pressure vessels with a minimum dry density of 1.45 g/cm3, but culturable SRB persisted in abundances higher than the as-received bentonite starting material in the outer layers of all pressure vessels throughout the full one-year experiment. Additional experiments were conducted to explain the increase in abundance of SRB, which had not previously been observed in other studies. These experiments suggested that the increase in abundance of SRB was likely not inherent to the gapfill material, nor was there evidence for it being a result of differences in SRB medium or in amounts of trace oxygen between studies. In both oxic and anoxic pressure vessels, an initial increase in abundance of culturable heterotrophs was observed prior to complete saturation, presumably as water became available but swelling pressure remained sufficiently low to allow for their growth. Although previously proposed to potentially be associated with a recovery from the viable but not culturable state rather than growth, a follow-up experiment suggested that the initial increase in abundance of culturable heterotrophs was likely a reflection of growth (i.e., cell division).

Dry density is an important consideration in DGR design, but it is not the only physical property that could influence microbial growth within a DGR. Temperature is expected to fluctuate from natural subsurface temperatures of <20°C to temperatures as high as 94°C, and little research has been conducted to explore the potential for microbial growth in bentonite at these elevated temperatures. This thesis includes experiments testing the abundance and community composition of microorganisms adapted to a variety of temperatures in as-received and hydrated bentonite. The results showed a low abundance of culturable microorganisms that survived incubation at 60°C, but 16S rRNA gene profiles dominated by presumably unculturable representatives of the thermophilic family Thermoactinomycetaceae. Hydrated bentonite was additionally incubated at temperatures of 75, 90, and 105°C, but DNA sequencing results did not show a shift in community composition from as-received bentonite, suggesting that the natural as-received bentonite microbial community may not include members adapted to these very high temperatures.

Lab-scale experiments allow for testing of very specific DGR-relevant conditions (e.g., dry density and temperature) and the effect these have on microbial community abundance and composition. However, a DGR is being designed to exist for upwards of one hundred thousand years, which is not a realistic timescale for any experiment. To circumvent this limitation, one approach is to couple lab-scale experiments to the study of natural analogues, which have naturally existed for DGR-relevant timeframes. This thesis presents a study of the Tsukinuno bentonite deposit in Japan, which can serve as a natural analogue to DGR bentonite. In this study, sequencing of DNA extracted from bentonite revealed microbial communities dominated by sequences associated with Thiobacillus, Hydrogenophaga, Comamonadaceae, and Pseudomonas. Although differences in community composition were observed between samples, microbial communities were relatively similar for all four studied cores and at all depths into the clay bed. A series of geochemical parameters were measured to help identify factors that may influence microbial community composition. The abundance of culturable anaerobic heterotrophs was positively correlated with the concentration of nitrate, which could be used by anaerobes for denitrification, and the abundance of culturable aerobic and anaerobic heterotrophs was negatively correlated with the abundance of the clay mineral montmorillonite, increased concentrations of which would increase the swelling capacity of the bentonite. The results presented throughout this thesis will together be useful for incorporation into future models of microbial activity within a DGR and can ultimately be used to inform DGR design.