[en] Highlights: • Source term analyses in a PWR of mitigated thermally-induced SGTR scenario performed. • Experimental ARTIST program results on aerosol scrubbing efficiency used in analyses. • Results demonstrate enhanced aerosol retention in a flooded steam generator. • High aerosol retention cannot be predicted by current theoretical scrubbing models. - Abstract: Integral source-term analyses are performed using MELCOR for a PWR Station Blackout (SBO) sequence leading to induced steam generator tube rupture (SGTR). In the absence of any mitigation measures, such a sequence can result in a containment bypass where the radioactive materials can be released directly to the environment. In some SGTR scenarios flooding of the faulted SG secondary side with water can mitigate the accident escalation and also the release of aerosol-borne and volatile radioactive materials. Data on the efficiency of aerosol scrubbing in an SG tube bundle were obtained in the international ARTIST project. In this paper ARTIST data are used directly with parametric MELCOR analyses of a mitigated SGTR sequence to provide more realistic estimates of the releases to environment in such a type of scenario or similar. Comparison is made with predictions using the default scrubbing model in MELCOR, as a representative of the aerosol scrubbing models in current integral codes. Specifically, simulations are performed for an unmitigated sequence and 2 cases where the SG secondary was refilled at different times after the tube rupture. The results, reflecting the experimental observations from ARTIST, demonstrate enhanced aerosol retention in the highly turbulent two-phase flow conditions caused by the complex geometry of the SG secondary side. This effect is not captured by any of the models currently available. The underlying physics remains only partly understood, indicating need for further studies to support a more mechanistic treatment of the retention process.

[en] Highlights: ► A new oxidation model for steam and air with N₂ as a catalyst has been developed at PSI. ► The model captures the initially protective effect of the oxide layer and transition to breakaway. ► The first stage of assessment was performed using detailed time-resolved data from separate-effect tests. ► The tests were performed under controlled isothermal conditions over a range of temperatures and gas compositions. ► The modelling approach allows extension to alternative cladding alloys. - Abstract: Exposure of nuclear fuel rods to air can lead to accelerated oxidation of the cladding, since the effect of nitrogen degrades the oxide layer which hence becomes a less effective barrier to the transport of oxygen to the metallic surface, resulting faster oxidation kinetics. The oxide layer typically becomes porous and can breakaway, a process known as breakaway oxidation. Exposure to air is most likely after a prior period of oxidation in steam. A new model has been developed which captures the initially protective effect of the oxide layer and transition to breakaway. The first stage of assessment was performed using detailed time-resolved data from separate-effect tests performed under controlled isothermal conditions over a range of temperatures and gas compositions. Following implementation into a new version of RELAP5/SCDAPSIM, a second stage of assessment is carried out, namely simulation of an independent integral air ingress transient experiment. This is the subject of Part 2 of this paper. The modelling approach allows extension to alternative cladding alloys such as those recently being deployed in reactor cores.

[en] Highlights: ► A new oxidation model for steam and air with N₂ as a catalyst has been assessed at PSI. ► It recognises breakway transition to linear from parabolic kinetics. ► It was successfully implemented in RELAP5/SCDAPSIM/Mod3.5. ► It was tested against PARAMETER-SF4 experimental data. ► The model successfully calculated the SF4 air ingress scenario. - Abstract: A new model, primarily aimed at air oxidation has been recently developed at PSI, assessed against separate effects tests data and implemented into a developmental version of RELAP5/SCDAPSIM. Oxygen is treated as an active species and nitrogen as a catalyst that promotes breakaway, so that the essential feature of the model is the transition from parabolic to breakaway oxidation. As part of the second stage of the model assessment, analysis is performed of the air ingress bundle experiment PARAMETER-SF4 which comprised pre-oxidation in steam, a period of air ingress, and reflood. The analysis focuses on the thermal response and possible breakaway during, the oxygen consumption. Of particular interest is the effect of oxygen starvation on the oxidation excursion during the subsequent reflood.

[en] Highlights: ► A new oxidation model for steam and air with N2 as a catalyst has been assessed at PSI. ► It recognises breakaway transition to linear from parabolic kinetics. ► The model was assessed against the QUENCH air ingress experiments Q10 and Q16. ► Breakaway prediction and timing are sensitive to the prior oxidation state of the bundle. ► Areas where model improvement is necessary were identified. - Abstract: A recently developed air oxidation model has been implemented in the code version SCDAPSim3.5 and used to analyse the air ingress experiments QUENCH-10 and QUENCH-16 as part of the model assessment. The experiments comprised pre-oxidation in steam, a period of air ingress, and reflood. The analysis focused on the thermal response and possible breakaway during the oxygen consumption. Of particular interest were the effect of the oxygen concentration on the kinetics and the effect of oxygen starvation on the oxidation excursion during the subsequent reflood. The large excursion observed in QUENCH-16 during reflood was not reproduced by the model; the possible causes were explored and indicated the need of further improvement in the model

[en] Highlights: • Accelerated oxidation due to nitrogen. • Steam and oxygen starvation phenomena. • Formation and oxidation of zirconium nitride. • Reflood excursion and quench. • Sensitivity to modelling assumptions - Abstract: A preliminary analysis of the bundle reflood experiment QUENCH-18 is performed with the SCDAPSim/Mod3.5/da code containing the PSI-developed model for oxidation in the presence of air. The simulation follows on from pre-test planning and prediction calculations using the same code and input model. The starting point for the post-test calculations differs from the pre-test only in respect of using the actual boundary conditions. Comparison with measured data enables several aspects of the experiment to be studied. Various treatments of steam and air oxidation kinetics investigate the effect of nitrogen on the oxidation and its continuing influence when air is no longer present. Concerning degradation, different assumptions on failure of the oxide crust indicate how the exposure of relocated metallic melt can enhance the oxidation excursion during reflood. Some modelling and knowledge limitations are identified, particularly regarding oxidation in steam-air mixtures, the roles of nitrogen and zirconium nitride as chemically active species. Several observed features of the facility operation remain unresolved. Simulations suggest that damage to the shroud affected the reflood progression. The bundle may also have been in a highly damaged state, with further impacts on reflooding. Interpretation is therefore provisional, pending more information on the bundle final state. However, the simulation results have significant implications for reactor calculations.

[en] Highlights: • Test with 20 el. heated, 2 unheated, 2 absorber rods under severe accident conditions. • Steam and oxygen starvation conditions. • Zirconium nitride formation and re-oxidation. • Failure of absorber rods with relocation of (Ag, In, Cd) melt and aerosol release. • Strong temperature escalation and oxidation of released cladding melt during reflood. The primary aims of the QUENCH-18 bundle test were to examine the oxidation of M5® claddings in air/steam mixture following a limited pre-oxidation in steam, and to achieve a long period of oxygen and steam starvations to promote interaction with the nitrogen. Additionally, the QUENCH-18 experiment investigated the effects of the presence of two Ag-In-Cd control rods, and two pressurized unheated rod simulators (6 MPa, He). The twenty low-pressurized heater rods (0.23 MPa, similar to the system pressure) were Kr-filled. In a first transient, the bundle was heated in an atmosphere of flowing argon and superheated steam by electrical power increase to the peak cladding temperature of 1400 K. During this heat-up, claddings of the two pressurized rods were burst at temperature of 1045 K. The attainment of 1400 K marked the start of the pre-oxidation stage to achieve a maximum cladding oxide layer thickness of about 80 µm. In the air ingress stage, the steam and argon flows were reduced, and air was injected. The first Ag-In-Cd aerosol release was registered at 1350 K and was dominated by Cd bearing aerosols. Later in the transient, a significant release of Ag was observed. A strong temperature escalation started in the middle of the air ingress stage. During the air ingress stage, a period of oxygen starvation occurred, which was followed by almost complete steam consumption and partial consumption of the nitrogen indicating formation of zirconium nitrides under oxygen starvation conditions. The temperatures continued to increase and stabilized at the melting temperature of Zr bearing materials until water injection. Almost immediately after the start of reflood there was a temperature excursion, leading to maximum measured temperatures of about 2430 K. Final quench was achieved after about 800 s. A significant quantity of hydrogen was generated during the reflood (238 g). Nitrogen release (>54 g) due to re-oxidation of nitrides was also registered. Residual zirconium nitrides were observed in the bundle middle. The metallographic investigations of the bundle show strong cladding oxidation and Zr melt formation. The Zr melt relocated downwards to the lower bundle part was strongly oxidized. Partially oxidized Zr-bearing melt was found down to elevation 160 mm; this elevation was the lowest with evidence of relocated pellet material. At the bundle bottom, only frozen metallic melt containing Zr, Ag, In and Cd was observed between several rods. The experiment exhibited a multiplicity of phenomena for which the data will be invaluable for code assessment and for indicating the direction of model improvements. Example of code application with SCDAPSim is given at the end of this paper.

[en] For over thirty years, IPSN and subsequently IRSN has played a major international role in the field of nuclear power reactor core melt accidents through the undertaking of important experimental programmes (the most significant being the Phebus-FP programme), the development of validated simulation tools (the ASTEC code that is today the leading European tool for modelling severe accidents), and the coordination of the SARNET (Severe Accident Research Network) international network of excellence. These accidents are described as 'severe accidents' because they can lead to radioactive releases outside the plant concerned, with serious consequences for the general public and for the environment. This book compiles the sum of the knowledge acquired on this subject and summarises the lessons that have been learnt from severe accidents around the world for the prevention and reduction of the consequences of such accidents, without addressing those from the Fukushima accident, where knowledge of events is still evolving. The knowledge accumulated by the Institute on these subjects enabled it to play an active role in informing public authorities, the media and the public when this accident occurred, and continues to do so to this day. Following the introduction, which describes the structure of this book and highlights the objectives of R and D on core melt accidents, this book briefly presents the design and operating principles (Chapter 2) and safety principles (Chapter 3) of the reactors currently in operation in France, as well as the main accident scenarios envisaged and studied (Chapter 4). The objective of these chapters is not to provide exhaustive information on these subjects (the reader should refer to the general reference documents listed in the corresponding chapters), but instead to provide the information needed in order to understand, firstly, the general approach adopted in France for preventing and mitigating the consequences of core melt accidents and, secondly, the physical phenomena, studies and analyses described in Chapters 5 to 8. Chapter 5 is devoted to describing the physical phenomena liable to occur during a core melt accident, in the reactor vessel and the reactor containment. It also presents the sequence of events and the methods for mitigating their impact. For each of the subjects covered, a summary of the physical phenomena involved is followed by a description of the past, present and planned experiments designed to study these phenomena, along with their modelling, the validation of which is based on the test results. The chapter then describes the computer codes that couple all of the models and provide the best current state of knowledge of the phenomena. Lastly, this knowledge is reviewed while taking into account the gaps and uncertainties, and the outlook for the future is presented, notably regarding experimental programmes and the development of modelling and numerical simulation tools. Chapter 6 focuses on the behaviour of the containment enclosures during a core melt accident. After summarising the potential leakage paths of radioactive substances through the different containments in the case of the accidents chosen in the design phase, it presents the studies of the mechanical behaviour of the different containments under the loadings that can result from the hazards linked with the phenomena described in Chapter 5. Chapter 6 also discusses the risks of containment building bypass in a core melt accident situation. Chapter 7 presents the lessons learned regarding the phenomenology of core melt accidents and the improvement of nuclear reactor safety. Lastly, Chapter 8 presents a review of development and validation efforts regarding the main computer codes dealing with 'severe accidents', which draw on and build upon the knowledge mainly acquired through the research programmes: ASTEC (IRSN and GRS), MAAP-4 (FAI (US)) and used by EDF and by utilities in many other countries, and MELCOR (SNL (US)) for the US Nuclear Regulatory Commission (US NRC)