A Radionuclide Release Model for Spent UO and MOX Fuel for Safety 2 Assessment with Application to Waste to be Disposed of in a Deep Geological Repository in Switzerland
Johnson, L.H., Curti, E. & Spahiu, K. (2023): A Radionuclide Release Model for Spent UO2 and MOX Fuel for Safety Assessment with Application to Waste to be Disposed of in a Deep Geological Repository in Switzerland. Nagra Arbeitsbericht NAB 23-10.
The main objective of the present report is to provide a model for the release of radionuclides from spent fuel for safety assessment calculations that support the general licence application for a repository for HLW in Switzerland, including the definition and justification of parameter values for the model. This goal is achieved based on a detailed literature review, auxiliary calculations and expert judgement. The review includes data from studies published in peerreviewed articles, as well as in public reports from waste management organizations and EURATOM projects dedicated to spent fuel, up to “FIRST-Nuclides”. With few exceptions, the results from the recently finalized DisCo project could not be included in the present review because most of the data produced have not yet been evaluated in a finalised form.
The heterogeneous distribution of radionuclides within spent fuel rods strongly influences the timing and kinetics of their release into pore water after eventual breaching of disposal canisters. This report provides both the basis for assessing the distribution of radionuclides associated with the spent fuel matrix and the model for the kinetics of their release.
The model for release of radionuclides considers the following processes:
- Release of the so-called IRF (instant release fraction) from the fuel – This applies to a small subset of volatile radionuclides and is defined as
IRF = IRFG + IRFGB,
where IRFG represents the fraction of the total inventory of the radionuclides in the fuel that is released from the gap (the interconnected void space in the fuel rod) and IRFGB is the fraction of the total inventory of the radionuclide in the fuel that is released from grain boundaries. Release from the gap is highly correlated to fission gas release (FGR) since gaseous species accumulate in open spaces at the end of fuel irradiation in the reactor. The term IRF is used in most of the waste management literature because the duration of release is very short (within a few months) relative to the duration of UO2 matrix dissolution, thus it is represented in safety assessment calculations as an instantaneous term. The assumption that IRFGB is also instantaneously released is pessimistic, as discussed in Section 4.2. The relatively rapid release upon exposure to water of a small fraction of the inventory of some dose-relevant radionuclides, in particular 129I, 14C, 36Cl, 135Cs, has significance in safety assessment of spent fuel disposal because of their long half-lives and potential high mobility in groundwater (Hummel 2017). - Fuel matrix dissolution – The largest fraction (> 97%) of the radionuclide inventory (the remaining fission products and activation products as well as the actinides) is released at the same fractional dissolution rate that the fuel matrix dissolves. This process is controlled by the geochemical conditions that exist when the canister is breached as well as the radiation field of the spent fuel.
- Release of activation products from Zircaloy and other metal parts of the fuel assemblies – This involves rapid release of a fraction of the 14C inventory associated with the Zircaloy oxide film, followed by slow release of activation products as a result of corrosion of the various alloys. The corrosion and radionuclide release model for Zircaloy and other metal parts of the fuel assemblies as derived in other reports is provided in Section 4.7.
The specific radionuclide inventories in the spent fuel matrix, Zircaloy cladding and fuel assembly structural materials are given in MIRAM-RBG Database (Nagra 2023a).
Chapter 2 of the report describes the initial state of the spent fuel at the time of its emplacement in disposal canisters. The various types of fuels are described including information on their composition, burn-up, radiation emissions and structure, as these features influence the release of radionuclides. Detailed information on fission gas release (FGR) for UO2 and MOX fuels as well as calculated data on FGR and burn-up for Swiss power reactor fuel are also presented. These data are required because FGR is influenced by fuel power history and because radionuclides that are semi-volatile under fuel operating conditions (e.g., 129I and Cs isotopes) exhibit an aqueous release that is proportional to FGR, as discussed in Chapter 4.
Chapter 3 presents information on the composition of the canister and the encapsulated materials (spent fuel, cladding and other structural materials) and their reactivity. This information is specific to the planned Swiss HLW repository and is the basis for the following thermodynamic modelling of the porewater chemistry on both sides of the canister, i.e. bentonite pore water (BPW) and “inside canister water” (ICW) in the context of the reference evolution scenario of the Swiss HLW repository. Predictions of the geochemical conditions are presented, i.e. pH, Eh, as well as concentrations of main solutes, radionuclides and redox-sensitive species (mainly Fe, H2) that strongly influence the U solubility limit under the strongly reducing conditions expected in the repository. The model calculations are based on the assumption (validated in Chapter 5) that hydrogen evolved by this process is protecting the fuel from radiolytic oxidation at all times after canister failure, which will occur at a time when radiolysis at the spent fuel surface will be weak (10'000 years after waste emplacement or later).
Chapter 4 presents a review of data on short-term aqueous release of various radionuclides. This is followed by a summary of the data that illustrates both the correlation of instant release fraction (IRF) of 129I, 36Cl and Cs isotopes with FGR and the absence of a correlation for some other radionuclides. The term IRF refers to release that is rapid relative to the long timeframe of radiological risk assessment. In practice, this is determined in a duration of a few months under oxic conditions and the IRF contribution is finished when a marked decrease in the cumulative fractional release of nuclides is observed. In recent years there have been many studies published that have provided both a large body of data on the radionuclides that are preferentially released and an understanding of how to estimate the releases from a repository. These data refer to spent fuel having a broad range of burn-up values from light water reactors (LWR), i.e. boiling water reactors (BWR) and pressurized water reactors (PWR). Also discussed are data on rapid release of 14C and 36Cl from fuel as well as the data on rapid release of 79Se, 99Tc and 107Pd. The recommended IRF values for the various fuels are then presented.
Chapter 5 discusses the dissolution of the UO2 matrix, including a review of studies of spent fuel, alpha-doped UO2 and relevant aspects of radiation chemistry. Based on this review, the recommended rate of matrix dissolution is derived.
Chapter 6 presents a literature review on geochemical studies of selected analogues of spent UO2 fuels, namely the Cigar Lake, Oklo and Witwatersrand deposits, focussing on the long-term stability of uraninite and the suppression of radiolytic oxidation by H2. Data on fission product mobility from the Oklo natural reactor are also briefly discussed. The collected evidence is used complementarily to laboratory data to support the long-term validity of the matrix dissolution rates derived in Chapter 5.
Chapter 7 presents the selected reference values for spent fuel dissolution rates and for the prompt release of nuclides upon first contact with water (Instant Release Fractions, or IRF). These values will be used as input in safety assessment calculations. They are based on currently available experimental and theoretical evidence, thoroughly discussed in Chapter 2 and Chapter 5. A short summary of supporting arguments is also included (Section 7.2).