Unclassified NEA/CSNI/R(2007)11 Organisation de Coopération et de Développement Economiques Organisation for Economic Co-operation and Development 17-Sep-2007 ___________________________________________________________________________________________

English text only NUCLEAR ENERGY AGENCY COMMITTEE ON THE SAFETY OF NUCLEAR INSTALLATIONS

OECD RESEARCH PROGRAMME ON FUEL-COOLANT INTERACTION STEAM EXPLOSION RESOLUTION FOR NUCLEAR APPLICATIONS - SERENA Final Report - December 2006

JT03232132

Document complet disponible sur OLIS dans son format d'origine Complete document available on OLIS in its original format

NE

A/C

SNI/R

(2007)11 U

nclassified

English text only

NEA/CSNI/R(2007)11

2

ORGANISATION FOR ECONOMIC CO-OPERATION AND DEVELOPMENT

The OECD is a unique forum where the governments of 30 democracies work together to address the economic, social and environmental challenges of globalisation. The OECD is also at the forefront of efforts to understand and to help governments respond to new developments and concerns, such as corporate governance, the information economy and the challenges of an ageing population. The Organisation provides a setting where governments can compare policy experiences, seek answers to common problems, identify good practice and work to co-ordinate domestic and international policies.

The OECD member countries are: Australia, Austria, Belgium, Canada, the Czech Republic, Denmark, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Japan, Korea, Luxembourg, Mexico, the Netherlands, New Zealand, Norway, Poland, Portugal, the Slovak Republic, Spain, Sweden, Switzerland, Turkey, the United Kingdom and the United States. The Commission of the European Communities takes part in the work of the OECD.

OECD Publishing disseminates widely the results of the Organisation’s statistics gathering and research on economic, social and environmental issues, as well as the conventions, guidelines and standards agreed by its members.

* * *

This work is published on the responsibility of the Secretary-General of the OECD. The opinions expressed and arguments employed herein do not necessarily reflect the official views of the Organisation or of the governments of its member countries.

NUCLEAR ENERGY AGENCY

The OECD Nuclear Energy Agency (NEA) was established on 1st February 1958 under the name of the OEEC European Nuclear Energy Agency. It received its present designation on 20th April 1972, when Japan became its first non-European full member. NEA membership today consists of 28 OECD member countries: Australia, Austria, Belgium, Canada, the Czech Republic, Denmark, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Japan, Luxembourg, Mexico, the Netherlands, Norway, Portugal, Republic of Korea, the Slovak Republic, Spain, Sweden, Switzerland, Turkey, the United Kingdom and the United States. The Commission of the European Communities also takes part in the work of the Agency.

The mission of the NEA is: − to assist its member countries in maintaining and further developing, through international co-operation, the scientific,

technological and legal bases required for a safe, environmentally friendly and economical use of nuclear energy for peaceful purposes, as well as

− to provide authoritative assessments and to forge common understandings on key issues, as input to government decisions on nuclear energy policy and to broader OECD policy analyses in areas such as energy and sustainable development.

Specific areas of competence of the NEA include safety and regulation of nuclear activities, radioactive waste management, radiological protection, nuclear science, economic and technical analyses of the nuclear fuel cycle, nuclear law and liability, and public information. The NEA Data Bank provides nuclear data and computer program services for participating countries.

In these and related tasks, the NEA works in close collaboration with the International Atomic Energy Agency in Vienna, with which it has a Co-operation Agreement, as well as with other international organisations in the nuclear field.

© OECD 2007 No reproduction, copy, transmission or translation of this publication may be made without written permission. Applications

should be sent to OECD Publishing: rights@oecd.org or by fax (+33-1) 45 24 99 30. Permission to photocopy a portion of this work should be addressed to the Centre Français d’exploitation du droit de Copie (CFC), 20 rue des Grands-Augustins, 75006 Paris, France, fax (+33-1) 46 34 67 19, (contact@cfcopies.com) or (for US only) to Copyright Clearance Center (CCC), 222 Rosewood Drive Danvers, MA 01923, USA, fax +1 978 646 8600, info@copyright.com.

NEA/CSNI/R(2007)11

3

COMMITTEE ON THE SAFETY OF NUCLEAR INSTALLATIONS

The NEA Committee on the Safety of Nuclear Installations (CSNI) is an international committee made of senior scientists and engineers, with broad responsibilities for safety technology and research programmes, and representatives from regulatory authorities. It was set up in 1973 to develop and co-ordinate the activities of the NEA concerning the technical aspects of the design, construction and operation of nuclear installations insofar as they affect the safety of such installations.

The committee’s purpose is to foster international co-operation in nuclear safety amongst the OECD member countries. The CSNI’s main tasks are to exchange technical information and to promote collaboration between research, development, engineering and regulatory organisations; to review operating experience and the state of knowledge on selected topics of nuclear safety technology and safety assessment; to initiate and conduct programmes to overcome discrepancies, develop improvements and research consensus on technical issues; to promote the co-ordination of work that serve maintaining competence in the nuclear safety matters, including the establishment of joint undertakings.

The committee shall focus primarily on existing power reactors and other nuclear installations; it shall also consider the safety implications of scientific and technical developments of new reactor designs.

In implementing its programme, the CSNI establishes co-operate mechanisms with NEA’s Committee on Nuclear Regulatory Activities (CNRA) responsible for the programme of the Agency concerning the regulation, licensing and inspection of nuclear installations with regard to safety. It also co-operates with NEA’s Committee on Radiation Protection and Public Health (CRPPH), NEA’s Radioactive Waste Management Committee (RWMC) and the NEA’s Nuclear Science Committee (NSC) on matters of common interest.

NEA/CSNI/R(2007)11

NEA/CSNI/R(2007)11

5

EXECUTIVE SUMMARY

Background

SERENA (Steam Explosion REsolution for Nuclear Applications) is a programme for the resolution of FCI remaining issues for LWRs.

The SERENA programme stems from concerns expressed by the Senior Group of Experts on Nuclear Safety Research Facilities and Programmes (SESAR/FAP) about de-emphasis of FCI research all over the world, while uncertainties still exist on some aspects of FCI. A restricted exploratory meeting involving FCI and reactor safety experts was held subsequently to discuss remaining research needs and advise CSNI regarding future steps in this area. Resolved the apha-mode failure from the risk perspective, the most important steam explosion remaining safety issues are failure of the lower head and damage to the cavity, with potential consequences for containment integrity. The group recognised that these issues are still partly unresolved and recommended that an international coordinated and well-controlled research project be carried out under the auspices of the CSNI to address the following aspects: pursuit of key fundamental understanding, in particular for judging the resistance of real corium material to explosions, realisation of fully verified codes for the purpose, realisation of special-purpose, separate-effects tests and finally, carrying out well-defined confirmatory testing.

CSNI recommended that reactor application be the focus of any new action in this area, and that prior to starting new research, existing knowledge base and analytical tools be carefully assessed to determine what would be strictly needed to bring predictability of FCI energetics to required levels for risk evaluation. A group of international FCI experts elaborated a programme in that sense, which was endorsed by CSNI in December 2001 and started January 2002. The programme had duration 3.5 years.

Objective of the work

The objective of the work was:

• Evaluating the capabilities of the current generation of FCI computer codes in predicting steam explosion induced loads,

• Identifying key FCI phenomena and associated uncertainties impacting predictability of steam explosion energetics in reactor situations,

• Proposing confirmatory research to reduce these uncertainties to acceptable levels for steam explosion risk assessment.

The objective was reached through comparative calculations of existing experiments and most relevant generic reactor situations by current FCI codes.

The following FCI codes participated in the exercise: ESPROSE-m, IDEMO, IFCI, IKEMIX, JASMINE, MATTINA, MC3D, PM-ALPHA, TEXAS-V, TRACER, VAPEX and VESUVIUS.

NEA/CSNI/R(2007)11

6

Method of working

First, generic situations corresponding to plausible melt relocation scenarios and having potential to produce damaging steam explosion were identified. At the same time, existing experiments which best address these conditions for both the pre-mixing and the explosion phases were selected.

Then, the selected pre-mixing and explosion tests were calculated and the results compared. This had the twofold objective of establishing a "setting to zero" of the codes and each participant starting with verification of their tools on a similar basis. It allowed partners verifying which parameters and model options were best suited to capture the essential features of experiments performed in so-called "realistic conditions", determining whether parameters and model options had to be adapted for reactor calculations, and discussing the impact of the various approaches on the results. Integrating information coming from this variety of hypotheses and backgrounds allowed identifying the common areas where uncertainties impacted the explosion results.

Reactor calculations had the scope of establishing whether the limitations, deficiencies and differences observed in the simulations of the experiments were important for FCI situations of interest for reactors, and identifying the key physics for these reactor situations. This was achieved by proper evaluation of codes based on the “fitness for purpose” concept. The “purpose” is to provide reasonably conservative estimates of energetics under conceivable modes of contact between molten core (fuel) and water (coolant) during key relocation events, both in- and ex-vessel, in representative melt progression sequences during severe accidents. The “fitness” is to express that such estimates are robust, i.e., energetics as represented by the calculated pressure-time history and loads on structures, are bounded. By definition, this task involved comparison of code predictions with a defined purpose, i.e., to determine what each code is capable of doing with regard to prediction of explosion loads on structures.

As a last step, a plan to move forward was proposed, with the aim to provide the confirmatory data that would allow codes reaching the required level of confidence for predicting explosion loads on reactor structures.

Calculations were accomplished by providing participants with same sets of initial conditions and reference data. They translated these initial conditions into adequate inputs for their codes. Participants were let free to set model options and parameters as they used to when doing code application for their own needs. However, they were asked to provide at least one calculation with using standard parameters, to document their choices and possibly make sensitivity calculations. Explosion phases of the experiments were calculated both for imposed and calculated pre-mixing whenever required. Comparison was made on a set of pre-established quantities, either for codes-to-data comparison or code-to-code comparison. Most important quantities were nodalisation, pressure and impulse, vaporisation/condensation rates, energy release, component fraction, debris characteristics.

Summary of results

Relevant reactor situations and experiments

Steam explosion occurring during multiple-jet discharge through the core support plate was considered as potentially the most challenging for the primary circuit. Steam explosion occurring during melt discharge through a large breach in the lower head into a flooded cavity was considered of most concern ex-vessel FCI, and a side pour the most challenging case for the structures. Therefore, work focussed at verifying the capabilities of the codes to predict these situations.. Noting that no relevant multi-pour experiment exists, the best that can be extracted today from existing experimental data base in relation with these reactor situations was found in the FARO/KROTOS (JRC-Ispra) and TROI (KAERI) programmes. Calculations were performed for FARO L-28 and FARO L-31 for pre-mixing, and KROTOS 44, TROI-13 and FARO L-33 for explosion, respectively.

NEA/CSNI/R(2007)11

7

Experiment calculations

The strength of a steam explosion is greatly dependant on the characteristics of the melt-water-steam mixture in which it propagates. It is therefore of paramount importance to verify first the code capabilities to describe this pre-mixing. When applied to the selected premixing experiments, most of the codes significantly overestimated void (steam and possibly non-condensable gases) fraction with respect to the data. Knowing the importance of void in pre-mixing for steam explosion behaviour, attention was largely focused on this issue. Several sensitivity calculations were done in addition to the base cases to identify the reasons for such differences. The break-up mechanisms were also investigated, particularly in the first phase of the premixing up to melt front-bottom contact, where spontaneous steam explosions often trigger in the experiments. However, resolution of this issue was not possible due to the variety of approaches and points of view that could not be resolved in the absence of detailed data on spatial distribution of corium and void in the existing experiments.

For the simulation of the explosion experiments, it was observed that FCI codes tend to noticeably overestimate the energetics for the tests performed with corium, while they predict reasonable behaviour for the test performed with alumina melt (KROTOS 44) . Noticing that the codes were tuned to the KROTOS-alumina data in their developmental stage, expert’s discussions largely focused on the material influence on the energetics, in view of the recent observations in TROI, complementing those already obtained with KROTOS-corium, that corium melts produce relatively mild steam explosions with respect to alumina melt. This aspect was judged to be one of the major uncertainties for predictions, in addition to the uncertainties on pre-mixing characteristics, as the characteristics of corium having low explosivity could not be generalised in the absence of data for a sufficiently large spectrum of prototypical corium compositions.

These issues, identified as the result of calculating the experiments, were not considered as critical for the purpose of evaluating the loads in reactor situations until verified for those situations.

Reactor calculations

Two standard configurations reflecting generic situations of most concern were selected for analysis by the computer codes, one for in-vessel steam explosion consisting of a collection of corium jets falling at once in the flooded lower head, and one for ex-vessel steam explosion consisting of a single large size jet falling in a flooded cavity.

For the in-vessel case, the results showed a rather reduced scatter of the predictions of the impulses despite the variety of the modelling approaches, ranging from some tens of kPa.s to ~200 kPa.s. It was recognised that such levels of loads would not challenge the integrity of the vessel in the absence of pre-existing thermal loads. The reasons for the low level of the loads stands essentially in the high voiding of the pre-mixing region, the reduced quantity of molten corium that can be mixed with water at whatever time the explosion triggers, the venting possibilities that exist in a large geometry. But it was also recognised that the calculated situation might not be the worst possible and that significant uncertainties affect void level and role. Nevertheless, there was a broad consensus between the partners that performing further research aiming at resolving these issues for in-vessel FCI was not a first priority.

For the ex-vessel case, all the predicted impulses were between 20 and 100 kPa.s. It was recognised that a level of loads of the order of some tens of kPa.s may induce some damage to the cavity, and in turn impact containment integrity. But, the large scatter of the predictions reflecting a lack of understanding of the mixing and explosive behaviour of prototypical melts, makes the level of potential damage to the cavity unpredictable. The partners agreed that research aiming at better quantification of safety margins for the containment and a more consistent prediction of ex-vessel FCI loads was required.

NEA/CSNI/R(2007)11

8

Conclusions and recommendations

One positive outcome of the programme is that whatever the modelling and numerical approaches all the codes were able to calculate the reactor situations of concern.

Another positive outcome is that, despite the different choices for setting the code parameters and options for the reactor applications, all the calculated loads were relatively low. Concerning in-vessel steam explosion the level of the loads was sufficiently below the capacity of the vessel to allow concluding that they would not induce damage to the lower head (in the absence of pre-existing thermal loads). For ex-vessel steam explosion all the calculated loads, even low, were above the capacity of a typical cavity walls. The scatter of the results raises the problem of the quantification of the safety margin for the containment in case of ex-vessel steam explosion.

The scattering of the predictions reflects the lack of understanding of prototypical corium steam explosion behaviour. The major sources of uncertainties concern the characteristics of the flow regime in the pre-mixing phase, especially void behaviour, and of the fragmentation of corium melts in the explosion phase (material effect). Convergence towards more consistent predictions requires resolving these uncertainties. Another issue pointed out by the partners is that extrapolation capabilities of the codes to multi-dimensional configurations suffered a lack of verification data base.

In line with these conclusions and recognising that:

1. Any new experimental assessment should focus on ex-vessel UO2-based corium,

2. A limited number of corium compositions have been used so far in steam explosion experiments (in particular, no data are available for corium having large liquidus-solidus intervals as expected in most accident scenarios),

3. The global measurements performed so far in any corium test do not allow to obtain the detailed data required to resolve the ”void issue”,

4. Both intrinsic FCI behaviour of corium jets and reactor-related FCI behaviour of corium pours should be assessed,

The group thinks that performing a limited number of well-designed well-instrumented tests using a minimum set of combinations of ex-vessel corium melts in different geometrical configurations, could provide sufficient data for the purpose. Instrumentation and analysis capabilities of the experimental programme should be such as to provide data to the required level of details to characterise the pre-mixing phase and the role of melt properties. The ultimate purpose of the tests being to improve and validate models and methods in order to be able to predict loads reliably and consistently, analytical work with this aim should be conducted in parallel.

Altogether, such a programme would make a decisive step in converging towards common understanding and thus, resolution of the steam explosion issue. In particular, confirmation for an extended range of corium melts that the void fractions in the mixing area and fragmentation characteristics are actually as those predicted and used in the present calculations, would confirm the “mild” character of explosion with reactor melts and significantly contribute to increasing the safety margins in relation to present and future SA management strategies.

The group thinks that using the complementary features of existing KROTOS (CEA) and TROI (KAERI) corium facilities with the help of their improved instrumentation for pre-mixing, namely, high energy X-ray radioscopy in KROTOS and electrical tomography in TROI, would provide the required data. KROTOS is more suited for investigating the intrinsic FCI characteristics of accident-scenario-related prototypical corium melts in a one-dimensional geometry, while TROI is more suited for testing the FCI

NEA/CSNI/R(2007)11

9

behaviour of these materials in reactor-like conditions by having more mass and multi-dimensional melt-water interaction geometry. Once the code models are validated against the KROTOS data, their capability to handle more reactor-like situations can be validated against the TROI test data.

The group recommends to undertake such a programme as the required complement of the present action in order to reach the objective of having codes sufficiently validated for evaluation of the risk of containment failure associated with ex-vessel steam explosions.

NEA/CSNI/R(2007)11

NEA/CSNI/R(2007)11

11

TABLE OF CONTENTS

Executive summary..................................................................................................................................... 5

1. Introduction ......................................................................................................................................... 15

2. Participating organisations................................................................................................................... 17

3. Participating FCI codes ....................................................................................................................... 19

4. Methodology........................................................................................................................................ 21

5. Relevant conditions for energetic FCI in NPP’s (Task 1) ................................................................... 23 5.1 Preliminary remarks ..................................................................................................................... 23 5.2 Generic situation........................................................................................................................... 23 5.3 In-vessel situations ....................................................................................................................... 23 5.4 Ex-vessel situations ...................................................................................................................... 24 5.5 Relevant experiments ................................................................................................................... 25

6. Calculation of pre-mixing experiments (Task 2)................................................................................. 27 6.1 Selected experiments .................................................................................................................... 27 6.2 Calculation conditions for pre-mixing experiments ..................................................................... 27 6.3 Comparison of pre-mixing calculation results.............................................................................. 27 6.4 Conclusion on calculation of pre-mixing experiments................................................................. 28

7. Calculation of explosion experiments (Task 3) ................................................................................... 29 7.1 Selected explosion experiments ................................................................................................... 29 7.2 Calculation conditions for explosion experiments ....................................................................... 29 7.3 Comparison of explosion calculation results................................................................................ 29

7.3.1 KROTOS K-44 ................................................................................................................. 29 7.3.2 TROI-13 and FARO L-33................................................................................................. 30

7.4 Conclusion on calculation of explosion experiments ................................................................... 30

8. Reactor calculations (Task 4) .............................................................................................................. 33 8.1 Selected geometry’s and conditions ............................................................................................. 33 8.2 Calculation conditions.................................................................................................................. 33 8.3 Comparison of the results............................................................................................................. 33

8.3.1 In-vessel case .................................................................................................................... 33 8.3.2 Ex-vessel case ................................................................................................................... 35

8.4 Discussion of reactor calculations and routes for confirmatory work.......................................... 36

9. General conclusion of phase 1............................................................................................................. 39

10. References ........................................................................................................................................... 41

List of SERA contributors .......................................................................................................................... 43

NEA/CSNI/R(2007)11

12

LIST OF TABLES

Table 1. SERENA Phase 1 time table ................................................................................................. 45 Table 2. Codes used by partners in Phase 1 ........................................................................................ 45 Table 3. SERENA partners’ interests for melt-into-water situations .................................................. 45 Table 4. Summary of melt-into-water most relevant reactor conditions............................................. 46 Table 5. Summary of conditions of selected FARO tests for pre-mixing calculation......................... 47 Table 6. Jet representation and break-up conditions considered

in the FARO tests simulation of the pre-mixing phase ......................................................... 48 Table 7. Conditions of pre-selected explosion tests ............................................................................ 49 Table 8. Summary of conditions used for the calculations of KROTOS-44....................................... 50 Table 9. Summary of conditions used for the calculations of FARO L-33......................................... 51 Table 10. Modelling/description of the trigger for KROTOS-44.......................................................... 52 Table 11. Modelling/description of the trigger for FARO L-33............................................................ 53 Table 12. Pre-mixing parameters used in reactor calculations with respect

to those used in the pre-mixing experiment calculations (1/2).............................................. 54 Table 13. Pre-mixing and explosion parameters used in reactor calculations

with respect to those used in explosion experiment calculation (1/5)................................... 56 Table 14. In-vessel case. Jet description and melt masses in the system at trigger............................... 61 Table 15. Ex-vessel case. Melt masses in the system at trigger ............................................................ 62 Table 16. Impact of using different assumptions on melt break-up on void and

melt distribution in pre-mixing phase ................................................................................... 63

NEA/CSNI/R(2007)11

13

LIST OF FIGURES

Figure 1. Schematic view of FARO test arrangement and initial component distribution in FARO L-28..................................................................... 65

Figure 2. Geometry and nodalisation used for FARO L-28 simulation ............................................. 66 Figure 3. Comparison of calculated vessel pressure and energy release

with data (bold curves) for FARO L-28.............................................................................. 69 Figure 4. Comparison of calculated vessel pressure and energy release

with data (bold curves) for FARO L-33.............................................................................. 71 Figure 5. Comparison of radially-averaged void fraction calculated by the codes

with global value calculated from experimental level swell (bold lines) for FARO L-28 at time 1.0 s and time 4.0 s, respectively................................................... 72

Figure 6. Comparison of global void fraction calculated by the codes and experimental value ........ 73 Figure 7. KROTOS-44 explosion pressures measured in water at six different elevations

at test section wall ............................................................................................................... 74 Figure 8. TROI-13 explosion pressures measured in water at two different elevations

at test section wall ............................................................................................................... 75 Figure 9. FARO L-33 explosion pressures measured in water at five different elevations

at test section wall ............................................................................................................... 76 Figure 10. Imposed pre-mixing geometry at trigger time for KROTOS-44 simulation ....................... 77 Figure 11. Comparison of dynamic pressure and corresponding impulse calculated

by the codes with experimental value (blue bold lines) at mid-height in the water pool for KROTOS-44...................................................................................... 78

Figure 12. Pressures and impulses in water with MC3D 3.4.02 (CEA-IRSN)..................................... 79 Figure 13. Pressures and impulses in water with IDEMO-2D (IKE) ................................................... 80 Figure 14. Pressures and impulses in water with JASMINE 3.0 (JAERI) ........................................... 81 Figure 15. Pressures and impulses in water with TEXAS-V (KAERI-KMU and UW)....................... 82 Figure 16. Pressures and impulses in water with IFCI 6.0 (KINS) ...................................................... 83 Figure 17. Pressure in water with VESUVIUS (NUPEC).................................................................... 84 Figure 18. Pressures and impulses in water with ESPROSE.m (UCSB) ............................................. 85 Figure 19. KROTOS-44 predicted histories of kinetic energy

(experimental evaluation is 0.153 MJ) ................................................................................ 86 Figure 20. Comparison of dynamic pressure and corresponding impulse calculated

by the codes with experimental value (blue bold lines) for TROI-13................................. 87 Figure 21. TROI-13 predicted histories of kinetic energy.................................................................... 88 Figure 22. Comparison of dynamic pressure and corresponding impulse calculated

by the codes with experimental value (bold lines) at level 1 390 mm in the water pool for FARO L-33 (highest value measured). Full calculations premixing + explosion ............................................................................. 89

Figure 23. Comparison of dynamic pressure and corresponding impulse calculated by the codes with experimental value (bold lines) at level 1 390 mm in the water pool for FARO L-33 (highest value measured). Common pre-mixing conditions.......................................................................................... 90

Figure 24. Pressures and impulses in water with MC3D 3.4.0.2 (CEA-IRSN).................................... 91

NEA/CSNI/R(2007)11

14

Figure 25. Pressures and impulses in water with IDEMO-2D (IKE) ................................................... 92 Figure 26. Pressures and impulses in water with JASMINE 3.0 (JAERI) ........................................... 93 Figure 27. Pressures and impulses in water with TEXAS-V (KAERI-KMU and UW)....................... 94 Figure 28. Pressures and impulses in water with IFCI 6.0 (KINS) ...................................................... 95 Figure 29. Pressures and impulses in water with VESUVIUS (NUPEC) ............................................ 96 Figure 30. Pressures and impulses in water with ESPROSE.m (UCSB) ............................................. 97 Figure 31. Result of calculations for FARO L-33 with IDEMO98

assuming difference vapour volume fractions in the mixing zone...................................... 98 Figure 32. FARO L-33 predicted histories of kinetic energy – full calculations

(experimental evaluation is 0.110 MJ) ................................................................................ 99 Figure 33. Calculated reactor situations ............................................................................................... 100 Figure 34. Geometry and nodalisation used for in-vessel case simulation (not to scale) ..................... 101 Figure 35. Maximum pressure calculated at vessel bottom (position indicated into brackets) ............ 104 Figure 36. Maximum impulse calculated at vessel bottom (position indicated into brackets) ............. 105 Figure 37. In-vessel axial distrubtions of radially averaged void fraction at trigger time

(radial averages made over a diameter of 1 m) ................................................................... 106 Figure 38. In-vessel global component fractions at melt-bottom contact averaged

over a cylinder of height the water depth, and diameter 1 m. Ordinate: component fraction in %; Abscissa codes........................................................... 107

Figure 39. In-vessel calculated impulses as a function of global void fraction at trigger time ............ 108 Figure 40. In-vessel evolution of fragmented melt mass during explosion.......................................... 109 Figure 41. Geometry and nodalisation used for ex-vessel case simulation (not to scale) .................... 111 Figure 42. Ex-vessel maximum pressure calculated at cavity bottom

(position indicated into brackets) ........................................................................................ 114 Figure 43. Ex-vessel maximum impulse calculated at cavity bottom

(position indicated into brackets) ........................................................................................ 116 Figure 44. Ex-vessel maximum pressure calculated at cavity wall

(position indicated into brackets) ........................................................................................ 118 Figure 45. Ex-vessel maximum impulse calculated at cavity wall

(position indicated into brackets) ........................................................................................ 120 Figure 46. Ex-vessel axial distributions of radially averaged void fraction at trigger time

(radial averages made over a diameter of 1 m) ................................................................... 122 Figure 47. Ex-vessel axial distributions of radially averaged melt fraction at trigger time

(radial averages made over a diameter of 1 m) ................................................................... 123 Figure 48. Ex-vessel global component fractions at melt-bottom contact averaged

over a cylinder of height the water depth, and diameter 1 m. Ordinate: component fraction in %; Abscissa codes........................................................... 124

Figure 49. Ex-vessel calculated impulses as a function of global void fraction at trigger time ........... 125 Figure 50. Ex-vessel fragmented melt mass during explosion ............................................................. 126

NEA/CSNI/R(2007)11

15

1. INTRODUCTION

SERENA (Steam Explosion REsolution for Nuclear Applications) is an international OECD programme for the resolution of FCI remaining issues in LWRs.

The SERENA programme stems from concerns expressed by the Senior Group of Experts on Nuclear Safety Research Facilities and Programmes (SESAR/FAP) of the Committee on the Safety of Nuclear Installations (CSNI) of Nuclear Energy Agency (NEA) of the Organisation for Economic Co-operation and Development (OECD). The group was worried by current de-emphasis of FCI research all over the world, while uncertainties still exist on some aspects of FCI (Nuclear Safety Research in OECD Countries, 2001, [1]).

An evaluation of remaining unresolved FCI issues and research needs was performed in a restricted exploratory meeting (held on 30-31 October 2000 at NEA Offices in Paris) involving FCI and reactor safety experts. A statement to CSNI was produced by the meeting [2] and discussed at the subsequent CSNI meeting. CSNI recommended to carefully analyse the existing knowledge base of FCI before re-opening investigations on fundamentals of steam explosion or carrying out new experimental programmes. In particular, CSNI recommended that the existing analytical tools be assessed to determine what would be strictly needed to strengthen reliability of estimates of the magnitude of FCI induced loadings in reactor conditions. Generally speaking, it was requested to establish a status of FCI understanding for reactor situations.

This led to defining a programme in two phases: a purely analytical phase to assess the predictive capabilities of existing tools for reactor cases (Phase 1), possibly followed by a phase to resolve the remaining uncertainties through performing complementary analytical work and confirmatory testing (Phase 2). A proposal for Phase 1 was matured during 2001 by a group of FCI experts and approved by CSNI in December 2001 [3,4].

It consisted of comparative calculations by available tools of selected existing experiments and reactor cases, in order to identify those areas where lack of understanding induced large uncertainties in the predictions of the loads in reactors. It was the first comparative exercise undertaken since ISP-39, which however concerned premixing only [5]. Phase 1 was divided into five tasks:

1. Identification of relevant conditions for FCI in reactor and selection of existing experiments which best address these conditions for both the pre-mixing and the explosion phases.

2. Comparative calculations of relevant pre-mixing experiments. 3. Comparative calculations of relevant explosion experiments. 4. Reactor applications. 5. Synthesis and recommendations for further work.

Work started January 2002 in the form of a coordinated action where each partner financed his own work and had duration 3.5 years until mid-2005 (Table 1). Comparisons and analyses were performed through e-mails exchanges and five meetings, one by task. A Technical Programme Committee assisted the Programme Coordinator and provided wide technical and international support to the action.

The present report makes the synthesis of the findings of Phase 1.

NEA/CSNI/R(2007)11

NEA/CSNI/R(2007)11

17

2. PARTICIPATING ORGANISATIONS

Participating organisations in Phase 1 were:

• IRSN (Institut de radioprotection et sûreté nucléaires) and CEA (Commissariat à l’énergie atomique), France (joint participation).

• FZK (ForschungsZentrum Karlsruhe), Germany. • IKE (Institut für Kernenergetik und Energiesysteme der Universität Stuttgart), Germany. • JAERI (Japan Atomic Energy Research Institute), Japan. • NUPEC (Nuclear Power Engineering Corporation), Japan.a • KAERI (Korea Atomic Energy Research Institute) and KMU (Korea Maritime University), Korea

(joint participation). • KINS (Korea Institute of Nuclear Energy), Korea. • NRC (Nuclear Regulatory Commission), United States of America.b • VTT (VTT Technical Research Centre), Finland.c • EREC (Electrogorsk Research and Engineering Centre on NPP Safety), Russian Federation.

The coordination was provided by CEA and the coordination costs supported jointly by CEA and IRSN.

a. NUPEC joined the project after completion of Task 1. b. Calculations entrusted to University of California, Santa Barbara (UCSB) and University of Wisconsin, Madison

(UW). c. VTT withdrew after completion of Task 1.

NEA/CSNI/R(2007)11

NEA/CSNI/R(2007)11

19

3. PARTICIPATING FCI CODES

Calculations were performed by using ESPROSE-m, IDEMO, IFCI, IKEMIX, JASMINE, MATTINA, MC3D, PM-ALPHA, TEXAS-V, TRACER, VAPEX and VESUVIUS codes. Some codes were used by different partners, see Table 2. All codes were applied 2-D axi-symmetric, except TEXAS (1-D) and PM-ALPHA (3-D). ESPROSE.m and IDEMO calculated the explosion phase only, starting from pre-mixing conditions calculated by the pre-mixing codes PM-ALPHA and IKEMIX, respectively. MATTINA has no typical explosion model and calculated the explosion phase by extending the validity of the pre-mixing heat transfer laws to the explosion time-scale and fragmentation characteristics. A comparative review of the codes is found in [9].

As being a status of the code capabilities to calculate FCI in reactor situations, it was not within the scope of SERENA to establish whether or not a specific code was qualified to be in. It was let to the responsibility of each partner to judge whether his code or the code he is using for FCI safety assessment had the required degree of qualification and verification to participate.

NEA/CSNI/R(2007)11

NEA/CSNI/R(2007)11

21

4. METHODOLOGY

First, generic situations corresponding to plausible melt relocation scenarios and capable to produce potentially damaging steam explosion were identified, and existing experiments which best address these conditions for both the pre-mixing and the explosion phases selected (Task 1).

Then, the selected pre-mixing and explosion tests were calculated and the results compared (Tasks 2 and 3). This had the twofold objective of establishing a “setting to zero” of the codes and each participant starting with verification of their tools on a similar basis (half of the partners had not full access to the data of the experiments – mainly coming form KROTOS and FARO experimental programmes – prior to SERENA. It allowed partners verifying which parameters and model options were best suited to capture the essential features of experiments performed in so-called “realistic conditions”, determining whether parameters and model options had to be adapted for reactor calculations, and discussing the impact of the various approaches on the results. Integrating information coming from this variety of hypotheses and backgrounds allowed identifying the common areas where uncertainties impacted the explosion results. Confirmation that those uncertainties were consequential to the predictability of the loads for reactors situations was assessed through calculations of selected reactor cases (Task 4).

Task 4 calculations allowed establishing whether the limitations, deficiencies and differences observed in the simulations of the experiments (in Tasks 2 and 3) were important for FCI situations of interest for reactors, and identifying the key physics for these reactor situations. This was achieved by proper evaluation of codes based on the “fitness for purpose” concept. The “purpose” is to provide reasonably conservative estimates of energetics under conceivable modes of contact between molten core (fuel) and water (coolant) during key relocation events, both in- and ex-vessel, in representative melt progression sequences during severe accidents. The “fitness” is to express that such estimates are robust, i.e., energetics as represented by the calculated pressure-time history and loads on structures, are bounded. By definition, this task involved comparison of code predictions with a defined purpose, i.e., to determine what each code is capable of doing with regard to prediction of explosion loads on structures.

Task 4 had the scope of delivering clear answer to the following questions:

• For both the pre-mixing and explosion phases, are the limitations and deficiencies of the various approaches observed for calculating the experiments relevant for reactor conditions?

• For both pre-mixing and explosion phases, are the differences of the various approaches important for calculating the loads resulting from explosions in reactors?

• What are the uncertainties on the results and are they acceptable? • What are the principal phenomena that may need further investigation to reduce the uncertainties if

they are not acceptable?

Tasks 2 to 4 were accomplished by providing participants with same sets of initial conditions and reference data. They translated these initial conditions into adequate inputs for their codes. Participants were let free to set model options and parameters as they used to. However, they were asked to provide at least one calculation with using standard parameters, to document their choices and possibly make sensitivity calculations. Explosion phases of the experiments were calculated both for imposed and calculated pre-

NEA/CSNI/R(2007)11

22

mixing whenever required. Comparison was made on a set of pre-established quantities, either for codes-to-data comparison or code-to-code comparison. These quantities included nodalisation, pressure and impulse, vaporization/condensation rates, energy release, component fraction, debris characteristics.

Analysis of the results consisted of:

• For each partner ̶ Discussing the reasons of the discrepancies of his results with the data, possibly with the help of

sensitivity calculations, ̶ Identifying which particular process(es) has(have) major impact on the predictions.

• For the group ̶ Same as above by cross-checking the individual analyses.

Task 5 was synthesising the findings of Phase 1, which is the subject of the present report.

NEA/CSNI/R(2007)11

23

5. RELEVANT CONDITIONS FOR ENERGETIC FCI IN NPP’S (TASK 1)

5.1 Preliminary remarks

Even though SERENA is an unprecedented coordinated worldwide effort for converging towards a common view of important FCI phenomena for reactor situations, the limited resources allocated to Phase 1 (approximately 1.2 person-year by organisation) did not make possible to cover all the spectrum of relevant situations. One way to overcome this difficulty was to research those conditions that bound most situations, independently of reactor types and designs.

However, finding bounding cases is far from being straightforward considering the variety of reactor designs and accident scenarios, and the large uncertainties that affect FCI understanding, which is the very reason why SERENA is undertaken. At the best, one could extrapolate from present knowledge by using one or another method (evaluation with existing tools in their present state of validation, expert judgement…), and the confirmation that the anticipated choices were right can come only at the end of the process, and possibly require some iterations.

When asked to identify the range of conditions which may result in an energetic FCI with potential damage to the reactor structures according to their own experience and/or peculiarities of running plants in their respective country, most partners used, either implicitly or explicitly, a similar approach, sometime complemented by probabilistic assessment.

5.2 Generic situation

There is, however, a generic situation that was considered by all participants as a bounding case for FCI in reactor conditions, whatever the LWR reactor type and design, and this situation is steam explosion occurring sometime during the penetration of a large amount of core melt into a pool of water.

There is one exception which concerns steam explosion occurring during reflooding a molten debris in a weakened lower head (sometimes referred to as “stratified steam explosion”), especially in relation to the in-vessel retention or IVR strategy. However, this case was not considered as a major FCI issue by most of the partners, and the melt-into-water case more critical anyway.

Consequently, SERENA addressed melt-into-water situations only, and the objective of the programme was considered reached if reached for this type of situation.

Table 3 summarises partners’ interest for melt-into-water situations as a function of reactor types and either in- or ex-vessel FCIs, and Table 4 lists the most relevant conditions for specific reactor types and safety requirements. A generic situation of most concern for both in- and ex-vessel situations was extracted from those tables.

5.3 In-vessel situations

There is a consensus among the experts that research on FCI has been successful in resolving the alpha-mode of containment failure (early containment failure resulting from an in-vessel steam explosion) from the risk perspective [6].

NEA/CSNI/R(2007)11

24

Considering that, in-vessel FCI remains of concern only with respect to lower head failure mode and consequences it may have on primary circuit (essentially damage on piping and steam generator) and debris distribution. It becomes of major concern when accident management strategy is based on in-vessel retention concept.

When integrating Table 4 data, gravity pours of UO2-ZrO2 melts in the lower head in a ~2-m-deep pool of saturated water at moderate system pressure are considered of most concern for in-vessel FCI among the plausible scenarios of in-vessel melt progression. One such situation was calculated in Task 4 (see Section 8).

Both single and multi-pours are of concern. Typically, a multiple-pour corresponds to a situation where a large breach forms in the crust of the in-core molten debris, giving sufficient melt flow to feed simultaneously several tens of core support plate holes (diameter of the holes of the order of 0.10 m). It is generally considered that the water in the lower head arrives up to the lower plate at the most.

In contrast with ex-vessel situations, in-vessel single pours are believed to be possible only through holes having a diameter of the same order than for multi-jet case. Thus, multi-jet configurations are considered by some experts to be potentially more energetic than single pours. It is not clear however whether this is the result of evaluations using available FCI tools or based on other considerations.

The reality is certainly not so straightforward and, depending on the distribution of the jets, some counter effects such as the production of highly voided regions inside the array of jets (if not completely voided), may decrease dramatically the potential for a very strong explosion.

In both cases, either single or multi-jets, jets arriving off-centre in the lower head are considered the most challenging for the vessel. Note, however, that side pours such as those through the downcomer, are not considered an issue with respect to steam explosion because of the limited amounts of melt which can be mixed with water at once.

5.4 Ex-vessel situations

From Table 3, it is seen that ex-vessel situations are of interest for all the partners and of major concern for most of them. The main reasons are one or more of the following:

• Failure of lower head in case of core melt down cannot be ruled out. • Absence of water in the cavity cannot be guaranteed at time of lower head failure. • Flooding the cavity in case of expected vessel failure is already part of some SAM strategies. • Use of water to cool down the debris in the cavity could be generalised if it is demonstrated that

energetic FCI has no consequences on the containment integrity.

When integrating Table 4 data, unconstrained large pours equivalent to some tens of cm in diameter of UO2-ZrO2-Zr-Steel melts in a deep pool of subcooled water at low system pressure and moderate driven pressure are of most concern for ex-vessel energetic FCI among the plausible scenarios of melt transfer into a flooded cavity. One such situation was calculated in Task 4 (see section 8).

Two categories of conditions can be established:

• Single pour of some tens of centimetres in diameter. • Multiple jets, each jet having a diameter of the order of 10 cm (in the case of penetrations in the lower

head).

NEA/CSNI/R(2007)11

25

In the first category, a side pour is believed to be most challenging for the cavity structure in case of steam explosion triggering “at the right time” (because of the vicinity of supporting structures). In the second category, jets falling at once as a compact array but non-interfering with each other is believed to be the worst case. But, at the present stage of knowledge, it is not possible to define which would be most challenging for the structures.

Note that melt quenching induced steaming (without steam explosion) is an issue for some reactor designs having containments of reduced volume. This situation can be considered as a sub-group of pre-mixing.

5.5 Relevant experiments

Noting that no relevant multi-pour experiment exists, the best that could be extracted today from experimental database in relation to the SERENA Phase 1 objectives and the relevant conditions for reactors was found in the FARO, KROTOS and TROI programmes. The specific tests selected and analysed are indicated in the corresponding sections.

NEA/CSNI/R(2007)11

NEA/CSNI/R(2007)11

27

6. CALCULATION OF PRE-MIXING EXPERIMENTS (TASK 2)

6.1 Selected experiments

Selection of pre-mixing experiments and pre-selection of explosions experiments was part of Task 1. According to the work plan for phase 1 of SERENA and taking into account the limited resources given to the programme, only two selected tests were fully analysed:

• FARO L-28 [7], which investigated the premixing phenomena only with the pouring of 175 kg of 80 wt% UO2 -20 wt% ZrO2 molten corium in nearly saturated water,

• FARO L-33 [8], which was an integral experiment covering both premixing and explosion with the pouring of 100 kg of 80 wt% UO2 -20 wt% ZrO2 molten corium in subcooled water.

Two other pre-mixing tests of interest for the material effect (PREMIX-16, performed with alumina in conditions similar to L-28) and pre-mixing in sub-cooled conditions (FARO L-31, performed in conditions similar to FARO L-33, for which pre-mixing debris data are not available due to the explosion), were let for analysis on a voluntary basis. Actually, only FARO L-31 was partly analysed by partners because it gave indication of the pre-mixing debris applicable to FARO L-33.

Table 5 summarises the specific conditions and available data of the FARO tests. A schematic view of FARO test vessel and initial conditions in FARO L-28 are shown in Figure 1.

6.2 Calculation conditions for pre-mixing experiments

Partners were given a common set of reference data on the conditions and results of the experiments. Figure 2 shows the various geometry approximations and nodalisations used in the simulations.

6.3 Comparison of pre-mixing calculation results

Figure 3 (a,b) and Figure 4 compare the various predictions for pressure and energy release with the experimental records for FARO L-28 and FARO L-33, respectively. Experimental energy data were calculated by using the pressure and temperature records. Time zero corresponds to start of melt delivery to the water. In FARO L-28 melt delivery duration is approximately 6 s, and the melt front reaches the bottom of the test section after about 1 s. In FARO L-33, pre-mixing ceases at about 1.1 s when the explosion is triggered, corresponding approximately to melt-bottom contact.

Figure 5 shows comparison between predicted radially-averaged void fraction and global value calculated from experimental level swell for FARO L-28 at time 1.0 s and time 4.0 s, respectively. Figure 6 shows a comparison of the global void fractions obtained from averaging the local values calculated by the various codes at 1s and 4s, respectively. Experimental values in Figure 5 and Figure 6 have been calculated by using the level swell records.

Figure 3 shows that the codes have difficulties to reproduce the pressure data both for the initial phase of the pressurisation up to melt-bottom contact and for the linear increase of the pressure, which roughly corresponds to a steady state phase. They tend to underestimate heat transfer (Figure 3) and overestimate

NEA/CSNI/R(2007)11

28

void (Figure 5) when compared to data. Void distribution in pre-premixing is a key issue as void is considered having a key limiting effect on explosion strength. Highly water depleted pre-mixtures (void fraction > 50%) are considered to be an obstacle to strong steam explosions. Therefore, the discussions in SERENA focused on finding the major processes impacting the level of void.

Different views were expressed about:

• Importance of jet break-up description. • Jet break-up (primary break-up) and particle break-up (secondary break-up) dominant mechanisms. • Importance of understanding the long-term phase of quenching in L-28, in consideration of the fact

that that most energetic explosions are supposed to occur if triggered when the melt front contacts the bottom of the water pool.

• Significance of far-field radiation heat transfer. • Contribution of gas generation from cooling of settled debris. • The reasons for the initial strong pressure increase in L-28 (see Figure 3a). • Choice of continuous/discontinuous phase transition, heat transfer parameters and cut-off limits. • Contribution of non-condensable gases (entrainment of cover gas and production of hydrogen).

Large differences exist on modelling these phenomena [9]. An example is given in Table 6 concerning the different ways initial jet conditions and jet/particle break-up are described in the present simulations. Number of these phenomena, such as far-field radiation, presence of non-condensable gases and gas generation form settled debris, are not modelled in some codes and no code describes all together at once. The scatter of the predictions, their distance to the data and the uncertainties which affect experimental data concerning the internals of the pre-mixing region made not possible to establish whether a model is more appropriate than another to describe a given phenomena.

Sensitivity calculations showed that reducing significantly interfacial steam/water friction, or changing the transition range between water and steam continuous regimes allowed reducing void to values comparable to the data. Reducing the initial particle diameter or forcing heat exchange in continuous steam flow regime improved the pressurisation curve. However, no physical basis exists, which could justify one or another modification. In addition, large uncertainties affect experimental data in the absence of detailed information of the pre-mixing zone internals: experimental void fraction is global value retrieved form water level swell measurements, which does not allow to identify where the void is located actually. The reasons for the initial strong pressure increase in L-28 and the contribution of the debris cooling to the overall void level are not well understood.

6.4 Conclusion on calculation of pre-mixing experiments

Concluding, one can say that the major uncertainty on pre-mixing identified from application to the selected experiments stands in component distribution., in particular void fraction (prediction and data), which is considered to have a key effect on explosion strength. The major question in relation to the scope of SERENA is whether the differences in modelling and the scatter of the predictions are relevant for reactor applications, i.e., do these discrepancies induce uncertainties on the prediction of loads in reactors such that one cannot conclude that these loads are calculated with reasonable safety margins? This question will be answered after analysis of reactor calculations. One can simply say for now that most codes overestimate void fraction with respect to the data, especially in saturated conditions.

NEA/CSNI/R(2007)11

29

7. CALCULATION OF EXPLOSION EXPERIMENTS (TASK 3)

7.1 Selected explosion experiments

In addition to FARO L-33, two other experiments were selected for testing the code performance for the explosion phase, namely, KROTOS-44 with alumina melt [10] and TROI-13 with 70 wt% UO2-30 wt% ZrO2 melt [11]. Table 7 summarises the main conditions of these tests. An experiment with alumina melt was chosen because it was a well-characterised 1-D steam explosion with well defined external trigger, producing a very energetic and unambiguous steam explosion, and thus well appropriate to test the explosion models.

Then, TROI-13 dealing with a similar quantity of corium and quasi 2-D axi-symmetric geometry (meaning that pre-mixture does not occupy all the cross-section) represented an extension to more realistic conditions.

Lastly, FARO L-33 with 25 kg of corium melt in water at the time of the trigger and 2-D geometry represented a step further in scale. Figure 7 to Figure 9 show test geometry’s and explosion pressures as measured in KROTOS-44, TROI-13 and FARO L-33, respectively.

7.2 Calculation conditions for explosion experiments

For FARO L-33 and TROI, partners performed calculation of the explosion phase both with pre-mixing as calculated by the codes and with imposed pre-mixing conditions. For KROTOS K-44, calculations were performed with imposed pre-mixing conditions only, see Figure 10. Note that a blind exercise was done also on a TROI test, subsequently performed as TROI-34. Results did not significantly differ from TROI-13 and are not discussed here.

Table 8 and Table 9 summarise the mesh used in the various codes for KROTOS K-44 and FARO L-33, respectively. Table 10 and Table 11 indicates how the trigger has been described or modelled in the various codes for KROTOS K-44 and FARO L-33, respectively.

Time zero in the figures is trigger activation.

7.3 Comparison of explosion calculation results

7.3.1 KROTOS K-44

Figure 11 shows comparison of predicted explosion pressure and corresponding impulse with experimental value at mid-height in the water pool for KROTOS-44 (pressure transducer K3 in Figure 7). It can be seen that most of the models applied to the same initial and boundary conditions were able to globally reproduce the strong event observed in the experiment even with differences in modelling of the key effects of fragmentation, and non-homogenous heat transfer to the coolant. Comparison with the other pressure records at different levels in the water (Figure 12 to Figure 18) shows a common interpretation of the experimental results as a propagating-escalating event. This is somehow not surprising since KROTOS-alumina data has been used as a basis to validate the models. In general, standard values of the parameters were used for the simulations. Note, however, that some predictions noticeably underestimate the loads and that the explosion energy is in all cases lower than the value calculated from the experimental data (Figure 19).

NEA/CSNI/R(2007)11

30

7.3.2 TROI-13 and FARO L-33

For TROI-13 (Figure 20) and FAROL-33 (Figure 22 to Figure 30) tests, the agreement seems of the same order than for KROTOS, despite the events in these experiments were significantly less energetic than in KROTOS K-44, with however a larger overestimate of the loads for TROI-13 than for L-33. Actually, TROI-13 was likely a series of local mild explosions and shock-wave reverberations (ringing) on walls (see Figure 8), which was approximated by calculating a coherent propagating event and may explain the larger differences with respect to the FARO L-33 simulations.

However, the calculated loads for both TROI and FARO have been obtained with reducing more or less arbitrarily key effects such as heat transfer and fragmentation, which makes their reliability questionable. In addition, the data concern two similar types of oxidic corium only, namely, 70 wt% UO2-30 wt% ZrO2 and 80 wt% UO2-20 wt% ZrO2. Therefore, it would be hazardous to extend the conclusion of “low explosivity” to other corium melt compositions before understanding the very reasons that led to mild explosions in the FARO, KROTOS-corium (not calculated here, but which demonstrate the same low energetics [10,12]) and TROI experiments.

Possible physical explanations for the observed reduced explosion energetics are melt freezing and hydrogen production during pre-mixing. But differences in test and/or pre-mixing geometry may also have a reducing effect because they allow venting during the explosion. Actually, visualisations performed in KROTOS has shown that differences in pre-mixing lateral extension exist between alumina and corium [12]: with alumina pre-mixture tends to occupy all the test section cross area giving a radial 1-D-like configuration, while with corium pre-mixture is more concentrated around the centre-line, giving more room for explosion venting.

Figure 22 and Figure 23 show the influence of the pre-mixing configuration when comparing the calculations performed with imposed and calculated pre-mixing for FARO L-33. Void fraction for imposed premixing calculations was 5%, while pre-mixing calculated values were higher in general. Actually, the level of loads are higher in the former case, with the exception of VESUVIUS (see impulses in Figure 24 to Figure 30). Sensitivity studies performed by IKE with IDEMO showed that a level of void up to 40% does not change very much the maximum explosion pressure (Figure 31). A value of 20% gives a better agreement with the experiment concerning the propagation velocity, which would corresponds to a void uniformly distributed over a diameter of 35 cm instead of 71 cm as supposed when giving 5% (value given for prescribed pre-mixing, deduced from level swell measurement assuming uniform void in the entire water volume). However, explosions with 60% void could not be triggered according to calculations performed by IKE with both IDEMO and MC3D.

The influence of the trigger strength on the explosion propagation and escalation has been assessed with IDEMO and ESPROSE.m. No significant influence has been found, provided the trigger is strong enough to initiate the explosion.

Large scatter of the predictions exists also on kinetic energy for both FARO L-33 and TROI-13 (Figure 21 and Figure 32, respectively).

7.4 Conclusion on calculation of explosion experiments

The major uncertainties on explosion as can be deduced form application to the selected experiments are related to the absence of clear understanding of pre-mixing behaviour, as already revealed by the scatter of results obtained in pre-mixing experiment simulations, and of melt behaviour during propagation of the explosion (material effect).

NEA/CSNI/R(2007)11

31

For the two basic descriptions of the explosion used in the codes, namely the micro-interaction and the non-equilibrium heat transfer concepts [9], a number of parameters have to be given for the fragmentation, heat release and heat partition between steam and water. These parameters govern escalation and propagation of the explosion for a given pre-mixture, which determines the strength of the explosion. Using the default values for these parameters, in general set up on alumina experiments, with different given or calculated pre-mixing configurations at the time of the trigger, significantly overpredicts the loads with respect to the data obtained with corium. Playing with these parameters allows in general finding back the order of magnitude of the data, but basic physical explanation is missing.

Again, reducing the above uncertainties is important relatively to the objective of SERENA as far as it is relevant for reactor estimates, and in particular if it may have impact on evaluation of safety margins. This question is addressed in the next section.

NEA/CSNI/R(2007)11

NEA/CSNI/R(2007)11

33

8. REACTOR CALCULATIONS (TASK 4)

8.1 Selected geometry’s and conditions

Figure 33 summarises the initial and boundary conditions used for in- and ex-vessel cases, respectively, according to the generic situations of most concern identified in Task 1 (see Chapter 5). When looking at these conditions, one has to keep in mind that the scope is not to calculate a specific scenario in a specific reactor geometry and draw conclusions about the FCI risk for that geometry. Instead, the scope is to verify whether the codes used by the partners as their tools for FCI analysis are able to calculate a plausible LWR accident situation, consistently with Task 1 outcome, and to compare the results in order to identify the differences and the necessary actions to understand and reduce them.

A gravity pour was considered for both in- and ex-vessel. For in-vessel, the multi-jet configuration is obviously fully 3-D while the codes are used 2-D axi-symmetric or 1-D (TEXAS-V) when applied to reactor situations, with the exception of PM-ALPHA, which performs 3-D calculations. This was made deliberately in order to include in the simulation all the issues related to code extrapolation to reactor cases, and, in particular, the simplifications that have to be made to reduce to 2-D or 1-D the actual 3-D situations.

The same oxide melt composition was used for both in- and ex-vessel situations, because: • Most codes are not calculating chemical reactions and subsequent hydrogen production. • Most codes have no model which takes into account the presence of inert gases. • In the present state of the codes, the impact of changing the melt physical properties when changing

from a pure oxide corium to a corium containing metallic phases in reactor prototypical was estimated of second order.

8.2 Calculation conditions

The calculations were made for both the pre-mixing and the explosion phases. A trigger was applied when the melt front reached the bottom of the vessel or cavity. For the in-vessel case, this occurs approximately after 1 s. It was planned to compare calculations with a trigger applied after 4 s. However, it was found that this case had little interest as practically no liquid water was calculated to be present in the mixing zone at that time, and in any case negligible loads were calculated.

It is important to note that the choice of parameters for explosion was not made consistently with respect to Task 3. Some partners used the reduced parameters and some the standard ones, according to whether or not they favoured corium-related or conservative assessment, respectively. Some differences in the choice of the parameters are noticed for pre-mixing also, essentially due to a necessary adaptation to the initial conditions. Table 12 and Table 13 show how parameters and options have been chosen for the calculation of the reactor cases relatively to the calculation of the pre-mixing and explosion experiments, respectively.

8.3 Comparison of the results

8.3.1 In-vessel case

Figure 34 shows the geometry’s and nodalisations used by the different codes for the in-vessel case simulation.

NEA/CSNI/R(2007)11

34

Figure 35 and Figure 36 show the dynamic pressure histories calculated at the wall and the corresponding impulses, respectively. For each code, it corresponds to the pressure history at the location where the maximum value was obtained at a time during the explosion. In general, those impulses were also the maximum obtained. In a few cases, the impulse was slightly higher at another location, but not such as to bias the conclusions that can be drawn from the exercise.

The results show noticeable differences in predicting the peak pressure at the RPV bottom (ranging from ~10 MPa to ~120 MPa, see Figure 35) and a rather reduced prediction range for the impulse (from some tens of kPa.s to ~200 kPa.s, see Figure 36). When looking at the impulses, some curves exhibit continued increase and/or maximum values do seem to have been reached. Actually, looking at the corresponding pressure curves shows that it is due either to a pressure level remaining high for a long time (MC3D) or to stopping the calculation after the pressure pulse of the explosion has been obtained (IDEMO). In both cases, the maximum impulse due the explosion has been reached actually, and is within the range indicated above. The question of why the pressure remains so high for a long time in some calculations would deserve more attention. As impulses have been obtained by integrating pressure over the time, care has to be taken when interpreting the results. Considering only the value at the end of the calculation time window may lead to overestimating the explosion load. For example, ESPROSE.m calculates pressure that remains long below the design pressure of the vessel, which induces “impulse” increasing continuously while not presenting any risk for the vessel integrity (Figure 36). MC3D PSA exhibits a pressure peak of 80 MPa at 3.5 ms, which vanishes at 8 ms, and the pressure level after that time does not present any challenge for the vessel. The impulse at 8 ms is 80 kPa.s against 150 kPa.s at the end of the time window considered (20 ms), respectively.

From the above considerations, it can be concluded that the effective loads are below the capacity of the vessel. Taking into account the large differences of the hypotheses, approaches, models and parameter setting philosophies, this is a remarkable result, which could allow concluding that the safety margins for in-vessel steam explosion are sufficient in a “fitness for purpose” sense. However, this conclusion is challenged by:

• The fact that the calculated case might not be the worst possible. Actually, the choice of a multi-jet configuration was made based on the expert judgment that it would have induced the largest loads due to the largest melt masses in pre-mixing with respect to a single melt jet typical in-vessel situation (see Section 5.3). But according to the calculations, the selected configuration induced large voiding of the pre-mixing region and relatively small fragmented melt masses (Figure 38 and Table 14), which in turn resulted in small fragmented masses in the explosion wave (Figure 40), and are the reasons for the reduced calculated loads. At the limit, the margins could be considered sufficient for the situation of concern.

• The uncertainties on the void predictions in the pre-mixing phase and on the influence of void on explosion propagation. All the codes calculate high void fractions, with values larger than 50% above elevation 1 m (Figure 37). In experiments simulations (Task 2), calculated void values were overestimated when compared to experimental values, which in turn are affected by significant uncertainties due their global character. As highly voided regions are considered to be an obstacle to the propagation of the explosion, the relatively mild calculated loads might be the consequence of a spurious mitigating effect induced by large voiding in the pre-mixing zone. But Figure 38 shows that concluding only on the basis of global void might be misleading: the lowest efficiencies are not always associated with the highest global voids. Therefore, removing the uncertainties on the level of void in the pre-mixture and clarifying its role in the propagation and efficiency of steam explosions explosion would be beneficial to consolidating the conclusion of relatively low loads.

• The scatter of the results. Any conclusion that can be drawn from the results requires that the predictions are within a reduced range as the first condition. How much reduced is reduced enough depends on considerations that are somehow subjective. A majority of partners were of the opinion that, if confirmed once resolved the remaining uncertainties (on void, essentially) the range of the

NEA/CSNI/R(2007)11

35

predictions as observed here for the impulses was sufficiently reduced as to validate the conclusion that the loads are not challenging for the vessel integrity. Some partners did not agree with this statement and were of the opinion that convergence of the models is required first.

8.3.2 Ex-vessel case

Figure 41 shows the geometry’s and nodalisations used by the different codes for the ex-vessel case simulation.

Figure 42 to Figure 45 show the dynamic pressure histories and impulses calculated at the cavity bottom and lateral walls. For each code, it corresponds to the pressure history at the location where the maximum value was obtained at a time during the explosion. Figure 45 shows the corresponding impulses. There were also the maximum obtained, except in a few cases where the impulse was slightly higher at another location, but not such as to bias the conclusions that can be drawn.

Figure 42 shows a large variety pressure curves calculated with the different codes for the cavity bottom. High pressures occur in earlier or later phases of pressure development (wave escalation and propagation), but over only small time spans (width of pressure peaks < 1 ms). Such peak pressures reach about 110 MPa (JASMINE F2) or are in a range of 60 – 80 MPa (JASMINE F1X, TEXAS UW, IFCI, ESPROSE -at later time-). Only very short pressure peaks with pressures in this range are obtained by IDEMO and MC3D. Longer lasting pressures do mostly not exceed 20 MPa. The very long lasting elevated pressure from the ESPROSE calculation in this range is of interest. It explains the continuous increase of impulse at the bottom from this calculation (see Figure 43). However, this increase only does harm if the long-lasting pressure load cannot be absorbed by the surrounding structures. Thus, as already noticed for the in-vessel case, looking only at the impulse is not sufficient and may be misleading, here overestimating loading.

Similarly, the results at the side wall show noticeable differences in the predictions for both the explosion pressure and the impulse, but with relatively lower loads. Maximum pressures (Figure 44) remain between 10 MPa (IDEMO) and 40 MPa (MC3D, ESPROSE, VESUVIUS), the latter again with relatively short pulses of < 2 ms. Again, ESPROSE yields a very long lasting pressure of 10 – 20 MPa over > 16 ms, IDEMO of 10 - 15 MPa over about 8 ms. The calculated impulses (Figure 45) vary from a few kPa.s to ~100 kPa.s, except for ESPROSE for the reasons indicated above.

The same conclusions as for the in-vessel case apply here concerning uncertainties related to void and the scatter of the results, with even an amplification of the scatter. Radially-averaged void fractions at trigger time (Figure 46) and global void values (Figure 48) shows values that are in most cases higher than for the in-vessel case, which does not necessarily result in lower values of the impulses at the bottom (Figure 49). But larger voids in pre-mixing are associated with larger pre-mixed melt (compare Figure 38 and Table 14 for in-vessel with Figure 48 and Table 15), which in turn resulted in larger fragmented masses in the explosion wave (compare Figure 40 for in-vessel with Figure 50 for ex-vessel) and partly explains that the level of loads are similar. Table 16 compares the impact of using different assumptions for representing melt penetration and break-up into water on void and melt distribution.

As already stated for the in-vessel case, void alone is not sufficient to explain the level of the loads. But obtaining such level of loads for voids as large as those calculated raises the question of what would be the loads for lower voids, taking into account that calculated void were considered overestimated when compared to experimental values, which in turn are affected by large uncertainties. Calculations for testing sensitivity of the code results to void would be helpful, but could not be addressed in the frame of Phase 1. It is not an easy task because major quantities are inter-dependent. In a more global (and easier?) way, it could be verified whether codes can predict pre-mixing with large quantities of melts and low void, and what would be the level of loads in those cases.

NEA/CSNI/R(2007)11

36

However, contrarily to the in-vessel case, all the codes calculate loads that are above the capacity of cavity walls for the situation of concern (may be not the most challenging case). This is safety relevant only in case it challenges the integrity of the containment. But its evaluation for any specific configuration requires having some confidence in the predictions of the loading of the cavity, which is not possible here due to the large scatter of the results. For in-vessel case one may judge that, despite the scatter of the results, the issue is resolved (for the situation of concern, at least) as no calculations predict overloading of the vessel. The closure of the in-vessel issue as a whole might therefore “just” require verifying that the calculated situation is bounding or confirm the results for another more challenging configuration.

For the ex-vessel case, this is not possible, neither it is possible to establish reliable safety margins of a given containment by using one or another code. It is therefore necessary to reduce the scatter of the predictions. This can only be accomplished by reducing the uncertainties on component distribution in pre-mixing and understanding the role of void and melt properties on steam explosion energetics.

8.4 Discussion of reactor calculations and routes for confirmatory work

A positive result of the exercise is that the different codes showed to be applicable for reactor calculations. The availability of the different codes for such analyses can be considered as a success. It is to be remarked that even complicated configurations as the multi-jet configuration were calculated by the codes, although in an approximate way. The scatter of the results, however, reveals a lack of understanding of some phenomena, especially concerning the limiting effects on the explosion strength.

One major issue revealed from reactor-oriented calculations is whether melt-rich mixtures of several tons of melt in water with liquid drop sizes in the range of ~ 1 cm diameter and void fractions below about 50 % are possible, which may produce stronger explosions. From the results of the calculations it can be concluded that in the ex-vessel (as well as the in-vessel) case such extended mixtures yielding strong explosions were not obtained, in spite of a large variety of approaches on melt break-up, various assumptions on initial break-up as well as application of different models on jet break-up, and a variety of modes of melt inflow into water. Although pressure peaks of up to 100 MPa were calculated, they were limited to local and short-time events, which were considered not to endanger the reactor pressure vessel and also not to cause extended damages in the containment. Whether, based on these results, damages and failure of structures in the cavity may yield relevant safety consequences remains a matter of debate, but is largely design-dependent and as such out of the scope of SERENA.

One question with respect to the possible establishment of large mixtures concerns the modelling of melt break-up depending on the scenario and specific conditions. Variation of initial break-up assumptions may be sufficient to show that, mainly due to void production, large explosive mixtures are not possible in realistic scenarios. Then, detailed modelling of jet break-up becomes less important. Such specific studies could not be performed in Phase 1 in a way that would have allowed to conclude.

Another question concerns the dynamics of jet break-up. The question is whether the continued break-up with time (secondary break-up according to the terminology used here), not just the drops as introduced initially, may yield temporally and locally melt-rich mixtures with less void. Actually, reactor calculations showed that high pressures could be calculated in spite of global high void, which may be partly explained by low void some regions of the pre-mixtures. Further clarification requires more specific inspection of the results. This would also help clarifying whether or not calculating secondary break-up is important: predictions with a same codes are made sometimes with calculating it, sometimes without calculating it (?).

However, should these sensitivity analyses have been performed at this stage and confirmed that large mixed masses with low void are not possible, the doubts that the codes calculate realistic voids would have remained. This is because application to experimental data in Task 2 were inconclusive in that respect,

NEA/CSNI/R(2007)11

37

mainly due to the fact that existing data have not the required level of precision. Therefore, performing confirmatory testing with improve instrumentation with the aim to remove the uncertainties of the detailed nature of the pre-mixing is required anyway.

A second major issue concerns the choice of modelling and model parameters with respect to the explosion phase. Choosing less strong fine fragmentation (fragmentation induced by the explosion wave) in the reactor applications based on the general argument that corium melt yields weaker explosions appears to be questionable since understanding is not provided. Even if a general trend to weaker explosions with corium is accepted, it remains to clarify the reasons and, at least, to which processes the effects have to be attributed. One such process may be solidification of part of corium melt fragments in pre-mixing, which would not be the case for, e.g., alumina. This solidified corium is then no longer available for fine fragmentation in a pressure wave, and/or obstacles the propagation the explosion, resulting in a reduced strength. Pysico-chemical behaviour is also invoked. No direct evidence of these effects has been provided so far. But there are also strong indications from the KROTOS experiments with corium, on which the argumentation of weaker corium explosions is usually based when compared with KROTOS experiments with alumina, that strong differences already occur in the mixing phase. This concerns the production with corium of smaller droplets, higher void and more restricted mixing zone giving more venting possibilities as compared to alumina. Therefore, reducing fine fragmentation by just tuning adequate parameters in the explosion models without physical explanation may be dangerous. The indications from KROTOS should at first be verified with respect to a possibly stronger break-up of corium in pre-mixing due to its higher density, considering that specific experimental conditions may have played an important role in the KROTOS experiments. This can be accomplished by performing confirmatory testing designed in such a way that any possible spurious effect due to the particular mode of running the experiments be eliminated (e.g., by controlling melt release conditions in KROTOS and by verifying that similar behaviour is observed in another experimental environment, e.g., TROI). These two experimental programmes are the only presently running FCI programmes using corium in the OECD countries.

Specific investigations with the codes directly addressing these questions are required, while the investigations in Phase 1 rather yielded an overview and impression of the status of the codes. Experimentally, further support is to be gained from continued KROTOS and TROI experiments allowing some more specific checks and especially checking over a range of conditions, especially with relevant changes of melt release into water (addressing specific experimental reasons for a different behaviour of corium and alumina) and variations of material composition, especially to confirm the observed low explosivity behaviour for ex-vessel corium, which have been poorly studied so far although ex-vessel situations reveal to be the most challenging for safety.

NEA/CSNI/R(2007)11

NEA/CSNI/R(2007)11

39

9. GENERAL CONCLUSION OF PHASE 1

Phase 1 of SERENA is the most accomplished international comparison exercise ever undertaken on steam explosion. The scope was to verify the predictive capability of the FCI codes when applied to reactor situations, to identify the major overall uncertainties which limit the confidences in those predictions and to explore confirmatory research routes which can lead to reduction of these uncertainties to acceptable levels for risk assessment. SERENA Phase 1 was also a state-of-the-art on the way people are using the codes for FCI induced load assessment.

One positive outcome of Phase 1 is that whatever the modelling and numerical approaches all the codes were able to calculate the reactor situations of concern, which was far from being evident at the start of the programme.

Another positive outcome is that, despite the large differences of the hypotheses, approaches, models, parameter setting philosophies and geometrical representations, no critical events with respect to reactor safety was predicted, especially for the in-vessel case with respect to lower head failure. Different hypotheses to explain these results have been actively discussed. With broad consensus the partners agreed that this was due to the fact that the codes calculated relatively limited melt mass in pre-mixture and high voids, and also to venting possibilities existing in large geometry. But it is not clear which effect is dominant.

However, it should be noted that only one case has been calculated for in-vessel and ex-vessel, respectively, that these cases might not be the worst possible and that only a few sensibility calculations could be performed in the time frame allocated to the project. In addition, there is the tendency to predict large void in premixing, which was judged to be an overestimation of voiding according to experimental data and is considered to have a mitigating effect, and to use “reduced” parameters to model the explosion, without firm physical reasoning, but which allows to reproduce experimental data. This altogether may induce underestimation of the loads. There is clearly a lack of knowledge of pre-mixing and explosion behaviour of corium melts, leading to a large scatter of the results, which further challenges the confidence one may have in those predictions.

Knowing these weaknesses, the next question is whether it is important to reduce them for safety assessment purposes. There was a broad consensus that this was not required for in-vessel FCI, i.e., that the safety margins for in-vessel FCI are large enough to encompass any possible underestimation of the predictions. On the contrary, partners agreed that performing research aiming at increasing the knowledge level of ex-vessel corium FCI behaviour, especially concerning the limiting effects on the explosion strength with the scope of reducing the scatter and increasing the confidence of the predictions, was required for ex-vessel FCI in order to be able to quantify the safety margin for the containment..

Partners agreed that performing a limited number of well-designed tests in particular in the existing KROTOS (CEA) and TROI (KAERI) corium facilities, accompanied with analytical work with the main aim of increasing the capability of the FCI models/codes for use in reactor case analyses and risk bounding, could significantly help reducing uncertainties on the pre-mixing and material influences on steam explosion energetics, and consequently, reducing the scatter of the prediction to acceptable levels for risk assessment. This effort is proposed to be undertaken as a second phase of SERENA.

NEA/CSNI/R(2007)11

NEA/CSNI/R(2007)11

41

REFERENCES

1. Nuclear Safety Research in OECD Countries: Major Facilities and Programmes at Risk, 2001. NEA#03145, ISBN: 92-64-18468-6, Available on the Web (in PDF) at: http://www.nea.fr/html/nsd/reports/nea3145-safetyresearch.pdf

2. Statement to CSNI on remaining research needs in the area of fuel/coolant interaction, Memorandum to CSNI, 13 November 2000, produced by the restricted exploratory meeting held 30-31 October 2000, NEA Headquarters, Paris, France.

3. D. Magallon et al., Detailed Work Plan for Phase 1 of the Programme, November 2001.

4. Magallon, D., Bang, K.-H., Basu, S., Berthoud, G., Buerger, M., Corradini, M.L., Jacobs, H., Meignen, R., Melikhov, O., Naitoh, M., Moriyama, K., Sairanen, R., Song, J-H., Suh, N., and Theofanous, T.G., OECD Programme SERENA (Steam Explosion Resolution for Nuclear Applications): Work Programme and First Results. NURETH-10, Seoul, Korea, October 5-9, 2003.

5. Annunziato, C. Addabbo, G. Leva, OECD/CSNI International Standard problem No. 39 on FARO Test L-14, Technical Note No. I.96.64, April, 1996, Joint Research Center and NEA/CSNI/R(1997) 31.

6. Basu S. and T. Ginsberg, A Reassessment of the Potential for Alpha-Mode Containment Failure and a review of the Current Understanding of Broader Fuel-Coolant Interaction (FCI) Issues, Second SERG Workshop, NUREG-1524 (1996).

7. Magallon D. et al., “MFCI Project—Final Report”, INV-MFCI(99)-P007, EUR 19567 EN (2000).

8. Magallon D. and Huhtiniemi I., Energetic event in fuel-coolant interaction test FARO L-33, ICONE-9 Conference, Nice, France, 8-12 April 2001.

9. Meignen R., Magallon D., Bang K.-H., Basu S., Berthoud G., Buerger M., Corradini M.L., Jacobs H., Melikhov O., Naitoh M., Moriyama K., Sairanen R., Song J-H., Suh N. and Theofanous T.G., Comparative Review of FCI Computer Models Used in the oecd-serena Program. Proceedings of ICAPP-05, Seoul, KOREA, May 15-19, 2005, Paper 5087.

10. Huhtiniemi, I., Magallon, D., and Hohmann, H., Results of recent KROTOS FCI tests: alumina versus corium tests, Nucl. Eng. Des. 189, 379-389 (1999).

11. Song, J.H., Park, I.K., Shin, Y.S., Kim, J.H., Hong,S.W., Min, B.T., and Kim, H.D., Fuel coolant interaction experiments in TROI using a UO2/ZrO2 mixture. Nucl. Eng. Des., 222, 1-15 (2003).

12. Huhtiniemi I., and Magallon, 1999, Insight into Steam Explosions with Corium Melts in KROTOS, Nucl. Eng. Des. 204, 391-400 (2001).

NEA/CSNI/R(2007)11

NEA/CSNI/R(2007)11

43

LIST OF SERENA CONTRIBUTORS

Main contributors to calculations and analysis

Finland ............................................ R. Sairanen (VTT) Task 1 France.............................................. G. Berthoud, G. Ratel (CEA-Grenoble)

R. Meignen (IRSN) Germany.......................................... H. Jacobs (FZ-Karlsrhue)

M. Bürger, M. Buck (IKE-Stuttgart) Japan ............................................... M. Moriyama (JAERI)

M. Naito (NUPEC) Korea............................................... J. H. Song (KAERI)

N. Suh (KINS) Russian Federation.......................... O. Melikhov (EREC) United States of America ................ S. Basu (NRC)

M. Corradini (University of Wisconsin) T. Theofanous, N. Dinh (University of California)

Programme Coordinator and Report Editor D. Magallon (CEA-Cadarache)

Secretariat J. Royen (OECD/NEA) up to March 2004 C. Vitanza (OECD/NEA)

NEA/CSNI/R(2007)11

NEA/CSNI/R(2007)11

45

TABLES

Table 1. SERENA Phase 1 time table

ID Task Name1 Task 1

2 End of Task 1 report

3 Task 2

4 End of Task 2 report

5 Task 3

6 End of Task 3 report

7 Task 4

8 End of Task 4 report

9 Task 5

10 Final report

Qtr 1 Qtr 2 Qtr 3 Qtr 4 Qtr 1 Qtr 2 Qtr 3 Qtr 4 Qtr 1 Qtr 2 Qtr 3 Qtr 4 Qtr 1 Qtr 2 Qtr 3Year 1 Year 2 Year 3 Year 4

Table 2. Codes used by partners in Phase 1

Partners Codes CEA/IRSN MC3D GRS (IKE) IKEMIX, MC3D/IKEJET, IDEMO FZK MATTINA KAERI/KMU TEXAS V, TRACER II NRC (UCSB, UW) PM-ALPHA, ESPROSE, TEXAS KINS IFCI 6.2 JAERI JASMINE EREC VAPEX NUPEC VESUVIUS

Table 3. SERENA Partners’ interests for melt-into-water situations

PWR’s/VVER’s BWR’s In-vessel Ex-vessel In-vessel Ex-vessel

CEA-IRSN X X EREC X FZK X X1 X X1 IKE X X X X JAERI X X KAERI/KMU X KINS X NRC X X X X VTT X

1. To a lesser extend

NEA/CSNI/R(2007)11

46

Table 4. Summary of melt-into-water most relevant reactor conditions EREC FZK IKE JAERI KAERI-KMU KINS VTT

Reference case In-vessel Ex-vessel Ex-vessel In-vessel Most critical Ex-vessel Ex-vessel Ex-vessel In-vessel Ex-vessel Ex-vessel

Lower head geometry Hemispherical Hemispherical

Pit geometry Cylindrical rather open

Vessel diameter, m 4.0 (inner) 4.4 (outer) 4.2-4.7 4.2-4.7Pit diameter, m 5 5.5 6.0-6.3Jet configuration 100 jets outer

holes of core plate

Single jet Central

Single jet Central

Multi-jet Multi-jet, if not interacting

Single jet Central

Multi-jet Multi-jet Single jet side pour

Single jet Side or bottom

Multi-jet

Jet diameter, m 0.07 0.20-0.80 0.60-1.50 0.05-0.08 0.05-0.15 0.05-1.00 0.072 0.072 0.1-0.5 0.071Jet entrance velocity, m/s < 20 < 20 <10 15-22Melt composition UO 2 -ZrO 2 or

UO 2- ZrO 2 -ZrUO2-ZrO2-Zr-Steel

UO2-ZrO2-Zr-Steel

UO2-ZrO2-metal

UO2-ZrO2-metal

Oxide Oxide-metal

UO2-ZrO2-Zr-Steel

Melt mass, 103 kg 10-50 20-40 50-170 100 109-270 66-110 66-110 1-severalMelt temperature, K 2320-2850 2000-3000 2900-3000 2700-2800 2800 2800Melt superheat, K 100-400 50-150 <50 100-200 100-200 100-200 High and

LowMelt flow rate, kg/s variable,

duration 5-15s1000 100's 233-5000

Injection pressure, MPa Melt head Melt head High and Low High and Low

High and Low

System pressure, MPa 0.2-2.0 0.1-0.3 0.1-0.5 0.2-0.5 <1.0 0.20-0.52 0.2-0.5 0.2-0.5 0.1Atmosphere (initially) Air Steam, Air,

N2, H2

Water temperature, K near T sat 293-353 T sat Saturated and subcooled

350-390 320 320 T sat 319

Water subcooling, K 0-20 20-50 80 0 Medium 71 (BWR) 3-43 (PWR)

0 depends on design

54 initially (sat. after 28s)

Water pool height, m 0.7-2.8 > 3 1.0-3.0 1.0-2.0 In-vess. < 10 Ex-vessel

5.0-8.2 1.5-6.85 1.5-6.85 8.3

Water mass, 103 kg 10 to 23Jet free fall in gas, m 0-1.1 < 1 2.5-4.5 0 0-4.5 0-5.35 0-5.35

CEA-IRSN NRC

NEA/CSNI/R(2007)11

47

Table 5. Summary of conditions of selected FARO tests for pre-mixing calculation

Experiment FARO L-28 FARO L-31 FARO L-33 Melt composition 80 wt% UO2

20 wt% ZrO2 80 wt% UO2

20 wt% ZrO2 80 wt% UO2

20 wt% ZrO2 Melt mass released 175 kg 92 kg 100 kg

(40 kg at trigger) Melt temperature 3053 K 3000 K 3070 K Melt superheat 203 K 150 K 220 K System pressure 0.51 MPa 0.22 MPa 0.41 MPa Water temperature 424 K 291 K 294 K Water subcooling 0 K 106 K 124 K Release diameter 0.05 m 0.05 m 0.05 m Δp melt delivery Gravity Gravity Gravity Free fall in gas space 0.89 m 0.77 m 0.77 m Water depth 1.44 m 1.45 m 1.62 m Test section diameter 0.71 0.71 m 0.71 m Free-board Closed volume (3.5 m3) Closed volume (3.5 m3) Closed volume (3.5 m3)

Available data • Melt jet characterisation in gas phase (visualisation).

• Pressures and temperatures in gas and water.

• Melt leading edge progression.

• Water level history. • Quenching rate. • Global void fraction. • H2 production. • Extent of mixing zone

(from visualisation). • Temperatures at

bottom plate. • Debris analysis.

• Melt jet characterisation in gas phase (visualisation).

• Pressures and temperatures in gas and water.

• Melt leading edge progression.

• Water level history. • Quenching rate. • Global void fraction. • H2 production. • Temperatures at

bottom plate. • Debris analysis.

• Melt jet visualisation in gas phase.

• Pressures and temperatures in gas and water.

• Melt leading edge progression.

• Water level history. • H2 production. • Propagation of

energetic event after triggering.

• Temperatures at bottom plate.

• Debris data.

NEA/CSNI/R(2007)11

48

Table 6. Jet representation and break-up conditions considered in the FARO tests simulation of the pre-mixing phase

Code (organisation)

Parameters

IFCI (KINS)

Collection of discrete particles. Initial particle diameter 0.020 m (range recommended in the code 0.020 – 0.040 m). Secondary break-up.a

IKEMIX (IKE)

Continuous jet with initial diameter and velocity as in the specifications. Diameter of particle generated from jet = 3 mm. No secondary break-up.

JASMINE (JAERI)

1-D continuous cylindrical liquid column located within the central cell with initialdiameter and velocity as in the specifications.

Diameter of particle generated from jet = 5 mm. Secondary break-up.

MATTINA (FZK)

Collection of macro- particles of diameter 0.050 m. Diameter of particle generated from the macro-particles = 10 mm. Secondary break-up.

MC3D 3.2 (IKE)

Continuous jet with initial diameter and velocity as in the specifications. Diameter of particle generated from jet = 4.5 mm in L-28 and 4 mm in L-31, L-33. Secondary break-up down to particles of 1 mm (imposed value).

MC3D 3.3.06 (CEA-IRSN)

Continuous jet with initial diameter and velocity as in the specifications. Diameter of particle generated from jet = 4 mm. Secondary break-up.

PM-ALPHA (UCSB)

Collection of discrete particles. Diameter of initial particles = 10 mm in FARO L-28 and 20 mm in FARO L-33. Secondary break-up in L-28 (final diameter imposed at 2 mm). No secondary break-up in L-33.

VAPEX-P (EREC)

Collection of macro-particles having same velocity and diameter 0.050 m. Diameter of particle generated from the macro-particles = 3 mm. No secondary break-up (only L-33 calculated).

TEXAS-V (KAERI-KMU)

Collections of discrete « blobs ». Initial diameter of blobs = mean diameter of discharge orifice. Secondary break-up with Rayleigh-Taylor + Kelvin-Helmholtz + Boundary layer

stripping (new model). TEXAS-V (UW)

Collections of discrete « blobs ». Initial diameter of blobs = mean diameter of discharge orifice. Secondary break-up: Rayleigh-Taylor instabilities only.

TRACER –II (KAERI-KMU)

Nab

VESUVIUS (NUPEC)

Continuous jet with initial diameter and velocity as in the specifications. Diameter of particle generated from jet = 1.2, 1.5, 2 mm (3 calculations). No secondary break-up.

a. The following terminology is used throughout the report for the pre-mixing phase:

• Primary break-up is jet break-up (there is no primary break-up for jet represented initially as a collection of discrete particles).

• Secondary break-up is further break-up of the particle stemming from the jet break-up (or break-up of the initial particles if jet represented as a collection of discrete particles).

b. Not available in the documentation provided.

NEA/CSNI/R(2007)11

49

Table 7. Conditions of pre-selected explosion tests

Experiment TROI-13 KROTOS-44 Melt and composition 70 wt% UO2 - 30 wt% ZrO2 100% Al2O3 Melt mass released 7.7 kg 1.45 kg Melt temperature ~3300 K 2673 K Melt superheat 359 K System pressure 0.1 MPa 0.1 MPa Water temperature 292 K 363 K Water subcooling 81 K 10 K Release diameter 0.02 m 0.03 Δp melt delivery Gravity Gravity after crucible impact Free fall in gas space 3.9 m 0.43 m Water depth 0.69 m 1.115m Test section diameter 0.60 m 0.20 Free-board Closed volume (8.03 m3) Closed volume (0.23 m3) Trigger No Yes (at bottom)

Available data Melt jet visualisation in gas phase • Pressures and temperatures

in gas and water. • Debris data.

• Pressures and temperatures in gas and water.

• Melt leading edge progression. • Water level history. • Propagation of energetic event

after triggering. • Temperatures at bottom plate. • Debris data.

NEA/CSNI/R(2007)11

50

Table 8. Summary of conditions used for the calculations of KROTOS-44

Partner CEA-IRSN FZK IKE JAERI KAERI-KMU KINS NUPEC UCSB UW

Code MC3D 3.4.02 MATTINA IDEMO-2D JASMINE3.0 TEXAS-V IFCI 6.0 VESUVIUS ESPROSE. m TEXAS -V

Walls Rigid Rigid Rigid Rigid Rigid Rigid Rigid Rigid Rigid 1-D/2-D 2-D

axi-symmetric 2-D axi-symmetric

2-D axi-symmetric

2-D axi-symmetric

1-D 2-D axi-symmetric

2-D axi-symmetric

2-D axi-symmetric

1-D

Mesh (rxz) 3×53 cells over 0.1x1.65 m

7×48 cells over 2 m in test vessel. 2×48 outer vessel

10×223 cells over 0.1× 2.23 m

4×33 cells over 1.8 m in test vessel. 2×33 outer vessel

32 1-D slices over a height of 2.15 m

5×26 cells over 2.11 m in test vessel. 4×26 outer vessel

7×78 over 0.1×1.56 m

5×85 over 0.1×1.7 m

30 1-D slices over a height of 2.15 m

Typicala dimension of cells rxz (cm)

From 4×1.5 at centre to 2.5×25 at wall. Outer vessel not modelled

1×2 to 2×2 1×1 for test section (closed)b. Outer vessel not modelled

2.5×5 in test vessel

7 cm high in test vessel. Total gas volume concentrated in upper region

2x5 in test vessel

From 1.6×2 centre to 0.5×2 periphery. Outer vessel not modelled

2×2 inner vessel. Outer vessel not modelled

6.7 high in test vessel up to elevation 1.55 m. Total gas volume concentrated in upper region (6 cells 10 cm high)

Time step (μs)

0.6 (average) Nac Na 4.8 (average) Na Na Na 1 Na

Trigger time in full calc. (s)

Not applicable 1.5d Not applicable Not applicable Not applicable Not applicable Not applicable 1.7e Not applicable

Deviation from prescribed initial conditions in interaction zone (ZI)

Nof Axially uniform distributions

Radially uniform distributions.

Radially variable and radially uniform distributions (2 calculations)

Axially (and radially) uniform distributions

Plug and bottom part, no deviation No indication for mixing zone → supposedly no

Case 1: no Case 2: initial particle size 5 mm

Radially uniform distributions

No indication → supposedly axially variable distributions

a. In most cases values are approximated. Gives information of the order of magnitude of the cell dimensions. b. Except for the two lowermost trigger cells which are of width 1 cm and height 1.2 cm. c. Not available from documentation provided. d. No digitised data provided. e. Pre-mixing calculated with 80% 15 mm and 20% 3 mm particles (no secondary break-up). f. "No" means variable axial and radial distributions as prescribed.

NEA/CSNI/R(2007)11

51

Table 9. Summary of conditions used for the calculations of FARO L-33 explosion phase

Partner CEA-IRSN FZK IKE JAERI KAERI-KMU KINS NUPEC UCSB UW Code MC3D 3.4.02a MATTINA IDEMO-2D JASMINE3.0 TEXAS-V IFCI 6.0 VESUVIUS ESPROSE. m TEXAS -V Walls Rigid Rigid Rigid Rigid Rigid Rigid Rigid Rigid Rigid 1-D/2-D 2-D

axi-symmetric 2-D axi-symmetric

2-D axi-symmetric

2-D axi-symmetric

1-D 2-D axi-symmetric

2-D axi-symmetric.

1-D

Mesh (rxz) 16×42 cells over 0.75×2.4 m

Not clear 10×98 cells over 0.355× 2.4 m

8×25 cells over 0.75×3.0 m

45 1-D slices over a height of ~ 3.2 m

9×34 over 0.74×2.31 m

9×60 over 0.75×3.5 m.

7x80 over 0.35x1.6 m.

35 1-D slices over a height of ~3.74m.

Typicalb dimension of cells rxz (cm)

1×5 to 7.5×5 inside test section; 15×5 outside

Not clear 3×2.5 up to r=15; 4×2.5 up to r=35.5; Outer vessel not modelled

5×10 Finer near inner vessel wall, water side

5-7 high FARO total gas volume concentrated in upper region.

6×6 to 6×8 inside test section; 13×6 to 13× 8 outside.

7×5 inside test section; 9×5 outside.

5x5 inner vessel; Outer vessel not modelled.

10 - 12.5 high FARO total gas volume concentrated in upper region.

Time step (μs)

Variable between 0.00001 and 1

200 Nac Variable 0.001 to 2 (FC) 0.001 to 10 (CC)

Na Na 0.5 1 Variable between 0.1 and 100.

Pre-mixing conditions in FCd / Task 2

Recalculated including non-condens. Gases.

Recalculated with updated version of code

As in Task 2 with MC3D.

As in Task .2 As in Task 2, but C(142) =0.2 instead of 0.6

As in Task 2 New: no L-33 calculation in Task 2

As in Task 2 with PM-ALPHA

Na

Trigger time in FC (s)

0.7 (MBCe

calculation) 1.13 (MBC experiment)

1.124 (MBC at 0.9 in pre-mixing calculation)

1.124 (melt has not reached the bottom in pre-mixing calcul.)

1.1 (MBC at 0.8 in pre-mixing calculation)

0.73 (MBC in pre-mixing calculation) or 1.2

1.097 (MBC in pre-mixing calculation)

1.324 (MBC in pre-mixing calculation)

~ 1.0 (MBC in pre-mixing calculation)

Na

Melt in water at trigger time in FC (kg)

30.8, time 0.7s 48.3, time 1.13

33 41.15 (32.5 in mixture, rest as a jet)

38 33 Na 38 20 Na

a. Includes non condensable gases. Version 3.3.06 in Task 2 without non-condensable gases. b. In most cases values are approximated. Gives information of the order of magnitude of the cell dimensions. c. Not available from documentation provided. d. FC: Full Calculation pre-mixing + explosion; CC: common pre-mixing conditions. e. MBC: melt front contact with bottom.

NEA/CSNI/R(2007)11

52

Table 10. Modelling/Description of the trigger for KROTOS-44

CEA-IRSN MC3D/3.4.02 • 12 MPa in two lowermost centreline cells (r=4 cm, total height=2 cm) located below elevation zero.

• Void fraction 0.50. • Vapour temperature 600 K. • Liquid temperature 304 K. • No melt.

FZK MATTINA • Numerical simulation. − 14.8 MPa gas filled chamber of 15 cm3 located

below the bottom of the facility and opened suddenly for prescribed pre-mixing.

− 14.8 MPa gas filled chamber of 150 cm3 located below the bottom of the facility and opened for 7 ms for full calculation.

IKE IDEMO-2D • 14.5 MPa in the two lowermost centreline cells (~15 cm3). • Void fraction 0.99. • Vapour temperature 1000 K. • Influence of trigger strength assessed.

JAERI JASMINE 3.0 • Numerical simulation − 14.8 MPa in lowermost centreline cell. − Volume 15.0 cm3 (void fraction 0.0153). − Expansion energy (pV) 222 J.

KAERI-KMU TEXAS-V • No information. KINS IFCI/6.0 • 4.0 MPa in bottom cell of pre-mixture area. NUPEC VESUVIUS • Trigger cell (15 cm3) filled with 15 MPa Argon gas. UCSB ESPROSE.m • Numerical simulation

− 1.5 cm3 gas volume (bottom centreline). − 2 kJ thermal energy.

UW TEXAS-V • Difficult to extract trigger data from input data set.

NEA/CSNI/R(2007)11

53

Table 11. Modelling/Description of the trigger for FARO L-33

CEA-IRSN MC3D/3.4.02 • CCa: 14.4 MPa in lowermost centreline cell (r=1 cm, z=5cm),Void fraction 0.50, Vapour temperature 1 000 K, Liquid temperature 304 K, No melt

• FC: Change of melt drops present in lowermost centreline cell into fragments, Fragment fraction 0.001, Fragment temperature 3000 K, Initial pressure 14 MPa

FZK MATTINA • Gas filled chamber of 4.8 cm diameter and 12 cm height located below the bottom of the facility

• Numerical simulation led to: − Pressure 800 MPa − Opening time of the chamber 15 μs

• Above values were used for prescribed pre-mixing • 200 MPa and 200 μs were used for full calculation

IKE IDEMO-2D • 14 MPa in lowermost centreline cell (r=3 cm, z=2.5 cm) • Void fraction 0.99 • Vapour temperature 1000 K • Influence of trigger strength assessed

JAERI JASMINE 3.0 • Numerical simulation − 35 MPa in lowermost centreline cell − Volume 28.6 cm3 − Void fraction 0.0046

• Energy (pV) 1 kJ KAERI-KMU TEXAS-V • High pressure (14 MPa) gas filled chamber located centreline

below the bottom of the facility KINS IFCI/6.0 • 12 MPa in lowermost centreline cell (r=5.9 cm, z=7.7 cm) NUPEC VESUVIUS • Numerical simulation:

− Lowermost centreline cell (r=7 cm, z=5 cm) − 2 kJ vaporisation energy − Pressure 25 MPa

UCSB ESPROSE.m • Numerical simulation led to: − 1 cm3 gas volume (bottom centreline) − 20 kJ thermal energy − Pressure 156 MPa − Temperature of m.fluid 6804 K − Coolant temperature 309 K − m.fluid fraction 0.058

UW TEXAS-V • Difficult to extract trigger data from input file

a. FC: Full Calculation pre-mixing + explosion; CC: common pre-mixing conditions

NEA/CSNI/R(2007)11

54

Table 12. Pre-mixing parameters used in reactor calculations with respect to those used in pre-mixing experiment calculations

Partner Code used Parameters FARO L-28 Parameters FARO L-31, L-33 Ex- and In-vessel reactor casesVersion 3.3.06 Version 3.306 Task 2

Version 3.4.02 Task 3 Version 3.3.06 CEA –

IRSN MC3D

Default values used Diameter of particles generated from jet: 4 mm Secondary break-up calculated

idem L-28 2 sets of calculations provided and used for explosion calculations: • 1 with pre-mixing para-

meters as used in Task 2. • 1 with pre-mixing

parameters as used for PSA studies at IRSN based on fitting FARO results (9 parameters varied).

Default value of critical We 13.75

idem L-28 Value of critical We: 30

Forced convection parts of film boiling heat transfer coefficients calculated with liquid conductivity on both sides of the interface

idem L-28 Forced convection in film boiling calculated with vapour conductivity

Initial diameter of discontinuous melt: 50 mm

idem L-28 idem L-28

FZK MATTINA

Secondary break-up calculated.

idem L-28 idem L-28

Values of parameters given, in particular: • Factor on total energy

leaving particle in Epstein-Hauser correlation: 1 (default) and 0.5

• Factors on radial and axial component of inter-phase drag coefficient: 1 (default) and 0.001

idem L-28 • Factor on total energy leaving particle in Epstein-Hauser correlation: ?

• Factors on radial and axial component of inter-phase drag coefficient: 0.001

Diameter of particles generated from jet: 3 mm

idem L-28 (for L-31) Diameter of particles generated from jet ?

IKE IKEMIX

Secondary break-up not calculated.

idem L-28 (for L-31) Secondary break-up calculated.

Version 2.1 Version 2.1 in Task 2 and 3.0 in Task 3

Version 3.2: Changes on jet break-up correlation

Particle from jet 5 mm in diameter

idem L-28 Particle from jet 3 mm in diameter in base case (P1). Sensitivity calculation with 10 mm (P2)

JAERI JASMINE

Secondary break-up calculated but found not significant

idem L-28 No secondary break-up of particles for ex-vessel Secondary break-up calculated except for central jet for in-vessel

NEA/CSNI/R(2007)11

55

Table 12. Pre-mixing parameters used in reactor calculations with respect to those used in pre-mixing experiment calculations (Cont’d)

Partner Code used Parameters FARO L-28 Parameters FARO L-31, L-33 Ex- and In-vessel reactor casesOld break-up model (TI only)a Old break-up model (TI only) Old break-up model (TI only) UW TEXAS-V NBREAK= 40 ( specifies the number of initially discrete masses)

NBREAK= 40 NBREAK= 2

Calculation of pre-mixing with new break-up model accounting for KHI, BLS, TI (TI only in old model)

Calculation of pre-mixing with new break-up model

Both break-up models used

NBREAK = 5 NBREAK= 0 NBREAK= 1

KAERI-KMU

TEXAS-V

Best fit of pressure curve obtained with KH break-up constant C(141) = 0.12 (standard value 0.01)

C(141) = 0.01 C(141) = 0.01

Initial diameter of molten corium drops in field 4: 20 mm

Unchanged Unchanged

GASCOEFF = 0.25 (maximum fraction of energy vapour can receive)

GASCOEFF = 0.1 GASCOEFF = 0.1

KINS IFCI/6.0

EPSA = 0.25 EPSA = 0.1 EPSA = 0.1 Jet represented as a collection of particles of initial diameter 10 mm

Initial particle diameter 20 mm in full calculations

Initially 80% 15 mm and 20% 3 mm particles

UCSB PM-ALPHA

Secondary break-up calculated No break-up No break-up

a. TI: Rayleigh-Taylor instabilities; KHI: Kelvin-Helmotz instabilities; BLS: Boundary Layer Stripping

NEA/CSNI/R(2007)11

56

Table 13. Pre-mixing and explosion parameters used in reactor calculations with respect to those used in explosion experiment calculations

Partner Code used Phase Parameters K-44 Parameters TROI-13, -34 Parameters FARO L-33 Ex-vessel reactor case Comments

Pre-mixing

No pre-mixing calculation required for K-44

Default values used: • Diameter of particles created from the jet: 4 mm • Velocity of these particles: 0.2 m/s axially and

radially

2 sets of calculations provided and used for explosion calculations: • 1 with pre-mixing

parameters as used in Tasks 2 and 3

• 1 with pre-mixing parameters as used for PSA studies at IRSN based on fitting FARO results.

CEA - IRSN

MC3D/3.3.06

Explosion Melt fragment diameter of 0.07 mm instead of 0.1 mm (default)

Melt fragment diameter 0.1 mm (default) Explosion parameters as in Task 3

• 2 base case calculations provided corresponding to the 2 different pre-mixing results

• Version 3.4.02 used in previous Tasks

No pre-mixing calculation required for K-44

No calculation provided Default value of critical We changed with respect to previous calculations: 30 instead of 13.75

FZK MATTINA

Pre-mixing and explosion

Values of some parameters given, in particular initial particle diameter 50 mm for pre-mixing calculation.

Heat transfer coefficients now use vapour conductivity instead of liquid conductivity for calculating heat transferred from the particles to the interface and from interface to liquid water

No specific explosion model. Explosion calculated with pre-mixing model

NEA/CSNI/R(2007)11

57

Table 13. Pre-mixing and explosion parameters used in reactor calculations with respect to those used in explosion experiment calculations

Partner Code used Phase Parameters K-44 Parameters TROI-13, -34 Parameters FARO L-33 Ex-vessel reactor case Comments IKEMIX

Pre-mixing

No pre-mixing calculation required for K-44

Explosion calculations performed with given conditions only.

• Task 3 performed with MC3D

• Values of parameters as in Task 2.

• Factor on total energy leaving particle in Epstein-Hauser correlation: 0.5 (default 1)

• Factors on radial and axial component of interphase drag coefficient: 0.001 (default 1)

• Diameter of particles generated from jet: 3 mm, no further break-up

• Factor on total energy leaving particle in Epstein-Hauser correlation?

• Factors on radial and axial component of interphase drag coefficient: 0.001

• Diameter of particles generated from jet ?

• Break-up of particles calculated

Standard values used, better results with heat transfer coefficient from the fragments of 5.105 against standard value of 5.104 W/m2.K

Standard values used, except for heat transfer coefficient from the fragments: 5.105 against 5.104 W/m2.K

Standard values used Use of 5.105 W/m2.K as heat transfer coefficient.

IKE

IDEMO-2D

Explosion

Some values indicated: • Fragment size 0.1 mm. • Factor for rate of

entrainment of cold liquid coolant into the heated part equal to 5.

• Fragmentation coefficient 0.13 (corresponds to dimensionless fragmentation time of 1.28).

Same values as for K-44 Same values as for K-44 Same values as for K-44.

Only hydrodynamic fragmentation in explosion

NEA/CSNI/R(2007)11

58

Table 13. Pre-mixing and explosion parameters used in reactor calculations with respect to those used in explosion experiment calculations

Partner Code used Phase Parameters K-44 Parameters TROI-13, -34 Parameters FARO L-33 Ex-vessel reactor case Comments

Pre-mixing

No calculation provided Default values Particle from jet 5 mm in diameter

• Version 3.2 of the code used instead of 2.1 for Task 2 and 3.0 for Task 3:

• Changes on jet break-up correlation

• Particle from jet 3 mm in diameter in base case (against 5 mm for FARO simulations)

• No secondary break-up of particles

3 pre-mixing calculations provided in addition to the base case (P2, P3, P4). Sensitivity tested on particle diameter, break-up length and two-phase friction factor

Default values used No calculation provided Default values in full calculation

Version 3.2 of the code used instead of 3.0: Change on switching scheme of fragmentation (too fast propagation with vers.3.0) and void effect in fragmentation

JAERI JASMINE

Explosion

Fragment size: 0.01 mm Fragment size: 0.01 mm Use of parameters tuned on KROTOS alumina experiments for base case explosion calculation (Same for L-33)

4 explosion calculation provided F1 to F4) corresponding to the 4 calculated pre-mixings, plus one with different calculation of void (F1X)

NEA/CSNI/R(2007)11

59

Table 13. Pre-mixing and explosion parameters used in reactor calculations with respect to those used in explosion experiment calculations

Partner Code used Phase Parameters K-44 Parameters TROI-13, -34 Parameters FARO L-33 Ex-vessel reactor case Comments No pre-mixing calculation required for K-44

Calculation of pre-mixing with new break-up model accounting for KHI, BLS, TI (TI only in old model)

Calculation of pre-mixing with new break-up model

Both break-up models used

Pre-mixing NBREAK= 0

(no free particle initially) NBREAK= 0 NBREAK= 1

• NBREAK specifies the number of initially discrete masses

• 3 calculations provided (C1, C2, C3): 2 with new model and different radial extension for pre-mixing; 1 with old model

Fragmentation constant C (CFR): 0.002

Fragmentation constant C (CFR): 0.002

Fragmentation constant C (CFR): 0.002

Fragmentation constant C (CFR): 0.002

TEXAS-V

Explosion Fragment radius (RFRAG): 0.02 mm

Fragment radius (RFRAG): 0.1 mm

Fragment radius (RFRAG): 0.1 mm

Fragment radius (RFRAG): 0.1 mm

3 calculations provided corresponding to the 3 pre-mixing cases above

KAERI-KMU

TRACER No calculation provided No calculation provided No calculation provided No value indicated No pre-mixing calculation required for K-44

No calculation provided Old break-up model (TI only)

Old break-up model (TI only) Pre-mixing

NBREAK= 40 NBREAK= 2

NBREAK specifies the number of initially discrete masses

Values as in L-33, except for fragmentation constant Cfr: 0.002 (empirically determined from KROTOS tests as being between 0.001 and 0.002)

Fragmentation constant Cfr: 0.001 and 0.00025

Fragmentation constant Cfr: 0.001

Fuel fragmentation time (=propagation time up to the nearest free surface): 1 ms

fuel fragmentation time (=propagation time up to the nearest free surface): 2 and 2.5 ms

fuel fragmentation time (=propagation time up to the nearest free surface): 1 ms

UW TEXAS-V

Explosion

Factor in mass fragmentation rate law F: form 1 to 0 at fuel fragmentation time

Factor in mass fragmentation rate law F: form 1 to 0 at fuel fragmentation time

Factor in mass fragmentation rate law F: form 1 to 0 at fuel fragmentation time

NEA/CSNI/R(2007)11

60

Table 13. Pre-mixing and explosion parameters used in reactor calculations with respect to those used in explosion experiment calculations

Partner Code used Phase Parameters K-44 Parameters TROI-13, -34 Parameters FARO L-33 Ex-vessel reactor case Comments

Values within the recommended range used in general

Values within the recommended range used in general

Default values used in general

KINS IFCI/6.0

Pre-mixing Initial diameter of molten corium drops in field 4: 15 mm

Initial diameter of molten corium drops in field 4: 15 mm

Initial diameter of molten corium drops in field 4: 20 mm

Initial diameter of molten corium drops in field 4: 20 mm

Explosion Fine fragment size: 100 μm Fine fragment size: 30 μm Fine fragment size: 100 μm

Fine fragment size: 10 μm

Initial particle size: 2 cases, 15 and 5 mm

Initial particle size: 3.19 mm as test data ?

Initial diameter of particles which separate from jet: 2.3 mm corresponding to Sauter Diam. of L-31

Initial diameter of particles which separate from jet: 3 mm

Pre-mixing Lower limit value of the

bubble diameter: DG= 0.6 mm

Secondary break-up calculated

NUPEC VESUVIUS

Explosion Fragment size: 0.075 mm (Sauter diameter of test debris)

Fragment size of 0.1 mm (Sauter diameter of test debris)

Fragment size: 0.1 mm No information given

PM-ALPHA

Pre-mixing

Calculated with 80% 15 mm and 20% 3 mm particles and no break-up : given values based on test data

Calculated with 50% 15 mm and 50% 3 mm particles and no break-up : given values based on test data

Particle diameter 20 mm constant in full calculation and 3.6 mm in given conditions

80% 15 mm and 20% 3 mm particles at trigger time

Fragmentation constant βf : 9

Fragmentation constant βf : 200

Fragmentation constant βf : 200

Fragmentation constant βf : 200

Thermal enhancement factor γt: stepwise decrease from 4 to 1 when pressure increases from 0.1 to 100 MPa

Thermal enhancement factor γt: stepwise decrease from 4 to 1 when pressure increases from 0.1 to 100 MPa

Thermal enhancement factor γt: constant value of 1

Thermal enhancement factor γt: constant value of 1

UCSB

ESPROSE.m

Explosion

Entrainment factor fe: 7 (conservative)

Entrainment factor fe: 7 (conservative)

Entrainment factor fe: 7 (conservative)

Entrainment factor fe: 7 (conservative)

NEA/CSNI/R(2007)11

61

Table 14. In-vessel case. Jet description and melt masses in the system at trigger

Jet description Total mass in water at MBC trigger (kg) [trigger time]

Total mass in water at 4.0 s trigger (kg) Comments

JAERI P1 2120 of which 1450 premixed [0.4 s]

20800 of which 2000 molten and 530 premixed

JAERI P2

1 central jet of diameter 8 cm 8 jets distributed at radius 32 cm 16 jets distributed at radius 64 cm P1: particles from jet 3 mm P2: particles from jet 10 mm

2120 of which 1450 premixed [0.4 s]

20800 of which 1870 molten and 550 pre-mixed

Change in particle diameter between P1 (3 mm) and P2 (10 mm) Mixed mass means averaged melt temperature above melting point and void fraction below 0.75.

KINS No information ~2040 [0.65 s] No calculation provided KAERI-KMU C1 697 [1.1.29 s] KAERI-KMU C2 No information 751 [1.49 s]

Not performed Mixing calculation failed after 2 s

C1: new break-up model (KI, KHI, BLS) C2: old break-up model (KI)

KAERI-KMU TR No calculation provided No calculation provided No calculation provided UW 25 simultaneous jet pour streams na [1.00 s] No calculation provided FZK 1 central jet of diameter 8 cm

8 jets distributed at radius 32cm 16 jets distributed at radius 64 cm

2348 of which 1020 fragmented [0.77 s]

15260 Calculation not performed

Practically no water in lower plenum at 4.0 s.

CEA-IRSN T2 ~1000 [0.6 s] ~2000 Fuel drop mass CEA-IRSN PSA

1 central jet of diameter 8 cm 2 curtains of thickness 4 cm at radii 16 and 32 cm

~1200 [0.7 s] ~2500

UCSB 24 jets of 8 cm over an area 120×160 cm2

~3800 [0.85 s] Not performed Most cells occupied by melt only at 4.0 s

IKE 1 central jet of diameter 8 cm 8 jets distributed at radius 16 cm 16 jets distributed at radius 64 cm

~3000 [0.8 s] ~15500

NEA/CSNI/R(2007)11

62

Table 15. Ex-vessel case. Melt masses in the system at trigger

triggering time / melt bottom contact (s)

Total mass out of vessel at trigger (kg)

Total mass in “water” at trigger (kg)

Mixed mass at trigger (kg)

Comments

JAERI P1 920 JAERI P2 2 900 JAERI P3 670 JAERI P4

0.5 / 0.5 4 400 3 130 (z < 4.1 m)

1 280

Mixed mass means averaged melt temperature above melting point and void fraction below 0.75.

KINS 1.2 / 1.12 (estimated) N.A. 7 000 N.A. KAERI-KMU C1 0.97 / ~1.00 4 710 N.A. C1: new model, 5.5 m mixing zone KAERI-KMU C2 0.97 / ~1.00 5 240 N.A. C2: new model, 3.85 m mixing zone KAERI-KMU C3 0.97 / ~1.20

5 760 4 710 N.A. C3: Old model, 5.5 m mixing zone

KAERI-KMU TR 0.6 / 0.6 N.A. 4 000 3 620 4000 from curve 3620 "Melt drop mass"

UW 1.00 / 1.00 N.A. N.A. N.A. FZK 1.36/ 1.36 7 373 7 373 6 400 Mixed mass means here “fragmented

mass”, i.e., drop diameter < 10 mm CEA-IRSN T2 0.95 / 1.00 8 000 6 400 1 400 CEA-IRSN PSA 1.00 / 1.00 8 000 6 400 1 600

Last column = fragmented melt

UCSB 1.36 / 1.38 N.A. N.A. 8 794 IKE 0.6 / 0.6 5 184 N.A. 1 300 1 300 = “fragmented melt” NUPEC 1.1 / 1.1 7 000 N.A. N.A.

NEA/CSNI/R(2007)11

63

Table 16. Impact of using different assumptions on melt break-up on void and melt distribution in pre-mixing phase

Calculation Description TRACER Jet modelled as drops of 0.5 m ∅ initially?

B.l. stripping → Decrease to only 0.4 m ∅ → Melt mass in form of very large drops, only little break-up → Small void → Only (coarsely) fragmented mass should contribute to explosion

jet mass (here simulated by [large] drops!) not to be considered

TEXAS UW Jet modelled by drops of 0.5 m ∅ initially Break-up at leading edge by Rayleigh-Taylor instabilities (RT hereafter) Parameter NBREAK=2 for drops at leading edge prone to break-up (in contrast to NBREAK=40 chosen for FARO L-28 and L-33, with argument that thinner jet there) → Large melt mass in mixture (Figure 47), but with relatively large diameters? → Relatively small void at bottom, but increasing strongly to ≈> 80% void above 2 m

(strong local variations, esp. in mass (Figure 47) (related to coarse break-up features?)

TEXAS KK Jet modelled by drops of 0.5 m ∅ initially

Case C1: In addition to RT break-up at leading edge, break-up by KH instabilities assumed along the jet, but with smaller rate than for L-28 and TROI (parameter C(141) = 0.01 instead of 0.12) → Weak break-up? → small melt surface → low void [further: distribution on whole cross-section, due to 1D model

→ Variation C2, but only reduction from 5.5 m to 3.85 m radius]

Case C3: Only RT break-up at leading edge → Even weaker break-up → small melt surface, → small void, even nearly no void,

as compared to C1 → Larger particles? particle sizes during premixing

PM-ALPHA Initially assumed drops: “two group” injection with 3 mm (20 % mass) and 15 mm (80%) diameter, no further break-up → Taken as broken mass → Instead of jet, the above drop injection is assumed → different problem → Large broken mass (initially assumed) → large void

MATTINA Initially assumed drops of 10 mm diameter Almost complete break-up (80 %) to diameter of about 4 mm obtained → high void in droplet range

IFCI

Jet m

odel

led

as d

rops

Initially assumed drops of 20 mm diameter Break-up to about 5 mm? Total drop mass taken for heat transfer in premixing and for fine fragmentation during explosion [i.e. like in PM-ALPHA, TRACER and MATTINA (with very different assumptions on initial drop size and break-up!), seemingly in contrast to TEXAS KK C1, C3 and also TEXAS UW (?)] → High void

NEA/CSNI/R(2007)11

64

Table 16. Impact of using different assumptions on melt break-up on void and melt distribution in pre-mixing phase (Cont’d)

Calculation Description MC3D PSA Smaller particles ≈1 mm than given 4 mm for T2, due to new jet break-up description in

feedback with surrounding mixture → More detailed checking and discussion would be important!

(e.g.: high void due to smaller particles, but reduced break-up from this higher void expected, also trend to decrease void by shift in flow pattern transition parameters)

→ Higher void than for T2: 80 % range MC3D T2 Similar mass stripped (to be explained by different relation between break-up and

surrounding mixture in old stripping model, as well as different parameters for transition flow pattern),but larger diameter (set to 4 mm) → Smaller void range : ≈ 10-20 %

IKEMIX New stripping model with new formulation of feedback to surrounding mixture, but again adapted to L-28 → Stripped mass in medium range, increasing to top like in JASMINE

(but stronger), in contrast to decrease to top from MC3D → Void higher than with MC3D T2, but lower in lower region than MC3D PSA;

mass and void parts higher than JASMINE JASMINE

Jet m

odel

led

as c

ontin

uous

, with

strip

ping

at s

urfa

ce d

ue to

shea

r flo

w in

stab

ilitie

s

Case P1: Similar range of mass stripped from jet as with MC3D and IKEMIX, less than IKEMIX, increasing to top as with IKEMIX, inversely to MC3D, particle diameter fixed: 3 mm, larger than with MC3D T2 and IKEMIX (1.5-2.5 mm) → Void rather small in lower range, then also increasing to ≈ 80% up to 3 m

Case P2: 10 mm drop diameter → Lower void, less solidification

Case P3: Stronger mass break-up of jet (factor 3 instead 1.7) → Shorter jet, more void in upper part, more freezing (?)

Case P4: Interfacial friction reduction with factor 0.01 → Significantly reduced void

NEA/CSNI/R(2007)11

65

Figure 1. Schematic view of FARO tests arrangement and initial component distribution in FARO L-28

NEA/CSNI/R(2007)11

66

Figure 2. Geometry and nodalisation used for FAROL-28 simulation

MC3D (CEA-IRSN) MC3D (IKE) IKEMIX (IKE)

Zone withocalculation

et injection

Inner Wa

Jet Injection

Zone without calculation

Inner wall

i=1 i=n k=1

0.74

1.44

2.33

0.0 cm

Outlet boundary condition: L28: (184+18) cm L31: (173 +18) cm

Initial water level: L31: 145 cm L28: 144 cm

Injection elevation: L28: 233 cmL31 22 cm

0.35 cm

304 cm

2-D axi-symmetric (r,z) 16×27 cells over 0.75×2.975 m. Variable mesh size axially and radially

2-D axi-symmetric (r,z) 10×26 cells over 0.74×2.33 m. Variable mesh size axially and radially.

2-D axi-symmetric (r,z) 7×21 cells over 0.35×3.04 m for inner vessel. Uniform mesh size of 7×12 cm in water. Annular space modelled as one single cell.

NEA/CSNI/R(2007)11

67

Figure 2. Geometry and nodalisation used for FAROL-28 simulation (Cont’d) MATTINA (FZK) JASMINE (JAERI) IFCI (KINS) VESUVIUS (NUPEC)

Radius [m]

Hei

ght[

m]

0 0.25 0.5 0.75 10

0.5

1

1.5

2

2.5

3

0.3531 0.7441

1.44

1.835

2.33

Δh1

Δh2

Δh3

rr

z i=10

k=57

i=7i=1

k=47

k=26

k=1

Drawing range for Drawing range for 2D contour figures2D contour figures

andand

Calc. domain of Calc. domain of the component the component

distribution.distribution.

The substitutive area of The substitutive area of the FAT vesselthe FAT vessel bottom.bottom.

rr

z

rr

z i=10

k=57

i=7i=1

k=47

k=26

k=1

Drawing range for Drawing range for 2D contour figures2D contour figures

andand

Calc. domain of Calc. domain of the component the component

distribution.distribution.

The substitutive area of The substitutive area of the FAT vesselthe FAT vessel bottom.bottom.

2-D axi-symmetric (r,z) 14×58 cells over 0.75×3.4 m. Variable mesh size axially and radially

2-D axi-symmetric (r,z) 8×25 cells over 0.66×3.0 m. Variable mesh size axially and radially

2-D axi-symmetric (r,z) 9×34 cells over 0.74×2.33 m. Variable mesh size axially and radially

2-D axi-symmetric (r,z) 10×57 cells over 0.75×3.5 m. Variable mesh size axially and radially.

Release vessel

NEA/CSNI/R(2007)11

68

Figure 2. Geometry and nodalisation used for FAROL-28 simulation (Cont’d)

PM-ALPHA (UCSB) TEXAS-V (UW) TEXAS-V (KAERI-KMU)

0.5 m

0.4m

1.4m

0.1m

0.6 m

0.7 m

2.9 m

Fuel injection

48474645444342414039383736353433323130292827262524232221201918171615141312111098765432

Area =0.3419 m 2(# 2~ # 5)H eight=0.0595m (# 3~ # 5) =0.0715m (# 2)

Area =0.3957 m 2(# 6~ # 25)H eight=0.0595m (# 6~ # 25)

Area =0.3957 m 2(# 6~ # 25)H eight=0.06(# 26~ # 38)

2.22m

1.44m

0.25

m

A rea =3.22m 2(# 39~ # 48)H eight=0.1m (# 39~ # 48)Just for m aking steam volum e of 3.22m 3

3-D calculation. Uniform parallelepipedic cells 10×10×10 cm over half the domain.

1-D calculation. 35 cells ~6.5 cm high over 2.33 m total height

1-D calculation. 35 cells ~6.0 cm high over 3.12 m total height

Material Fuel Volumes

K=19 &20

K=14

K=1 x=0

NEA/CSNI/R(2007)11

69

Figure 3 a. Comparison of calculated vessel pressure and energy release with data (bold curves) for FARO L-28 (Pre-mixing of 175 kg of corium melt in 1.5-m-deep saturated water pool; melt-bottom contact at ~1 s)

KAERI-KMU calculations labelled c1 and c2 correspond to two different settings of parameters affecting Kelvin-Helmhotz break-up and thick/thin film criteria.

L-28. Cover gas pressure

0.5

1.0

1.5

2.0

2.5

3.0

0 1 2 3 4 5 6 7 8

Time (s)

Pres

sure

(MPa

)

Experiment

JASMINE-pre/2.1VESUVIUS

IFCI/6.0TRACER-II/2.0

TEXAS-V c1(kaeri)TEXAS-V c2 (kaeri)

MATTINAIKEMIX

MC3D/3.2 (ike)MC3D/3.3.06 (cea-irsn)

TEXAS-V (uw)PM-ALPHA.L.3D

L-28. Total energy release

0.0

20.0

40.0

60.0

80.0

100.0

120.0

140.0

160.0

180.0

200.0

0 1 2 3 4 5 6 7 8

Time (s)En

ergy

(MJ)

NEA/CSNI/R(2007)11

70

Figure 3 b. Idem Figure 3a over the first 1.2 s (melt-bottom contact at ~1 s)

L-28. Cover gas pressure

0.5

0.6

0.7

0.8

0.0 0.2 0.4 0.6 0.8 1.0 1.2

Time (s)

Pres

sure

(MPa

)

Experiment

JASMINE-pre/2.1VESUVIUS

IFCI/6.0TRACER-II/2.0

TEXAS-V c1(kaeri)TEXAS-V c2 (kaeri)

MATTINAIKEMIX

MC3D/3.2 (ike)MC3D/3.3.06 (cea-irsn)

TEXAS-V (uw)PM-ALPHA.L.3D

L-28. Total energy release

0.0

10.0

20.0

30.0

0.0 0.2 0.4 0.6 0.8 1.0 1.2

Time (s)En

ergy

(MJ)

NEA/CSNI/R(2007)11

71

Figure 4. Comparison of calculated vessel pressure and energy release with data (bold curves) for FARO L-33 (1.5-m-deep subcooled water pool; triggering at 1.1 s in the experiment, i.e.,

approximately at melt bottom contact with 25 kg of corium melt in pre-mixing)

L-33. Cover gas pressure

0.40

0.42

0.44

0.46

0.48

0.50

0.0 0.2 0.4 0.6 0.8 1.0 1.2

Time (s)

Pres

sure

(MPa

)

Experiment

JASMINE-pre/2.1IFCI/6.0

TRACER-II/2.0TEXAS-V (kaeri)

MATTINAMC3D/3.2 (ike)MC3D/3.3.06 (cea-irsn)PM-ALPHA.L.3D

EREC 1-DEREC 2-D

L-33. Total energy release

0.0

5.0

10.0

15.0

20.0

25.0

0.0 0.2 0.4 0.6 0.8 1.0 1.2

Time (s)En

ergy

(MJ)

NEA/CSNI/R(2007)11

72

Figure 5. Comparison of radially-averaged void fraction calculated by the codes with global value calculated from experimental level swell

(bold lines) for FARO L-28 at time 1.0 s and time 4.0 s, respectively.

L-28. Void fraction radially averaged at time 1.0 s

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

Fraction

Elev

atio

n (m

)

JASMINEVESUVIUSIKEMIXMC3D (ike)PM-ALPHA

L-28. Void fraction radially averaged attime 4.0 s

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0Fraction

Elev

atio

n (m

)

VESUVIUSIKEMIX

MC3D (ike)JASMINE

NEA/CSNI/R(2007)11

73

Figure 6. Comparison of global void fraction calculated by the codes and experimental value

- - - Experiment 1s..…. Experiment 4s.

0

0.25

0.5

0.75

1

JASMIN

EVESUVIU

S

IFCITRACERTEXAS K

KMATTIN

AMC3D

CEA

MC3D IK

E

IKEMIX

TEXAS UW

PM-ALP

HAMea

n valu

eStd

dev.

Calculated Void 1s.Calculated Void 4s.

NEA/CSNI/R(2007)11

74

Figure 7. KROTOS-44 explosion pressures measured in water at six different elevations at test section wall

NEA/CSNI/R(2007)11

75

Figure 8. TROI-13 explosion pressures measured in water at two different elevations at test section wall

NEA/CSNI/R(2007)11

76

Figure 9. FARO L-33 explosion pressures measured in water at five different elevations at test section wall (pt 490 is pressure transducer located 490 mm form bottom plate, and so on)

1.124 1.126 1.128 1.130 1.132 1.134 1.136 1.138Time (s)

50

0

50

00

50

00

50

00

XXX FAT33F pt 490

X X X X X X

XX

X XX

X XX X X

X X X X

YYY pt 715

Y Y Y Y Y YY

YY

YY

Y YY Y

Y Y YY Y

ZZZ pt 940

Z Z

Z

Z Z Z ZZ Z Z

ZZ

Z Z ZZ Z Z Z Z

VVV pt 1165

V V V V V V VV

V VV

VV

VV V V V

V V

JJJ pt 1390

J J J J J J J J JJ

JJ J

J

J JJ J

JJ

propagation velocity 370 m/s

1.124 1.126 1.128 1.130 1.132 1.134 1.136 1.138Time (s)

50

0

50

00

50

00

50

00

XXX FAT33F pt 490

X X X X X X

XX

X XX

X XX X X

X X X X

YYY pt 715

Y Y Y Y Y YY

YY

YY

Y YY Y

Y Y YY Y

ZZZ pt 940

Z Z

Z

Z Z Z ZZ Z Z

ZZ

Z Z ZZ Z Z Z Z

VVV pt 1165

V V V V V V VV

V VV

VV

VV V V V

V V

JJJ pt 1390

J J J J J J J J JJ

JJ J

J

J JJ J

JJ

propagation velocity 370 m/s

1.124 1.126 1.128 1.130 1.132 1.134 1.136 1.138Time (s)

50

0

50

00

50

00

50

00

XXX FAT33F pt 490

X X X X X X

XX

X XX

X XX X X

X X X X

YYY pt 715

Y Y Y Y Y YY

YY

YY

Y YY Y

Y Y YY Y

ZZZ pt 940

Z Z

Z

Z Z Z ZZ Z Z

ZZ

Z Z ZZ Z Z Z Z

VVV pt 1165

V V V V V V VV

V VV

VV

VV V V V

V V

JJJ pt 1390

J J J J J J J J JJ

JJ J

J

J JJ J

JJ

propagation velocity 370 m/s

1.124 1.126 1.128 1.130 1.132 1.134 1.136 1.138Time (s)

50

0

50

00

50

00

50

00

XXX FAT33F pt 490

X X X X X X

XX

X XX

X XX X X

X X X X

YYY pt 715

Y Y Y Y Y YY

YY

YY

Y YY Y

Y Y YY Y

ZZZ pt 940

Z Z

Z

Z Z Z ZZ Z Z

ZZ

Z Z ZZ Z Z Z Z

VVV pt 1165

V V V V V V VV

V VV

VV

VV V V V

V V

JJJ pt 1390

J J J J J J J J JJ

JJ J

J

J JJ J

JJ

propagation velocity 370 m/s

NEA/CSNI/R(2007)11

77

Figure 10. Imposed pre-mixing geometry at trigger time for KROTOS-44 simulation

ZI

plug

Covergas

0.15m

0.7m

0.379m

αv = 0αl = 1αm = 0

αv = 0.165αl = 0.845αm = 0

αv av. = 0,09αl av. = 0,884αm av. = 0,026

r=0.1

Taulukko 1. Linear variation in ZI gives: − At elevation 150 mm αm = 0.032 ; αv = 0.090 ; αl = 0.878 − At elevation 850 mm αm = 0.020 ; αv = 0.090 ; αl = 0.890 − Diameter of pre-mixed particles at time of the explosion = 15 mm (according to KROTOS alumina pre-mixing test data)

NEA/CSNI/R(2007)11

78

Figure 11. Comparison of dynamic pressure and corresponding impulse calculated by the codes with experimental value (blue bold lines) at mid-height in the water pool for KROTOS-44

(All calculations with imposed pre-mixing conditions, except ESPROSE FC where pre-mixing has been calculation with PM-ALPHA. VESUV-1 and -2 correspond to initial particle size of 15 mm -as specified- and 5 mm, respectively).

0.0

20.0

40.0

60.0

80.0

100.0

0.000 0.002 0.004 0.006 0.008 0.010Time (s)

Pres

sure

(MPa

)

K3MC3DJASMI-AVESUV-1VESUV-2IDEMOESPROSE CCESPROSE FCTEXAS (kaeri)TEXAS (uw)IFCI

0.0

30.0

60.0

90.0

120.0

150.0

0.000 0.002 0.004 0.006 0.008 0.010

Time (s)Im

puls

e (k

Pa.s

)

NEA/CSNI/R(2007)11

79

Figure 12. Pressures and Impulses in water with MC3D 3.4.02 (CEA-IRSN) (K-44 – Pressures and impulses in water with MC3D)

0.0

20.0

40.0

60.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Pres

sure

(MPa

)

K5

MC3D

0.0

20.0

40.0

60.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Pres

sure

(MPa

)

K4

MC3D

0.0

20.0

40.0

60.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Pres

sure

(MPa

)

K3

MC3D

0.0

20.0

40.0

60.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Pres

sure

(MPa

)

K2

MC3D

0.0

20.0

40.0

60.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Pres

sure

(MPa

)

K1

MC3D

0.0

20.0

40.0

60.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Pres

sure

(MPa

)

K0

MC3D

0.0

20.0

40.0

60.0

80.0

100.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Impu

lse

(kPa

.s)

K5

MC3D

0.0

20.0

40.0

60.0

80.0

100.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Impu

lse

(kPa

.s)

K4

MC3D

0.0

20.0

40.0

60.0

80.0

100.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006

Time (s)

Impu

lse

(kPa

.s)

K3

MC3D

0.0

20.0

40.0

60.0

80.0

100.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Impu

lse

(kPa

.s)

K2

MC3D

0.0

20.0

40.0

60.0

80.0

100.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Impu

lse

(kPa

.s)

K1

MC3D

0.0

20.0

40.0

60.0

80.0

100.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Impu

lse

(kPa

.s)

K0

MC3D

5

NEA/CSNI/R(2007)11

80

Figure 13. Pressures and Impulses in water with IDEMO-2D (IKE) (K-44 – Pressures and impulses in water with IDEMO-2D)

0.0

20.0

40.0

60.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Pres

sure

(MPa

)

K5

IDEMO

0.0

20.0

40.0

60.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Pres

sure

(MPa

)

K4

IDEMO

0.0

20.0

40.0

60.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Pres

sure

(MPa

)

K3

IDEMO

0.0

20.0

40.0

60.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Pres

sure

(MPa

)

K2

IDEMO

0.0

20.0

40.0

60.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Pres

sure

(MPa

)

K1

IDEMO

0.0

20.0

40.0

60.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Pres

sure

(MPa

)

K0

IDEMO

0.0

20.0

40.0

60.0

80.0

100.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Impu

lse

(kPa

.s)

K5

IDEMO

0.0

20.0

40.0

60.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Impu

lse

(kPa

.s)

K4

IDEMO

0.0

20.0

40.0

60.0

80.0

100.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Impu

lse

(kPa

.s)

K3

IDEMO

0.0

20.0

40.0

60.0

80.0

100.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Impu

lse

(kPa

.s)

K2

IDEMO

0.0

20.0

40.0

60.0

80.0

100.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Impu

lse

(kPa

.s)

K1

IDEMO

0.0

20.0

40.0

60.0

80.0

100.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Impu

lse

(kPa

.s)

K0

IDEMO

NEA/CSNI/R(2007)11

81

Figure 14. Pressures and Impulses in water with JASMINE 3.0 (JAERI) (K-44 – Pressures and impulses in water with JASMINE)

JASMI A stands for prescribed conditions (UNI A in partner terminology) and JASMI B for uniform radial distributions of components (UNI B), respectively

0.0

20.0

40.0

60.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Pres

sure

(MPa

)

K5

JASMI-A

JASMI-B

0.0

20.0

40.0

60.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Pres

sure

(MPa

)

K4

JASMI-A

JASMI-B

0.0

20.0

40.0

60.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Pres

sure

(MPa

)

K3

JASMI-A

JASMI-B

0.0

20.0

40.0

60.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Pres

sure

(MPa

)

K2

JASMI-A

JASMI-B

0.0

20.0

40.0

60.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Pres

sure

(MPa

)

K1

JASMI-A

JASMI-B

0.0

20.0

40.0

60.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Pres

sure

(MPa

)

K0

JASMI-A

JASMI-B

0.0

20.0

40.0

60.0

80.0

100.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Impu

lse

(kPa

.s)

K5

JASMI-A

JASMI-B

0.0

20.0

40.0

60.0

80.0

100.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Impu

lse

(kPa

.s)

K4

JASMI-A

JASMI-B

0.0

20.0

40.0

60.0

80.0

100.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Impu

lse

(kPa

.s)

K3

JASMI-A

JASMI-B

0.0

20.0

40.0

60.0

80.0

100.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Impu

lse

(kPa

.s)

K2

JASMI-A

JASMI-B

0.0

20.0

40.0

60.0

80.0

100.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Impu

lse

(kPa

.s)

K1

JASMI-A

JASMI-B

0.0

20.0

40.0

60.0

80.0

100.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Impu

lse

(kPa

.s)

K0

JASMI-A

JASMI-B

NEA/CSNI/R(2007)11

82

Figure 15. Pressures and Impulses in water with TEXAS-V (KAERI-KMU and UW) (K-44 – Pressures and impulses in water with TEXAS-V)

0.0

20.0

40.0

60.0

80.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Pres

sure

(MPa

)

K5

KAERI

UW

0.0

20.0

40.0

60.0

80.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Pres

sure

(MPa

)

K4

KAERI

UW

0.0

20.0

40.0

60.0

80.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Pres

sure

(MPa

)

K3

KAERI

UW

0.0

20.0

40.0

60.0

80.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Pres

sure

(MPa

)

K2

KAERI

UW

0.0

20.0

40.0

60.0

80.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Pres

sure

(MPa

)

K1

KAERI

UW

0.0

20.0

40.0

60.0

80.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Pres

sure

(MPa

)

K0

UW

0.0

20.0

40.0

60.0

80.0

100.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Impu

lse

(kPa

.s)

K5

KAERI

UW

0.0

20.0

40.0

60.0

80.0

100.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Impu

lse

(kPa

.s)

K4

KAERI

UW

0.0

20.0

40.0

60.0

80.0

100.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Impu

lse

(kPa

.s)

K3

KAERI

UW

0.0

20.0

40.0

60.0

80.0

100.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Impu

lse

(kPa

.s)

K2

KAERI

UW

0.0

20.0

40.0

60.0

80.0

100.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Impu

lse

(kPa

.s)

K1

KAERI

UW

0.0

20.0

40.0

60.0

80.0

100.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Impu

lse

(kPa

.s)

K0

UW

NEA/CSNI/R(2007)11

83

Figure 16. Pressures and Impulses in water with IFCI 6.0 (KINS) (K-44 – Pressures and impulses in water with IFCI)

0.0

20.0

40.0

60.0

80.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Pres

sure

(MPa

)

K5

IFCI

0.0

20.0

40.0

60.0

80.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Pres

sure

(MPa

)

K4

IFCI

0.0

20.0

40.0

60.0

80.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Pres

sure

(MPa

)

K3

IFCI

0.0

20.0

40.0

60.0

80.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Pres

sure

(MPa

)

K2

IFCI

0.0

20.0

40.0

60.0

80.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Pres

sure

(MPa

)

K1

IFCI

0.0

20.0

40.0

60.0

80.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Pres

sure

(MPa

) K0

0.0

20.0

40.0

60.0

80.0

100.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Impu

lse

(kPa

.s)

K5

IFCI

0.0

20.0

40.0

60.0

80.0

100.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Impu

lse

(kPa

.s)

K4

IFCI

0.0

20.0

40.0

60.0

80.0

100.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Impu

lse

(kPa

.s)

K3

IFCI

0.0

20.0

40.0

60.0

80.0

100.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Impu

lse

(kPa

.s)

K2

IFCI

0.0

20.0

40.0

60.0

80.0

100.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Impu

lse

(kPa

.s)

K1

IFCI

0.0

20.0

40.0

60.0

80.0

100.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Impu

lse

(kPa

.s)

K0

NEA/CSNI/R(2007)11

84

Figure 17. Pressure in water with VESUVIUS (NUPEC) (K-44 – Pressures and impulses in water with VESUVIUS)

0.0

20.0

40.0

60.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Pres

sure

(MPa

)

K5

VESUV-1

VESUV-2

0.0

20.0

40.0

60.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Pres

sure

(MPa

)

K4

VESUV-1

VESUV-2

0.0

20.0

40.0

60.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Pres

sure

(MPa

)

K3

VESUV-1

VESUV-2

0.0

20.0

40.0

60.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Pres

sure

(MPa

)

K2

VESUV-1

VESUV-2

0.0

20.0

40.0

60.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Pres

sure

(MPa

)

K1

VESUV-1

VESUV-2

0.0

20.0

40.0

60.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Pres

sure

(MPa

)

K0

VESUV-1

VESUV-2

0.0

20.0

40.0

60.0

80.0

100.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Impu

lse

(kPa

.s)

K5

VESUV-1

VESUV-2

0.0

20.0

40.0

60.0

80.0

100.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Impu

lse

(kPa

.s)

K4

VESUV-1

VESUV-2

0.0

20.0

40.0

60.0

80.0

100.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Impu

lse

(kPa

.s)

K3

VESUV-1

VESUV-2

0.0

20.0

40.0

60.0

80.0

100.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Impu

lse

(kPa

.s)

K2

VESUV-1

VESUV-2

0.0

20.0

40.0

60.0

80.0

100.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Impu

lse

(kPa

.s)

K1

VESUV-1

VESUV-2

0.0

20.0

40.0

60.0

80.0

100.0

0.000 0.001 0.002 0.003 0.004 0.005 0.006Time (s)

Impu

lse

(kPa

.s)

K0

VESUV-1

VESUV-2

Case 1 (VESUV-1) is with initial particle size of 15 mm as specified Case 2 (VESUV-2) is with initial particle size of 5 mm

NEA/CSNI/R(2007)11

85

Figure 18. Pressures and Impulses in water with ESPROSE.m (UCSB) CC stands for prescribed conditions. FC is for pre-mixing conditions as calculated by PM-ALPHA

0.0

40.0

80.0

120.0

160.0

0.000 0.002 0.004 0.006 0.008 0.010Time (s)

Pres

sure

(MPa

)

K5ESPROSE CCESPROSE FC

0.0

40.0

80.0

120.0

160.0

0.000 0.002 0.004 0.006 0.008 0.010Time (s)

Pres

sure

(MPa

)

K4ESPROSE CCESPROSE FC

0.0

40.0

80.0

120.0

160.0

0.000 0.002 0.004 0.006 0.008 0.010Time (s)

Pres

sure

(MPa

)

K3ESPROSE CCESPROSE FC

0.0

40.0

80.0

120.0

160.0

0.000 0.002 0.004 0.006 0.008 0.010Time (s)

Pres

sure

(MPa

)

K2ESPROSE CCESPROSE FC

0.0

40.0

80.0

120.0

160.0

0.000 0.002 0.004 0.006 0.008 0.010Time (s)

Pres

sure

(MPa

)

K1ESPROSE CCESPROSE FC

0.0

40.0

80.0

120.0

160.0

0.000 0.002 0.004 0.006 0.008 0.010Time (s)

Pres

sure

(MPa

)

K0

0.0

40.0

80.0

120.0

160.0

0.000 0.002 0.004 0.006 0.008 0.010Time (s)

Pres

sure

(MPa

)

K5ESPROSE CCESPROSE FC

0.0

40.0

80.0

120.0

160.0

200.0

0.000 0.002 0.004 0.006 0.008 0.010Time (s)

Impu

lse

(kPa

.s)

K4ESPROSE CCESPROSE FC

0.0

40.0

80.0

120.0

160.0

200.0

0.000 0.002 0.004 0.006 0.008 0.010Time (s)

Impu

lse

(kPa

.s)

K3ESPROSE CCESPROSE FC

0.0

40.0

80.0

120.0

160.0

200.0

0.000 0.002 0.004 0.006 0.008 0.010Time (s)

Impu

lse

(kPa

.s)

K2ESPROSE CCESPROSE FC

0.0

40.0

80.0

120.0

160.0

200.0

0.000 0.002 0.004 0.006 0.008 0.010Time (s)

Impu

lse

(kPa

.s)

K1ESPROSE CCESPROSE FC

0.0

40.0

80.0

120.0

160.0

200.0

0.000 0.002 0.004 0.006 0.008 0.010Time (s)

Impu

lse

(kPa

.s)

K0

NEA/CSNI/R(2007)11

86

Figure 19. KROTOS-44 predicted histories of kinetic energy (experimental evaluation is 0.153 MJ)

K-44. History of kinetic energy

0.00

0.03

0.06

0.09

0.12

0.15

0 2 4 6 8Time (ms)

Kin

etic

ene

rgy

(MJ)

JASMINE A JASMINE BVESUVIUS 1VESUVIUS 2TEXAS KAERI TEXAS UW IDEMO MC3D

(Experimental evaluation 0.153 MJ)K-44. History of kinetic energy

0.00

0.03

0.06

0.09

0.12

0.15

0 2 4 6 8Time (ms)

Kin

etic

ene

rgy

(MJ)

JASMINE A JASMINE BVESUVIUS 1VESUVIUS 2TEXAS KAERI TEXAS UW IDEMO MC3D

(Experimental evaluation 0.153 MJ)

NEA/CSNI/R(2007)11

87

Figure 20. Comparison of dynamic pressure and corresponding impulse calculated by the codes with experimental value (blue bold lines) for TROI-13

(Imposed pre-mixing conditions except, MC3D FC: full calculation pre-mixing + explosion with MC3D; KAERI FC: full calculation with TEXAS; ESPR FC: pre-mixing calculation with PM-ALPHA, explosion calculation with ESPROSE, 07 kJ is trigger energy 0.7 kJ).

-2.0

0.0

2.0

4.0

6.0

8.0

10.0

12.0

0.000 0.005 0.010 0.015 0.020Time (s)

Pres

sure

(MPa

)

IVDP102IDEMOMC3D AMC3D FCVESUV-3KAERI DP02KAERI FC DP02IFCIESPROSE CCESPR FC 07kJ

0.0

5.0

10.0

15.0

20.0

25.0

30.0

0.000 0.005 0.010 0.015 0.020Time (s)

Impu

lse

(kPa

.s)

NEA/CSNI/R(2007)11

88

Figure 21. TROI-13 predicted histories of kinetic energy

(Experimental evaluation is 0.052 MJ. MC3D A, B, correspond to two different imposed component distributions. MC3D FC correspond to full calculation pre-mixing + explosion with the code)

TROI-13. History of kinetic energy

0.00

0.10

0.20

0.30

0.40

0 2 4 6 8 10Time (ms)

Kin

etic

ene

rgy

(MJ) VESUVIUS 3

TEXAS KAERI IDEMO MC3D AMC3D BMC3D FC

(Experimental evaluation 0.052 MJ )TROI-13. History of kinetic energy

0.00

0.10

0.20

0.30

0.40

0 2 4 6 8 10Time (ms)

Kin

etic

ene

rgy

(MJ) VESUVIUS 3

TEXAS KAERI IDEMO MC3D AMC3D BMC3D FC

(Experimental evaluation 0.052 MJ )

NEA/CSNI/R(2007)11

89

Figure 22. Comparison of dynamic pressure and corresponding impulse calculated by the codes with experimental value (bold lines) at level 1390 mm in the water pool for FARO L-33

(highest value measured). Full calculations premixing+explosion.

(MC3D R07 and R113 correspond to triggering at 0.7 s –calculated melt/bottom contact- and 1.13 s –experimental melt/bottom contact-, resp.)

0.0

4.0

8.0

12.0

16.0

20.0

0.000 0.005 0.010 0.015Time (s)

Pres

sure

(MPa

)

PT.1390

MC3D R07

MC3D R113

JASMINE

VESUVIUS

IDEMO

TEXAS (kaeri)

TEXAS (uw) BASE

TEXAS (uw) FIT

IFCI

ESPROSE

0.0

10.0

20.0

30.0

40.0

50.0

0.000 0.005 0.010 0.015Time (s)

Impu

lse

(kPa

.s)

NEA/CSNI/R(2007)11

90

Figure 23. Comparison of dynamic pressure and corresponding impulse calculated by the codes with experimental value (bold lines) at level 1390 mm in the water pool for FARO L-33 (highest value measured). Common pre-mixing conditions.

0.0

5.0

10.0

15.0

20.0

0.000 0.005 0.010 0.015Time (s)

Pres

sure

(MPa

)

PT.1390

MC3D CC

JASMI-B CC

VESUV CC

IDEMO CC

TEXAS (kaeri) CC

IFCI CC

ESPRO CC

0.0

40.0

80.0

120.0

0.000 0.005 0.010 0.015Time (s )

Impu

lse

(kPa

.s)

NEA/CSNI/R(2007)11

91

Figure 24. Pressures and Impulses in water with MC3D 3.4.02 (CEA-IRSN) “R07” and “R113” stand for trigger applied at time 0.7 s

(Melt front Bottom Contact in pre-mixing calculation) and 1.13 s, respectively.

L-33 - Pressures and Impulses in water - MC3D

0.0

4.0

8.0

12.0

16.0

20.0

0.000 0.002 0.004 0.006 0.008 0.010T (s)

Pres

sure

(MPa

)

PT.1615MC3D CCMC3D R07MC3D R113

0.0

4.0

8.0

12.0

16.0

20.0

0.000 0.002 0.004 0.006 0.008 0.010T (s)

Pres

sure

(MPa

)

PT.1390MC3D CCMC3D R07MC3D R113

0.0

4.0

8.0

12.0

16.0

20.0

0.000 0.002 0.004 0.006 0.008 0.010T (s)

Pres

sure

(MPa

)

PT.1165MC3D R07MC3DMC3D R113

0.0

4.0

8.0

12.0

16.0

20.0

0.000 0.002 0.004 0.006 0.008 0.010T (s)

Pres

sure

(MPa

)

PT.0940MC3D CCMC3D R07MC3D R113

0.0

4.0

8.0

12.0

16.0

20.0

0.000 0.002 0.004 0.006 0.008 0.010T (s)

Pres

sure

(MPa

)

PT.0715MC3D CCMC3D R07MC3D R113

0.0

4.0

8.0

12.0

16.0

20.0

0.000 0.002 0.004 0.006 0.008 0.010T (s)

Pres

sure

(MPa

)

PT.0490MC3D CCMC3D R07MC3D R113

0.0

10.0

20.0

30.0

40.0

50.0

0.000 0.005 0.010 0.015 0.020

Impu

lse

(kPa

.s)

0.0

10.0

20.0

30.0

40.0

50.0

0.000 0.005 0.010 0.015 0.020

Impu

lse

(kPa

.s)

0.0

10.0

20.0

30.0

40.0

50.0

0.000 0.005 0.010 0.015 0.020

Impu

se (k

Pa.s

)

0.0

10.0

20.0

30.0

40.0

50.0

0.000 0.005 0.010 0.015 0.020

Impu

lse

(kPa

.s)

0.0

10.0

20.0

30.0

40.0

50.0

0.000 0.005 0.010 0.015 0.020

Impu

lse

(kPa

.s)

0.0

10.0

20.0

30.0

40.0

50.0

0.000 0.005 0.010 0.015 0.020Time (s)

Impu

lse

(kPa

.s)

NEA/CSNI/R(2007)11

92

Figure 25. Pressures and Impulses in water with IDEMO-2D (IKE)

L-33 - Pressures and Impulses in water - IDEMO

0.0

2.0

4.0

6.0

8.0

10.0

0.000 0.002 0.004 0.006 0.008 0.010T (s)

Pres

sure

(MPa

)

PT.1615IDEMO CCIDEMO

0.0

2.0

4.0

6.0

8.0

10.0

0.000 0.002 0.004 0.006 0.008 0.010T (s)

Pres

sure

(MPa

)

PT.1390IDEMO CCIDEMO

0.0

2.0

4.0

6.0

8.0

10.0

0.000 0.002 0.004 0.006 0.008 0.010T (s)

Pres

sure

(MPa

)

PT.1165IDEMO CCIDEMO

0.0

2.0

4.0

6.0

8.0

10.0

0.000 0.002 0.004 0.006 0.008 0.010T (s)

Pres

sure

(MPa

)

PT.0940IDEMO CCIDEMO

0.0

2.0

4.0

6.0

8.0

10.0

0.000 0.002 0.004 0.006 0.008 0.010T (s)

Pres

sure

(MPa

)

PT.0715IDEMO CCIDEMO

0.0

2.0

4.0

6.0

8.0

10.0

0.000 0.002 0.004 0.006 0.008 0.010T (s)

Pres

sure

(MPa

)

PT.0490IDEMO CCIDEMO

0.0

10.0

20.0

30.0

40.0

50.0

0.000 0.005 0.010 0.015 0.020T (s)

Impu

lse

(kPa

.s)

0.0

10.0

20.0

30.0

40.0

50.0

0.000 0.005 0.010 0.015 0.020T (s)

Impu

lse

(kPa

.s)

0.0

10.0

20.0

30.0

40.0

50.0

0.000 0.005 0.010 0.015 0.020T (s)

Impu

lse

(kPa

.s)

0.0

10.0

20.0

30.0

40.0

50.0

0.000 0.005 0.010 0.015 0.020T (s)

Impu

lse

(kPa

.s)

0.0

10.0

20.0

30.0

40.0

50.0

0.000 0.005 0.010 0.015 0.020T (s)

Impu

lse

(kPa

.s)

0.0

10.0

20.0

30.0

40.0

50.0

0.000 0.005 0.010 0.015 0.020T (s)

Impu

lse

(kPa

.s)

NEA/CSNI/R(2007)11

93

Figure 26. Pressures and Impulses in water with JASMINE 3.0 (JAERI) JASMI A stands for Base Case (UNI A in partner terminology) and JASMI B for Fit Case (UNI B), respectively

L-33 - Pressures and Impulses in water - JASMINE

0.0

4.0

8.0

12.0

16.0

20.0

0.000 0.002 0.004 0.006 0.008 0.010T (s)

Pres

sure

(MPa

)

PT.1615JASMINEJASMI-A CCJASMI-B CC

0.0

4.0

8.0

12.0

16.0

20.0

0.000 0.002 0.004 0.006 0.008 0.010T (s)

Pres

sure

(MPa

)

PT.1390JASMINEJASMI-A CCJASMI-B CC

0.0

4.0

8.0

12.0

16.0

20.0

0.000 0.002 0.004 0.006 0.008 0.010T (s)

Pres

sure

(MPa

)

PT.1165JASMINEJASMI-A CCJASMI-B CC

0.0

4.0

8.0

12.0

16.0

20.0

0.000 0.002 0.004 0.006 0.008 0.010T (s)

Pres

sure

(MPa

)

PT.0940JASMINEJASMI-A CCJASMI-B CC

0.0

4.0

8.0

12.0

16.0

20.0

0.000 0.002 0.004 0.006 0.008 0.010T (s)

Pres

sure

(MPa

)

PT.0715JASMINEJASMI-A CCJASMI-B CC

0.0

4.0

8.0

12.0

16.0

20.0

0.000 0.002 0.004 0.006 0.008 0.010T (s)

Pres

sure

(MPa

)

PT.0490JASMINEJASMI-A CCJASMI-B CC

0.0

40.0

80.0

120.0

160.0

200.0

0.000 0.005 0.010 0.015 0.020T (s)

Impu

lse

(kPa

.s)

0.0

40.0

80.0

120.0

160.0

200.0

0.000 0.005 0.010 0.015 0.020T (s)

Impu

lse

(kPa

.s)

0.0

40.0

80.0

120.0

160.0

200.0

0.000 0.005 0.010 0.015 0.020T (s)

Impu

lse

(kPa

.s)

0.0

40.0

80.0

120.0

160.0

200.0

0.000 0.005 0.010 0.015 0.020T (s)

Impu

lse

(kPa

.s)

0.0

40.0

80.0

120.0

160.0

200.0

0.000 0.005 0.010 0.015 0.020T (s)

Impu

lse

(kPa

.s)

0.0

40.0

80.0

120.0

160.0

200.0

0.000 0.005 0.010 0.015 0.020T (s)

Impu

lse

(kPa

.s)

NEA/CSNI/R(2007)11

94

Figure 27. Pressures and Impulses in water with TEXAS-V (KAERI-KMU and UW)

L-33 - Pressures and Impulses in water - TEXAS-V

0.0

4.0

8.0

12.0

16.0

20.0

0.000 0.002 0.004 0.006 0.008 0.010T (s)

Pres

sure

(MPa

)

PT.1615

0.0

4.0

8.0

12.0

16.0

20.0

0.000 0.002 0.004 0.006 0.008 0.010T (s)

Pres

sure

(MPa

)

PT.1390KAERI CCKAERIUW BASEUW FIT

0.0

4.0

8.0

12.0

16.0

20.0

0.000 0.002 0.004 0.006 0.008 0.010T (s)

Pres

sure

(MPa

)

PT.1165KAERI CCKAERIUW BASEUW FIT

0.0

4.0

8.0

12.0

16.0

20.0

0.000 0.002 0.004 0.006 0.008 0.010T (s)

Pres

sure

(MPa

)

PT.0940KAERI CCKAERIUW BASEUW FIT

0.0

4.0

8.0

12.0

16.0

20.0

0.000 0.002 0.004 0.006 0.008 0.010T (s)

Pres

sure

(MPa

)

PT.0715KAERI CCKAERIUW BASEUW FIT

0.0

4.0

8.0

12.0

16.0

20.0

0.000 0.002 0.004 0.006 0.008 0.010T (s)

Pres

sure

(MPa

)

PT.0490KAERI CCKAERIUW BASEUW FIT

0.0

4.0

8.0

12.0

16.0

20.0

0.000 0.005 0.010 0.015 0.020T (s)

Impu

lse

(kPa

.s)

0.0

40.0

80.0

120.0

160.0

200.0

0.000 0.005 0.010 0.015 0.020T (s)

Impu

lse

(kPa

.s)

0.0

40.0

80.0

120.0

160.0

200.0

0.000 0.005 0.010 0.015 0.020T (s)

Impu

lse

(kPa

.s)

0.0

40.0

80.0

120.0

160.0

200.0

0.000 0.005 0.010 0.015 0.020T (s)

Impu

lse

(kPa

.s)

0.0

40.0

80.0

120.0

160.0

200.0

0.000 0.005 0.010 0.015 0.020T (s)

Impu

lse

(kPa

.s)

0.0

40.0

80.0

120.0

160.0

200.0

0.000 0.005 0.010 0.015 0.020T (s)

Impu

lse

(kPa

.s)

NEA/CSNI/R(2007)11

95

Figure 28. Pressures and Impulses in water with IFCI 6.0 (KINS)

L-33 - Pressures and Impulses in water - IFCI

0.0

2.0

4.0

6.0

8.0

10.0

0.000 0.005 0.010 0.015 0.020Time (s)

Pres

sure

(MPa

)

PT.1615

0.0

2.0

4.0

6.0

8.0

10.0

0.000 0.005 0.010 0.015 0.020Time (s)

Pres

sure

(MPa

)

PT.1390

IFCI CC

IFCI

0.0

2.0

4.0

6.0

8.0

10.0

0.000 0.005 0.010 0.015 0.020Time (s)

Pres

sure

(MPa

)

PT.1165

IFCI CC

IFCI

0.0

2.0

4.0

6.0

8.0

10.0

0.000 0.005 0.010 0.015 0.020Time (s)

Pres

sure

(MPa

)

PT.0940

IFCI CC

IFCI

0.0

2.0

4.0

6.0

8.0

10.0

0.000 0.005 0.010 0.015 0.020Time (s)

Pres

sure

(MPa

)

PT.0715

0.0

2.0

4.0

6.0

8.0

10.0

0.000 0.005 0.010 0.015 0.020Time (s)

Pres

sure

(MPa

)

PT.0490IFCI CCIFCI

0.0

10.0

20.0

30.0

40.0

50.0

0.000 0.005 0.010 0.015 0.020Time (s)

Impu

lse

(kPa

.s)

0.0

10.0

20.0

30.0

40.0

50.0

0.000 0.005 0.010 0.015 0.020Time (s)

Impu

lse

(kPa

.s)

0.0

10.0

20.0

30.0

40.0

50.0

0.000 0.005 0.010 0.015 0.020Time (s)

Impu

lse

(kPa

.s)

0.0

10.0

20.0

30.0

40.0

50.0

0.000 0.005 0.010 0.015 0.020Time (s)

Impu

lse

(kPa

.s)

0.0

10.0

20.0

30.0

40.0

50.0

0.000 0.005 0.010 0.015 0.020Time (s)

Impu

lse

(kPa

.s)

0.0

10.0

20.0

30.0

40.0

50.0

0.000 0.005 0.010 0.015 0.020Time (s)

Impu

lse

(kPa

.s)

NEA/CSNI/R(2007)11

96

Figure 29. Pressures and Impulses in water with VESUVIUS (NUPEC)

L-33 - Pressures and Impulses in water - VESUVIUS

0.0

2.0

4.0

6.0

8.0

10.0

0.000 0.002 0.004 0.006 0.008 0.010T (s)

Pres

sure

(MPa

)

PT.1615VESUV CCVESUVIUS

0.0

2.0

4.0

6.0

8.0

10.0

0.000 0.002 0.004 0.006 0.008 0.010T (s)

Pres

sure

(MPa

)

PT.1390VESUV CCVESUVIUS

0.0

2.0

4.0

6.0

8.0

10.0

0.000 0.002 0.004 0.006 0.008 0.010T (s)

Pres

sure

(MPa

)

PT.1165VESUV CCVESUVIUS

0.0

2.0

4.0

6.0

8.0

10.0

0.000 0.002 0.004 0.006 0.008 0.010T (s)

Pres

sure

(MPa

)

PT.0940VESUV CCVESUVIUS

0.0

2.0

4.0

6.0

8.0

10.0

0.000 0.002 0.004 0.006 0.008 0.010T (s)

Pres

sure

(MPa

)

PT.0715VESUV CCVESUVIUS

0.0

2.0

4.0

6.0

8.0

10.0

0.000 0.002 0.004 0.006 0.008 0.010T (s)

Pres

sure

(MPa

)

PT.0490VESUV CCVESUVIUS

0.0

10.0

20.0

30.0

40.0

50.0

0.000 0.005 0.010 0.015 0.020T (s)

Impu

lse

(kPa

.s)

0.0

10.0

20.0

30.0

40.0

50.0

0.000 0.005 0.010 0.015 0.020T (s)

Impu

lse

(kPa

.s)

0.0

10.0

20.0

30.0

40.0

50.0

0.000 0.005 0.010 0.015 0.020T (s)

Impu

lse

(kPa

.s)

0.0

10.0

20.0

30.0

40.0

50.0

0.000 0.005 0.010 0.015 0.020T (s)

Impu

lse

(kPa

.s)

0.0

10.0

20.0

30.0

40.0

50.0

0.000 0.005 0.010 0.015 0.020T (s)

Impu

lse

(kPa

.s)

0.0

10.0

20.0

30.0

40.0

50.0

0.000 0.005 0.010 0.015 0.020T (s)

Impu

lse

(kPa

.s)

NEA/CSNI/R(2007)11

97

Figure 30. Pressures and Impulses in water with ESPROSE.m (UCSB)

L-33 - Pressures and Impulses in water - ESPROSE.m

0.0

4.0

8.0

12.0

16.0

20.0

0.000 0.005 0.010 0.015 0.020 0.025 0.030T (s)

Pres

sure

(MPa

)

PT.1615

0.0

4.0

8.0

12.0

16.0

20.0

0.000 0.005 0.010 0.015 0.020 0.025 0.030T (s)

Pres

sure

(MPa

)

PT.1390ESPRO CCESPROSE

0.0

4.0

8.0

12.0

16.0

20.0

0.000 0.005 0.010 0.015 0.020 0.025 0.030T (s)

Pres

sure

(MPa

)

PT.1165ESPRO CCESPROSE

0.0

4.0

8.0

12.0

16.0

20.0

0.000 0.005 0.010 0.015 0.020 0.025 0.030T (s)

Pres

sure

(MPa

)

PT.0940ESPRO CCESPROSE

0.0

4.0

8.0

12.0

16.0

20.0

0.000 0.005 0.010 0.015 0.020 0.025 0.030T (s)

Pres

sure

(MPa

)

PT.0715ESPRO CCESPROSE

0.0

4.0

8.0

12.0

16.0

20.0

0.000 0.005 0.010 0.015 0.020 0.025 0.030T (s)

Pres

sure

(MPa

)

PT.0490ESPRO CCESPROSE

0.0

4.0

8.0

12.0

16.0

20.0

0.000 0.005 0.010 0.015 0.020 0.025 0.030T (s)

Impu

lse

(kPa

.s)

0.0

40.0

80.0

120.0

160.0

200.0

0.000 0.005 0.010 0.015 0.020 0.025 0.030T (s)Im

puls

e (k

Pa.s

)

0.0

40.0

80.0

120.0

160.0

200.0

0.000 0.005 0.010 0.015 0.020 0.025 0.030T (s)

Impu

lse

(kPa

.s)

0.0

40.0

80.0

120.0

160.0

200.0

0.000 0.005 0.010 0.015 0.020 0.025 0.030T (s)

Impu

lse

(kPa

.s)

0.0

40.0

80.0

120.0

160.0

200.0

0.000 0.005 0.010 0.015 0.020 0.025 0.030T (s)

Impu

lse

(kPa

.s)

0.0

40.0

80.0

120.0

160.0

200.0

0.000 0.005 0.010 0.015 0.020 0.025 0.030T (s)

Impu

lse

(kPa

.s)

NEA/CSNI/R(2007)11

98

Figure 31. Result of calculations for FARO L-33 with IDEMO assuming different vapour volume fractions in the mixing zone

NEA/CSNI/R(2007)11

99

Figure 32. FARO L-33 predicted histories of kinetic energy - full calculations (experimental evaluation is 0.110 MJ)

L-33. History of kinetic energy

0.0

0.2

0.4

0.6

0.8

1.0

0 5 10 15 20 25 30Time (ms)

Kin

etic

ene

rgy

(MJ)

JAERI VESUVIUSTEXAS KAERI TEXAS UW BASETEXAS UW FITIFCIIDEMO MC3D 0.7MC3D 1.13ESPROSE

(Experimental evaluation 0.110 MJ)L-33. History of kinetic energy

0.0

0.2

0.4

0.6

0.8

1.0

0 5 10 15 20 25 30Time (ms)

Kin

etic

ene

rgy

(MJ)

JAERI VESUVIUSTEXAS KAERI TEXAS UW BASETEXAS UW FITIFCIIDEMO MC3D 0.7MC3D 1.13ESPROSE

(Experimental evaluation 0.110 MJ)

NEA/CSNI/R(2007)11

100

Figure 33. Calculated reactor situations

In-vessel Ex-vesselIn-vessel Ex-vessel

NEA/CSNI/R(2007)11

101

Figure 34. Geometry and nodalisation used for in-vessel case simulation (not to scale)

MC3D (CEA-IRSN) MATTINA (FZK) IFCI (KINS)

Melt ejection calculated 1 central jet, 2 annular jets. See Table 14

Melt ejection calculated 1 central jet, 2 annular jets. See Table 14

Melt ejection calculated 1 central jet, 2 annular jets. See Table 14

NEA/CSNI/R(2007)11

102

Figure 34. Geometry and nodalisation used for in-vessel case simulation (not to scale) (Cont’d)

IKEMIX (IKE) JASMINE (JAERI) PM-ALPHA (UCSB)

Melt ejection prescribed. See Table 14 Melt ejection prescribed. See Table 14 Melt ejection prescribed. See Table 14

3-D uniform grid size dx=dy=dz= 15cm

NEA/CSNI/R(2007)11

103

Figure 34. Geometry and nodalisation used for in-vessel case simulation (not to scale) (Cont’d)

TEXAS-V (UW) TEXAS-V (KAERI-KMU)

Melt ejection prescribed. See Table 14 1-D simulation

Melt ejection prescribed. See Table 14 1-D simulation

NEA/CSNI/R(2007)11

104

Figure 35. Maximum pressure calculated at vessel bottom (position indicated into brackets)

In-vessel. Maximum pressure at bottom

0.0

20.0

40.0

60.0

80.0

100.0

120.0

0.0 4.0 8.0 12.0 16.0 20.0Time (ms)

Pres

sure

(MPa

)

JASMINE F1 (r=0.1)

VESUVIUS (trigger cell)

IFCI (centre)

TEXAS UW (G1)

TEXAS KK C1

TEXAS KK C2

MATTINA (centre)

IDEMO (r=0.007)

IDEMO (r=0.908)

MC3D T2 (centre)

MC3D PSA (centre)

ESPROSE (centre)

NEA/CSNI/R(2007)11

105

Figure 36. Maximum impulse calculated at vessel bottom (position indicated into brackets)

In-vessel. Maximum impulse at bottom

0.0

50.0

100.0

150.0

200.0

250.0

0.0 4.0 8.0 12.0 16.0 20.0Time (ms)

Impu

lse

(kPa

.s)

JASMINE F1 (r=0.1)

IFCI (centre)

TEXAS UW (G1)

TEXAS KK C1

TEXAS KK C2 )

IDEMO (r=0.007)

IDEMO (r=0.241)

MC3D T2 (centre)

MC3D PSA (centre)

ESPROSE (centre)

MATTINA ( centre)

NEA/CSNI/R(2007)11

106

Figure 37. In-vessel axial distributions of radially averaged void fraction at trigger time (radial averages made over a diameter of 1 m)

In-vessel. Void fraction radially averaged atmelt bottom contact

0.0

1.0

2.0

3.0

4.0

0.0 0.2 0.4 0.6 0.8 1.0Fraction

Elev

atio

n (m

)

JASMINE P1

IKE IKEMIX

TEXAS UW

TEXAS KK C1

TEXAS KK C2

IFCI

MC3D T2

MC3D PSA

PM-ALPHA

NEA/CSNI/R(2007)11

107

Figure 38. In- vessel global component fractions at melt-bottom contact averaged over a cylinder of height the water depth, and diameter 1 m. Ordinate: component fraction in %; Abscissa: codes.

PM

-A /

ES

PR

OS

E

MC

3D -

T2

MC

3D -

PS

A

IFC

I

IKE

MIX

/ I

DE

MO

JAS

MIN

E

TE

XA

S U

W

TE

XA

S K

K

MA

TT

INA

0

1 0

2 0

3 0

4 0

5 0

6 0

Volu me fractio ns

Melt Drop Fra ctionVoid fra ction

NEA/CSNI/R(2007)11

108

Figure 39. In-vessel calculated impulses as a function of global void fraction at trigger time

0

0.5

1

1.5

2

2.5

0 10 20 30 40 50 60

Vo id fr actio n

Imp

uls

e at

bot

tom

(b

ar.s

)

PM-A / ESPROSE

MC3D - T2

MC3D - PSAIFCI

IKEMIX / IDEMO

JASMINETEXAS UWTEXAS KK

MA TTINA

NEA/CSNI/R(2007)11

109

Figure 40 a. In-vessel evolution of fragmented melt mass during explosion

In-vessel. Fragmented melt mass during explosion

0.0

400.0

800.0

1 200.0

1 600.0

2 000.0

0 5 10 15 20 25 30Time (ms)

Frga

men

ted

mas

s (k

g) JASMINE F1

TEXAS UW

TEXAS KK C1

TEXAS KK C2

IDEMO

MC3D T2

MC3D PSA

MATTINA

NEA/CSNI/R(2007)11

110

Figure 40 b. In-vessel evolution of fragmented melt mass during explosion (reduced mass scale)

In-vessel. Fragmented melt mass during explosion

0.0

10.0

20.0

30.0

40.0

50.0

0 5 10 15 20 25 30Time (ms)

Frga

men

ted

mas

s (k

g) JASMINE F1

TEXAS UW

TEXAS KK C1

TEXAS KK C2

IDEMO

MC3D T2

MC3D PSA

MATTINA

NEA/CSNI/R(2007)11

111

Figure 41. Geometry and nodalisation used for ex-vessel case simulation (not to scale)

MC3D (CEA-IRSN) MATTINA (FZK) IFCI (KINS)

Melt ejection calculated. Melt ejection calculated. Melt ejection calculated.

NEA/CSNI/R(2007)11

112

Figure 41. Geometry and nodalisation used for ex-vessel case simulation (not to scale) (Cont’d)

IKEMIX (IKE) JASMINE (JAERI) PM-ALPHA (UCSB) TRACER (KAERI-KMU)

Water

Steam

Melt

V

2.75 m

4 m

5 m

zr

Melt ejection prescribed Melt ejection prescribed Melt ejection prescribed 3-D uniform grid size dx=dy=dz= 15cm

Melt ejection prescribed dr = 0.275 m, dz = 0.25 m

NEA/CSNI/R(2007)11

113

Figure 41. Geometry and nodalisation used for ex-vessel case simulation (not to scale) (Cont’d)

VESUVIUS (NUPEC) TEXAS-V (UW) TEXAS-V (KAERI-KMU)

Melt ejection prescribed Melt ejection prescribed

1-D simulation Melt ejection prescribed 1-D simulation

NEA/CSNI/R(2007)11

114

Figure 42 a. Ex-vessel maximum pressure calculated at cavity bottom (position indicated into brackets)

Ex-vessel. Maximum pressure at bottom

0.0

20.0

40.0

60.0

80.0

100.0

120.0

0.0 4.0 8.0 12.0 16.0 20.0Time (ms)

Pres

sure

(MPa

)

JASMINE F1 (r=0.2)

JASMINE F1X (r=0.2)

JASMINE F2 (r=0.2)

VESUVIUS (trigger cell)

IFCI (centre)

TEXAS UW (G1)

TEXAS KK C1

TEXAS KK C3

MATTINA (r=0.56)

IDEMO(r=0.025)

MC3D T2 (centre)

MC3D PSA (centre)

ESPROSE (centre)

NEA/CSNI/R(2007)11

115

Figure 42 b. Ex-vessel maximum pressure calculated at cavity bottom (reduced pressure scale)

Ex-vessel. Maximum pressure at bottom

0.0

5.0

10.0

15.0

20.0

25.0

30.0

0.0 4.0 8.0 12.0 16.0 20.0Time (ms)

Pres

sure

(MPa

)

VESUVIUS (trigger cell)

MATTINA (r=0.56)

MC3D PSA (centre)

NEA/CSNI/R(2007)11

116

Figure 43 a. Ex-vessel maximum impulse calculated at cavity bottom (position indicated into brackets)

Ex-vessel. Maximum impulse at bottom

0.050.0

100.0150.0200.0250.0300.0350.0400.0

0.0 4.0 8.0 12.0 16.0 20.0Time (ms)

Impu

lse

(kPa

.s)

JASMINE F1 (r=0.2)

JASMINE F1X (r=0.2)

IFCI (centre)

TEXAS UW (G1)

TEXAS KK C1

TEXAS KK C3

IDEMO(r=0.025)

MC3D T2 (centre)

MC3D PSA (centre)

ESPROSE (centre)

MATTINA (r=0.56)

NEA/CSNI/R(2007)11

117

Figure 43 b. Ex-vessel maximum impulse calculated at cavity bottom (reduced impulse scale)

Ex-vessel. Maximum impulse at bottom

0.0

20.0

40.0

60.0

80.0

100.0

0.0 4.0 8.0 12.0 16.0 20.0Time (ms)

Impu

lse

(kPa

.s)

JASMINE F1 (r=0.2)

JASMINE F1X (r=0.2)

IFCI (centre)

TEXAS UW (G1)

TEXAS KK C1

TEXAS KK C3

IDEMO(r=0.025)

MC3D T2 (centre)

MC3D PSA (centre)

ESPROSE (centre)

MATTINA (r=0.56)

NEA/CSNI/R(2007)11

118

Figure 44 a. Ex-vessel maximum pressure calculated at cavity wall (position indicated into brackets)

Ex-vessel. Maximum pressure at wall

0.0

10.0

20.0

30.0

40.0

50.0

0.0 4.0 8.0 12.0 16.0 20.0Time (ms)

Pres

sure

(MPa

)

VESUVIUS (z=0.3m)

IFCI (z=3.2 m)

MATTINA (z=1.95m)

IDEMO (z=0.48m)

MC3D T2 (bottom)

MC3D PSA (bottom)

ESPROSE (z=0.25m)

TRACER (0.5 m)

NEA/CSNI/R(2007)11

119

Figure 44 b. Ex-vessel maximum pressure calculated at cavity wall (reduced pressure scale)

Ex-vessel. Maximum pressure at wall

0.0

2.0

4.0

6.0

8.0

10.0

0.0 4.0 8.0 12.0 16.0 20.0Time (ms)

Pres

sure

(MPa

)

IFCI (z=3.2 m)MATTINA (z=1.95m)

MC3D PSA (bottom)TRACER (0.5 m)

NEA/CSNI/R(2007)11

120

Figure 45 a. Ex-vessel maximum impulse calculated at cavity wall (position indicated into brackets)

Ex-vessel. Maximum impulse at wall

0.0

50.0

100.0

150.0

200.0

0.0 5.0 10.0 15.0 20.0 25.0 30.0Time (ms)

Impu

lse

(kPa

.s)

VESUVIUS (z=0.3m)

IFCI (z=3.2m)

IDEMO (z=0.48m)

MC3D T2 (z=0)

MC3D PSA (z=0)

ESPROSE (z=0.25m)

TRACER (z=0.5m)

MATTINA (r=1.95 m)

NEA/CSNI/R(2007)11

121

Figure 45 b. Ex-vessel maximum impulse calculated at cavity wall (reduced impulse scale)

Ex-vessel. Maximum impulse at wall

0.0

20.0

40.0

60.0

80.0

100.0

0.0 5.0 10.0 15.0 20.0 25.0 30.0Time (ms)

Impu

lse

(kPa

.s)

VESUVIUS (z=0.3m)

IFCI (z=3.2m)

IDEMO (z=0.48m)

MC3D PSA (z=0)

TRACER (z=0.5m)

MATTINA (z=1.95 m)

NEA/CSNI/R(2007)11

122

Figure 46. Ex-vessel axial distributions of radially averaged void fraction at trigger time (radial averages made over a diameter of 1 m)

Ex-vessel. Void fraction radially averaged attrigger time

0.00.5

1.01.5

2.02.5

3.03.54.04.55.0

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0Fraction

Elev

atio

n (m

)

JASMINE P1IKE IKEMIXTEXAS UWTEXAS KK C1TEXAS KK C3TRACERIFCIMC3D T2MC3D PSAPM-ALPHA

NEA/CSNI/R(2007)11

123

Figure 47. Ex-vessel axial distributions of radially averaged melt fraction at trigger time (radial averages made over a diameter of 1 m)

Ex-vessel. Melt fraction radially averaged attrigger time

0.00.51.01.52.02.53.03.54.04.55.0

0.00 0.05 0.10 0.15 0.20 0.25 0.30Fraction

Elev

atio

n (m

)

JASMINE P1IKEMIX (J+D)IKEMIX (D)TEXAS UWTEXAS KK C1TEXAS KK C3TRACERIFCIMC3D T2 (M+D)MC3D T2 (D)MC3D PSA (M+D)MC3D PSA (D)PM-ALPHA

NEA/CSNI/R(2007)11

124

Figure 48. Ex- vessel global component fractions at melt-bottom contact averaged over a cylinder of height the water depth, and diameter 1 m. Ordinate: component fraction in %; Abscissa: codes

PM

-A /

ESP

RO

SE

MC

3D

- T

2

MC

3D

- P

SA

IFC

I

IKE

MIX

/ ID

EM

O

JAS

MIN

E

TE

XA

S U

W

TE

XA

S K

K

TR

AC

ER

MA

TT

INA

VE

SU

VIU

S

0

10

20

30

40

50

60

70

80

Volume fractions

Melt Drop FractionVoid fraction

NEA/CSNI/R(2007)11

125

Figure 49. Ex-vessel calculated impulses as a function of global void fraction at trigger time

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

0 10 20 30 40 50 60 70 80

Void fraction

Imp

uls

e at

bot

tom

(b

ar.s

)

PM-A / ESPROSE

MC3D - T2

MC3D - PSA

IFCIIKEMIX / IDEMO

JASMINE

TEXAS UW

TRACER

NEA/CSNI/R(2007)11

126

Figure 50. Ex-vessel fragmented melt mass during explosion

Ex-vessel. Fragmented melt mass during explosion

0.0

300.0

600.0

900.0

1 200.0

1 500.0

0 5 10 15 20 25 30Time (ms)

Frga

men

ted

mas

s (k

g)

JASMINE F1

JASMINE F1X

VESUVIUS

TEXAS UW

TEXAS KK C1

TEXAS KK C3

IDEMO

MC3D T2

MC3D PSA

MATTINA