Results of Phase 1 of OECD Programme SERENA Steam Explosion REsolution for Nuclear Applications

1
Results of Phase 1 of OECD Programme SERENA
(Steam Explosion REsolution for Nuclear
Applications) 

• Presented by
• D. MagallonProgramme coordinator

2
Participants

• K-H Bang, Korea Maritime University, Korea,
• S. Basu, Nuclear Regulatory Commission, USA,
• G. Berthoud, Commissariat à l'Énergie Atomique,
  France,
• M. Bürger, Institute für Kernenergetik und
  Energiesysteme, Germany,
• M.L. Corradini, University of Wisconsin, Madison,
  USA,
• H. Jacobs, Forschungszentrum Karlsruhe, Germany,
• R. Meignen, Institut de Radioprotection et de
  Sûreté Nucléaire, France,
• O. Melikhov, Electrogorsk Research and
  Engineering Centre, Russia,
• K. Moriyama, Japan Atomic Energy Research
  Institute, Japan,
• M. Naitoh, Nuclear Power Engineering Corporation,
  Japan,
• J-H. Song, Korea Atomic Energy Research
  Institute, Korea,
• N. Suh, Korea Institute of Nuclear Safety, Korea,
  
• T.G. Theofanous, University of California, Santa
  Barbara, USA.

3
Scope of SERENA

• Bring together FCI experts to make a status of
  code capabilities to predict steam explosion
  induced dynamic loading of the reactor structures
  (first joint exercise since ISP-39 on pre-mixing
  in 1996)
• Reach consensus on important FCI phenomena for
  reactor application
• Carry out confirmatory research required to
  reduce uncertainties on these phenomena

4
Objectives of the programme

• Phase 1
• Identify areas where discrepancies of calculation
  results by the different FCI codes induce large
  uncertainties in the prediction of loads in
  reactors
• Identify additional research needed to increase
  the level of confidence of the predictions
• Phase 2
• Reduce these uncertainties to acceptable level
  for risk assessment by carrying out confirmatory
  analytical and/or experimental research

5
Organisation Phase 1

• Coordinated action
• Coordinator financed by CEA and IRSN
• Each partner finances his own work
• 1.2 p.y for the planned 3.5 year duration of the
  programme
• Technical Committee in charge of
• Assisting the Programme Coordinator
• Advising the programme on all matters it judges
  appropriate to insure that the objectives will be
  reached in due time
• Exchange with existing national programmes
• Discuss results and make suggestions for test
  matrix
• Started with TROI and will be extended to other
  programmes (ECO, KROTOS).

6
Phase 1 Tasks

• Task 1 Identification of relevant conditions for
  FCI in NPP's and selection of relevant
  experiments
• Task 2 Code application to relevant pre-mixing
  experiments
• Task 3 Code application to relevant explosion
  experiments
• Task 4 Code application to reactor
  configurations
• Task 5 Synthesis of the findings of Phase 1 and
  proposal to move forward, if appropriate

7
Overall schedule for Phase 1

• Started January 2002
• Final meeting June 2005
• Reports a CD containing all the reports (one by
  task) will be issued beginning 2006

8
Codes used in Phase 1
9
Methodology

• Participants were given same sets of initial
  conditions and reference data
• They translated these initial conditions into
  adequate inputs for their codes
• Explosion phases of the experiments were
  calculated both for imposed and calculated
  pre-mixing
• Participants were let free to set model options
  and parameters as they used to, but had to
  document their choices and possibly make
  sensitivity calculations
• Comparison was made on a set of pre-established
  quantities

10
Relevant reactor conditions (T1)

• Ex-vessel large pour of corium melt with metallic
  phases in sub-cooled water at low pressure
• In-vessel multi-jets (10 cm diameter) into lower
  head and saturated water at moderate pressure
• Selected as generic situations of most interest

11
Calculated pre-mixing experiments (T2)

• Integral experiment FARO L-33 (JRC)
• Triggered steam explosion after penetration of
  25 kg of UO2-ZrO2 melt jet in 1.6-m-depth
  sub-cooled water at 0.4 MPa
• FARO-28 (JRC)
• Quenching of 175 kg of UO2-ZrO2 melt jet in
  1.4-m-depth saturated water at 0.5 MPa (jet
  duration 6 s)
• FARO L-31 (JRC)
• Quenching of 92 kg of UO2-ZrO2 melt jet in
  1.4-m-depth subcooled water at 0.2 MPa
• PREMIX-16 (FZK)
• Pre-mixing of 60 kg of alumina melt in
  1.3-m-depth saturated water at 0.5 MPa (similar
  to FARO L-28)

12
FARO Experimental arrangement
13
PREMIX Experimental arrangement
14
Data to be provided for pre-mixing (1)

• Code version
• Code description, in particular
• Models and computational methods for jet-break-up
  and mixing
• Values of the various parameters used and their
  justification
• For test data/code comparison
• Melt progression rate
• Vessel pressure history
• Level swell history
• Energy release to steam and water as a function
  of time
• Final particle size distribution and mean size
• Final fragmented melt area

15
Data to be provided for pre-mixing (2)

• For code/code comparison
• Nodalisation
• Vaporisation/condensation rate
• Axial component distribution radially averaged
• Axial component distribution at centreline
• Particle size distribution at melt-bottom contact
• Melt surface area history
• 2-D contours of pre-mixing area based on 1 void
  fraction every 100 ms from 0 to 1 s
• One 2-D contour at 4s for L-28

16
Results on pre-mixing (1)
17
Results on pre-mixing (2)
18
Results on pre-mixing (3)
19
Void predictions for FARO L-28
20
Example of pre-mixing evolution
21
Summary results on pre-mixing (T2)

• Physics and robustness of codes have been
  improved since ISP-39 ? Low pressure cases can
  be calculated
• Differences between code predictions and between
  code predictions and data are large
• Tendency to underestimate heat transfer and
  overestimate void when compared to data
• Main cause description of flow regime
• Melt length-scale evolution
• Non-equilibrium phase change
• Heat transfer mechanisms (radiation)

22
Calculated explosion experiments (T3)

• FARO L-33
• Imposed and calculated pre-mixing conditions
• KROTOS K-44 (JRC)
• Triggered steam explosion with 1.4 kg of alumina
  melt and 10K-subcooled water at 0.1 MPa.
• TROI-13 (KAERI)
• Spontaneous steam explosion with 1.14 kg of
  UO2-ZrO2 melt and 80K-subcooled water at 0.1 MPa
• TROI blind (KAERI)
• Triggered steam explosion with UO2-ZrO2 melt and
  50K-subcooled water at 0.1 MPa

23
KROTOS and TROI set up
TROI test assembly
KROTOS 44 test section
24
Data to be provided for explosion (1)

• Code version
• Code description, in particular
• Models and computational methods for jet-break-up
  and mixing
• Values of the various parameters used and their
  justification
• For test data/code comparison
• Explosion pressures at vessel wall in water
• Vessel pressure history
• Debris characteristics

25
Data to be provided for explosion (2)

• For code/code comparison
• Nodalisation
• Pressure distributions at different times
• Distribution of volume fractions as a function of
  time
• Energy transfer to steam and water
• Evolution of melt surface area
• Evolution of fragmented mass
• History of kinetic energy

26
Results on explosion (1)
27
Results on explosion (2)
28
Results on explosion (3)
29
K-44 explosion pressure history by IDEMO
30
F-33 explosion pressure history by IDEMO
31
Summary of results on explosion (T3)

• Differences between code predictions are large
• In order to get the order of magnitude of the
  data, key effects such as heat transfer and
  fragmentation parameters had to be (more or less
  arbitrarily) reduced from those tuned for alumina
• Possible physical explanations for the
  experimentally observed reduced explosion
  energetics are
• Melt freezing during premixing
• Hydrogen production during pre-mixing

32
Calculated reactor situations (T4)
In-vessel
Ex-vessel
33
Data to be provided for reactor calc. (1)

• All necessary information on how the initial and
  boundary conditions are treated
• Values of parameters and justification
• Nodalisation
• Melt progression data, melt in water at trigger
  times
• Axial distributions of radially averaged
  component fractions at trigger times
• History of melt surface area       

34
Data to be provided for reactor calc. (2)

• History of melt surface area
• Particle distribution at trigger times
• Vaporisation/condensation rates
• History of energy release to steam and water
• History of fragmented melt mass during explosion
• History of kinetic energy
• Impulse (and pressure) at bottom wall and maximum
  calculated values for both in- and ex-vessel
  cases

35
In-vessel impulses
36
Ex-vessel impulses
37
Ex-vessel fragmented melt mass
38
In-vessel pre-mixing with MC3D
39
In_vessel impulses and void fraction
40
Ex-vessel pre-mixing with MC3D
41
Ex-vessel impulses and void fractions
42
Summary results on reactor cases (T4)

• Whatever the modelling and numerical approaches,
  all the codes were able to calculate the reactor
  situations of concern.
• There is still a tendency to predict large void
  in premixing, that was judged to be an
  overestimation of voiding in Task 2
• While the scatter of the results is large for
  both pre-mixing and explosion, the results
  indicate clear tendencies
• For the in-vessel case, the calculated loads were
  far below the capacity of a typical intact vessel
• For the ex-vessel case, the calculated loads were
  above the capacity of typical cavity walls (no
  specific design considered).
• However
• Only one case has been calculated for in-vessel
  and ex-vessel, respectively,that might not be the
  worst possible.
• In some cases reduced parameters have been used
  to model the explosion without firm physical
  reasoning. 
• Only a few variations of the parameters have been
  performed

43
Overall Conclusions (1)

• First international exercise to compare steam
  explosion modelling approaches and calculation
  results for reactor applications
• There were strong discussions on the FCI issues
  that are relevant for reactor situations
• Different views were expressed regarding some
  issues
• Dominant break-up mechanisms during pre-mixing,
  especially during melt fall into water up to melt
  front bottom contact
• Need to calculate jet (and particle) break-up for
  reactor application.
• These issues could not be resolved in the frame
  of Phase 1.
• On the basis of available experiments, no
  definite conclusions could be drawn on the
  reasons for the reduced energetics observed for
  corium melts with respect to alumina melts

44
Overall Conclusions (2)

• It was recognised that further checking of the
  codes by performing separate effect calculations
  would be helpful to improve understanding of the
  importance of major phenomena on the explosion
  strength.
• But it was recognised also that this is a hard
  task to perform because compensating effects
  would inevitably be present and obscure the
  conclusions.
• In spite of the variety of modelling and
  approaches (and of large pre-mixed masses), the
  predicted explosion pressures were relatively low
  for the reactor situations considered in the
  exercise
• The major reason for this was considered to be
  the high void calculated in regions containing
  relevant amounts of melt

45
Overall Conclusions (3)

• A clarification about the impact of void on
  explosion strength is necessary.
• e.g., parametric calculations with variation of
  void could be done to identify the level of void
  which yields sufficient reduction of explosion
  strength. These calculations were not
  consistently made in Phase 1.
• Summarising, it was agreed that the major issues
  that limit the confidence in the results are
• The large scatter of the results,
• The uncertainties on the pre-mixing flow
  patterns, especially on void and pre-mixed melt
  mass predictions (missing detailed data on these
  issues),
• The missing physical justification for the use of
  reduced heat transfer and fragmentation
  parameters for corium melts in the explosion
  phase thereby affecting energetics.

46
Overall Conclusions (4)

• Consequently, actions should be undertaken to
• Confirm characteristics of corium energetics
  (material effect)
• Reduce uncertainties on void predictions
• Reduce the spread of the predictions
• These issues are proposed to be investigated in a
  second phase with the support of an experimental
  programme of confirmatory nature using KROTOS and
  TROI facilities (next presentation)