Abstract
Increasing global energy demands have intensified reliance on nuclear energy, as the transition away from fossil fuels and the limited capacity of renewable sources struggle to meet demand, highlighting the need for clear nuclear waste disposal pathways. Cement has been widely used for nuclear waste immobilisation, particularly for low-level and intermediate-level wastes, because it is a versatile, inexpensive, relatively simple process, has high radiation stability, and is fire and impact-resistant [1]. However, cementation increases the volume of the waste, and Portland cement may be considered suboptimal for incorporating organic liquid waste; the addition of various types of organics at high loadings will result in strong retardation and the cement may not set. Geopolymers are a subset of the broader class of alkali-activated materials (AAMs), which contain little to no calcium content [2].
The organic liquids which are the focus of this research are Nevastane and TBP (tributyl phosphate); both of which are organic oils. Oils are used as lubricants in fuel processing and manufacturing, and become classified as secondary nuclear waste due to radiological contamination. Nevastane is a generic motor oil and is used as a representation of the oil waste produced by the nuclear industry. TBP is used in the nuclear process during the PUREX process to extract valuable elements from spent nuclear fuel to be reused [3]. During the PUREX process, the spent TBP (30%, diluted with 70% kerosene, OK) is irradiated and forms radiolysis products - the majority of which is dibutyl phosphate (DBP), monobutyl phosphate (MBP), organic acids and butanol. Radioactive waste loadings of 300 μg/mL of uranium, thorium, plutonium, and other radionuclides release alpha radioactivity that has been measured at levels as high as 9.67 Bq/mL [3]. The liquid nature of the spent TBP/OK from the PUREX process means that it needs to be treated for disposal.
The NNL (National Nuclear Laboratories, UK) collaboration highlighted that geopolymer cement with immobilised oil would require additional investigation into stopping the geopolymer reaction without removal of the immobilised oil via oven drying, desiccation and freeze-drying. This was necessary to understand how the oil is interacting with the geopolymer matrix. Freeze-drying was concluded to be the best as it removed both the free and pore water without removing the oil. However, it did cause the sample to crumble so this route would only be used if the sample could not be measured at the set time (i.e. seven, 28 and 90 days).
The nature of the organic oils within the geopolymer matrix caused traditional technical used in cement analysis to be re-investigated and validated before the routes of immobilisation could be investigated and compared. For example, leaching tests could not be done with ICP-MS as the oil would damage the instrument. Instead, calibration graphs based on pH values, total organic carbon analysis (TOC) with various surfactants and adaptions to the leaching setup were investigated, but due to the nature of the oils were deemed unacceptable.
Organic wastes such as oils can be immobilised in cement by three different techniques: direct incorporation, preemulsification, and solid impregnation [4]. Direct incorporation is the simplest of the routes and was used in the validation of suitable analytical techniques. It found that 10 – 30 wt% Nevastane oil could be immobilised in potassium geopolymer cement without liquid separation, whilst, TBP/Dodecane had liquid separation at all loadings. Therefore, the two routes of pre-emulsification and impregnation were investigated for TBP/Dodecane immobilisation.
A Pickering emulsion with silica fume was used for the pre-emulsification route for immobilisation of Nevastane and TBP/Dodecane. A Pickering emulsion works by creating a layer of solid particles between the liquid and the oil to minimise tension and cause the emulsion to have greater stability [5]. The solid particles attach to the interface so provide superior stability to the emulsion than a surfactant [6].
The impregnation route involves using a solid precursor with organic oil and has been less investigated in the literature compared to direct and emulsification routes [4-6]. This route could be advantageous for the immobilisation of organic oils as the absorption of the oils into a precursor before addition to the cement allows for an additional layer of encapsulation. The additional layer of protection would be especially suitable for highly leachable toxic compounds that could be a potential issue in contaminated organic oil. Furthermore, this technique does not require the formation of a stable emulsion compared to direct incorporation and pre-emulsification.
The collaboration with INL focused on understanding how radiation affects the ability of the geopolymer matrix to immobilise the organic oil initially and during long-term storage. This was done in a number of ways:
• Irradiating the geopolymer cement with no oil, 20 wt% Nevastane oil and 20 wt% TBP/Dodecane at 0 – 40 kGy, 1 MPa, 5 MPa and 10 MPa and the headspace was analysed by GC for hydrogen, carbon dioxide and alkanes
• The precursors (metakaolin, organic oils and surfactants) used to create geopolymer cement with immobilised oil were individually irradiated from 0 – 40 kGy in argon and in air. Argon was used for the maximum gas generation whilst the air environment would be more realistic of a long-term storage facility
• Nevastane oil, TBP/Dodecane and PowerPozz metakaolin were all irradiated to 40 kGy, and used via a design of experiments methodology to produce geopolymer cement with irradiated precursors. The effect of radiation of the initial and final setting, compressive strength at seven and 28 days, chemical bonds and crystalline phases were investigated
The collaboration with INL was able to investigate work that had not been covered in depth in the literature. For example, the irradiation of Nevastane degraded to hydrogen, methane, ethane and numerous products that could not be identified. When these were identified by GC-MS, the majority of the products were highly flammable, had low explosion limits, high vapour density and hazards to health and the environment. The work discovered that if Nevastane waste was to be immobilised in geopolymer cement in long-term storage, detectors would have to be installed at the ground level of the facility. This would be to monitor chemicals released from the samples to prevent fires and/or explosions.
In conclusion, this PhD focused on creating a robust geopolymer cement for the immobilisation of radioactive waste via various routes of immobilisation, whilst also investigating the effects of radiation during mixing, initial and final setting of the cement in terms of the implications for long-term storage or disposal of the waste packages.
