Periodic Reporting for period 3 - DISCO (Modern spent fuel dissolution and chemistry in failed container conditions)
Reporting period: 2020-06-01 to 2021-11-30
The DisCo project aimed to improve the scientific understanding of the basis of the safety cases for Spent Nuclear Fuel repositories. The issue is whether the spent fuel dissolution process is affected by the composition and characteristics of the spent fuel itself. The behaviour of modern fuels after being used in a nuclear reactor may be different from of the traditional spent nuclear fuel. The objectives of the DisCo project were 1) to enhance our understanding of spent fuel matrix dissolution under conditions representative of failed containers in reducing repository environments; and 2) to assess whether MOX and doped fuels behave in a similar manner to conventional fuels. Experimental and modelling tasks were defined to achieve the project objectives. The results are important to Waste Management Organisations but also of interest for researchers in universities and higher education.
One main conclusion of the project is that matrix dissolution behaviour of MOX, fuels doped with Cr and Al, and standard fuel types is similar. Therefore, the modern fuel types studied in this project can be accepted for the planned repository designs without any obvious negative effect on the safety assessments. Project results have emphasized the importance of a reducing environment and the presence of corroding iron and hydrogen in the repositories. Models that take these important features and processes into account have been developed and are now available for use in safety cases. Overall, knowledge has been gained of the crystallography, mineral chemistry and microstructure of the uranium oxide that constitutes the matrix of most standard and modern nuclear fuels.
One main conclusion of the project is that matrix dissolution behaviour of MOX, fuels doped with Cr and Al, and standard fuel types is similar. Therefore, the modern fuel types studied in this project can be accepted for the planned repository designs without any obvious negative effect on the safety assessments. Project results have emphasized the importance of a reducing environment and the presence of corroding iron and hydrogen in the repositories. Models that take these important features and processes into account have been developed and are now available for use in safety cases. Overall, knowledge has been gained of the crystallography, mineral chemistry and microstructure of the uranium oxide that constitutes the matrix of most standard and modern nuclear fuels.
Due to the Covid-19 pandemic the project was extended by 6 months and an extra project meeting was arranged 6 months before the end of the project. Communication, training activities and dissemination of results have proceeded successfully.
In WP2, the preparation and characterisation of a range of samples and experimental systems was coordinated. The samples included real spent fuels, synthetic uranium dioxide systems with and without dopants including alpha-emitters. Much effort was devoted to the synthesis of optimised model systems. Samples were characterized before and after experiments, to investigate uranium solid state chemistry and effects of chemical components on the crystalline characteristics and microstructure. The work in WP2 produced uranium oxide pellets with varying content of additives and alpha-emitting Pu, as well as optimised synthesizing methods. Included in WP2 was a study concerning the effects of prolonged water contact on damaged fuel. The results show negligible effect on the spent fuel matrix, but further analyses are needed.
Dissolution experiments involving spent nuclear fuel were performed in WP3. The focus was on experiments under reducing conditions, with hydrogen overpressure, to study the radionuclide release from spent standard fuels, MOX fuels and fuels doped with Cr and with Cr+Al. Most experiments used a simple solution with bicarbonate and sodium chloride, while a few experiments used a more complex solution mimicking a young cement water with Ca. With regards to spent fuel matrix dissolution the project has provided results from experiments that have run for at least two years. Data produced in WP3 data is informative for rapid radionuclide release, disconnected to the spent fuel matrix dissolution. For the long-term experiments performed under hydrogen overpressure, it was found that the hydrogen inhibits oxidative dissolution of the uranium oxide matrix, but it appears a slow dissolution continues beyond ca two years under strongly reducing conditions. This indicates there is more to learn about the hydrogen effect but more importantly, the results show that the effects of dopants (Cr, Cr+Al, Pu) are insignificant with regards to the long-term dissolution of spent nuclear fuel.
In WP4, samples with varying complexity and compositions ranging from pure uranium dioxide to uranium oxide with Cr and Al, and Cr and Pu-238, and varying amount of Th (to mimic MOX) have been dissolved using a range of aqueous solutions, Eh and pH. After the experiments were concluded, samples were studied and compared with the as-prepared samples. Dissolution data were collected via an extensive experimental programme and related to microstructure, doping and reactive surface area. The results indicate insignificant effects on the fuel matrix by adding chromium to the uranium oxide. The carefully characterised model materials studied in WP4 provided insights regarding the behaviour of spent nuclear fuels under repository conditions that are complimentary to the results from spent fuel experiments.
Both the dry solid state and the dissolution and transport processes in various heterogeneous systems were studied in WP5 by modelling. Numerical models were developed to study effect of dopants in terms of oxygen potential as well as on the precipitation of solids inside a failed iron-based canister. Development of a matrix dissolution model incorporated reactions between matrix, separate phases, radiolysis and hydrogen. Radiolytical, chemical and reactive transport modelling involved thermodynamics and kinetics and, in one case, electrochemistry. The MOX system focused on effects of Pu on radiolysis and of anoxic Fe corrosion. When applicable, experimental data were used, for model calibration and parametrization. Modelling results have clarified the effect of additives in fuel and the effect of near-field materials, specifically Fe. The models support the hypothesis and experimental data indicating that the effect of Cr-doping on both fuel oxygen potential and dissolution process is insignificant. Modelling performed in the project has enhanced the understanding of important processes expected to occur inside a failed canister, especially concerning the effect of near-field materials, such as Fe.
In WP2, the preparation and characterisation of a range of samples and experimental systems was coordinated. The samples included real spent fuels, synthetic uranium dioxide systems with and without dopants including alpha-emitters. Much effort was devoted to the synthesis of optimised model systems. Samples were characterized before and after experiments, to investigate uranium solid state chemistry and effects of chemical components on the crystalline characteristics and microstructure. The work in WP2 produced uranium oxide pellets with varying content of additives and alpha-emitting Pu, as well as optimised synthesizing methods. Included in WP2 was a study concerning the effects of prolonged water contact on damaged fuel. The results show negligible effect on the spent fuel matrix, but further analyses are needed.
Dissolution experiments involving spent nuclear fuel were performed in WP3. The focus was on experiments under reducing conditions, with hydrogen overpressure, to study the radionuclide release from spent standard fuels, MOX fuels and fuels doped with Cr and with Cr+Al. Most experiments used a simple solution with bicarbonate and sodium chloride, while a few experiments used a more complex solution mimicking a young cement water with Ca. With regards to spent fuel matrix dissolution the project has provided results from experiments that have run for at least two years. Data produced in WP3 data is informative for rapid radionuclide release, disconnected to the spent fuel matrix dissolution. For the long-term experiments performed under hydrogen overpressure, it was found that the hydrogen inhibits oxidative dissolution of the uranium oxide matrix, but it appears a slow dissolution continues beyond ca two years under strongly reducing conditions. This indicates there is more to learn about the hydrogen effect but more importantly, the results show that the effects of dopants (Cr, Cr+Al, Pu) are insignificant with regards to the long-term dissolution of spent nuclear fuel.
In WP4, samples with varying complexity and compositions ranging from pure uranium dioxide to uranium oxide with Cr and Al, and Cr and Pu-238, and varying amount of Th (to mimic MOX) have been dissolved using a range of aqueous solutions, Eh and pH. After the experiments were concluded, samples were studied and compared with the as-prepared samples. Dissolution data were collected via an extensive experimental programme and related to microstructure, doping and reactive surface area. The results indicate insignificant effects on the fuel matrix by adding chromium to the uranium oxide. The carefully characterised model materials studied in WP4 provided insights regarding the behaviour of spent nuclear fuels under repository conditions that are complimentary to the results from spent fuel experiments.
Both the dry solid state and the dissolution and transport processes in various heterogeneous systems were studied in WP5 by modelling. Numerical models were developed to study effect of dopants in terms of oxygen potential as well as on the precipitation of solids inside a failed iron-based canister. Development of a matrix dissolution model incorporated reactions between matrix, separate phases, radiolysis and hydrogen. Radiolytical, chemical and reactive transport modelling involved thermodynamics and kinetics and, in one case, electrochemistry. The MOX system focused on effects of Pu on radiolysis and of anoxic Fe corrosion. When applicable, experimental data were used, for model calibration and parametrization. Modelling results have clarified the effect of additives in fuel and the effect of near-field materials, specifically Fe. The models support the hypothesis and experimental data indicating that the effect of Cr-doping on both fuel oxygen potential and dissolution process is insignificant. Modelling performed in the project has enhanced the understanding of important processes expected to occur inside a failed canister, especially concerning the effect of near-field materials, such as Fe.
The dissolution data and model developments generated in this project have improved the understanding of processes controlling the oxidative dissolution of uranium dioxide and spent fuel. The results allow the comparison between the redox response of traditional versus modern fuels. By systematically investigating the effects of individual dopants in a uranium dioxide matrix, details of the reactions involved in oxidative dissolution were revealed.
An outcome of this project is improved knowledge on the effects of dopants on solid state characteristics such as solubility and oxygen potential of the solid. One question that now has been answered is if changes in spent fuel chemistry invoked by the dopants may also change the radionuclide release pattern from these new types of fuels. The project results have clarified that, for the studied doped fuels, this effect is insignificant. To calculate the dissolution rate of spent fuel relevant for very long times (up to one million years) models are needed: this project developed chemical models of the dissolution for this purpose.
An outcome of this project is improved knowledge on the effects of dopants on solid state characteristics such as solubility and oxygen potential of the solid. One question that now has been answered is if changes in spent fuel chemistry invoked by the dopants may also change the radionuclide release pattern from these new types of fuels. The project results have clarified that, for the studied doped fuels, this effect is insignificant. To calculate the dissolution rate of spent fuel relevant for very long times (up to one million years) models are needed: this project developed chemical models of the dissolution for this purpose.