TITRE DE LA PRESENTATION

1
European Review Meeting on Severe Accident
Research June 12-14, 2007 Karlsruhe (Germany)
Preliminary analyses of the Phebus FPT3
experiment using Severe Accident Codes
(ATHLET-CD, ICARE/CATHARE, MELCOR) Georges
Repetto, Olivier De Luze (IRSN), Tilman Drath,
Thorsten Hollands, Marco K. Koch
(RUB-LEE), Christine Bals, Klaus Trambauer,
Henrique Austregesilo (GRS) Jon Birchley (PSI),

2
OUTLINE Introduction Modelling and
boundary conditions Code results
Thermal evolution Hydrogen release
Control rod rupture B4C oxidation and gas
releases Bundle Degradation
Conclusion
3
Introduction
Pre test calculation were performed using
several S. A codes such as ICARE2 (IRSN),
ATHLET-CD (GRS) and MELCOR (PSI and JNES)
This benchmark was published in the NUCLEAR
TECHNOLOGY (Feb.07)
This presentation is devoted to FPT3 post test
calculations performed as a Joint Activity, by

• IRSN (ICARE2)
• GRS
• RUB
• PSI (MELCOR)

(ATHLET-CD)
FPT3
in the framework of the Phebus Bundle
Interpretation Circle and the SARNET CORIUM WP
9.2 early phase and B4C effects
4
Part I Brief review of the codes modelling and
the FPT3 boundary conditions
5
Codes and modelling options
6
Modelling options (Zry oxidation)
Best fitted coming from the COLOSS project
MELCOR PSI
ICARE2 IRSN
ATHLET-CD GRS-RUB
7
Codes and modelling options
8
Modelling options (B4C-SS interaction)
ICARE2 IRSN
ATHLET-CD RUB
ATHLET-CD GRS
domain
9
Codes and modelling options
10
Modelling options B4C oxidation (1/2)
ATHLET-CD GRS
ATHLET-CD RUB
11
Modelling options B4C oxidation (1/2)
For ICARE2 (IRSN) the kinetics of solid B4C
oxidation based on IRSN (VERDI) and FzK (BOX)
experiments
(mole/m2/s)
Steinbrück et al. (FZKA 6979), for ATHLET-CD
FzK
VERDI
(119 experiments)
This work performed in the frame of the WP9.2
activities
12
FPT3 boundary conditions (versus FPT2)
Steam flow rate
Oxidation phase
Heat up phase
Calibration phase
Similar Bundle power transient as FPT2 except for
the P4 Power plateau and the maximum power (which
is lower of about 27 ) Constant steam mass
flowrate (0.5 g/s) (as for FPT2)
13
FPT3 boundary conditions (Bundle Power transient)
Test
Code input deck
P4
As for previous Phebus tests, to achieve good
thermal results during the calibration phase and
the P4 plateau, the bundle power has been reduced
(-15 to -25)
14
Part II Results of the simulations
15
Thermal results temperatures evolution (1/3)
Assuming the bundle power adjustment, rather good
thermal results were obtained up to the end of
the transient
0.5 m elevation
Test (TUS)
Test (Tc)
bundle
Test (Tc)
Shroud
16
Thermal results temperatures evolution (2/3)
Rather good thermal results with however
overestimation of temperature during oxidation
0.7 m elevation
C
Code overestimation
Test
bundle
Shroud
17
Thermal results temperatures evolution (3/3)
An other elevation shows the same behavior which
seems to be linked to overestimation of the steam
starvation duration given by the codes
C
2600 K
0.6 m elevation
Code overestimation
bundle
Test
Shroud
18
Hydrogen generation (1/2)
H2 (g)
Rather good results obtained with both code
simulations regarding the total hydrogen
generation
The lower value produced with MELCOR is due to
the under- estimation of the H2 during the main
oxidation period (mainly coming from B4C)
experiment
6 g from the sample
19
Hydrogen generation (2/2)
360 s
The total H2 produced is correct Nevertheless,
the rather short starvation period observed in
the experiment is not simulated by both codes

• the pure starvation period during the test was
  confirmed to be shorter than what was foreseen by
  pre-calculations
• (1200 /-200 s)

20
Control Rod rupture (1/2)
ATHLET-CD
ICARE2
B4C-SS interaction
Rupture of the Control rod at t 9490 s / 9370 s
values consistent with the experiment
observation (9550 s)
In the simulation, it happens when the SS clad is
completely dissolved by interaction with B4C (in
ICARE2 and ATHLET-CD)
(no specific model for the Guide Tube rupture in
both S.A codes)
21
Control Rod rupture (2/2)
ATHLET-CD (GRS)
0.3 kg
Control rod rupture
Clad rupture
Metallic melt and crust from control rod
(B,C, SS, Zry) is relocated to lower elevations
with ICARE2, whereas, it remained in place with
ATHLET-CD code This should increase the oxidation
of the Control rod materials in ATHLET-CD versus
ICARE2 results
Example of mass amount involved in the Control
rod degradation with ATHLET-CD code
22
B4C oxidation and Hydrogen generation
ATHLET-CD (GRS)
ICARE2
H2
70
15
CR Rupture
About 15 solid B4C remained intact, which is
consistent with the experiment
23
Gas release coming from B4C oxidation (test
results) Carbonaceous species
0.64 mole of CO
mainly during the starvation phase most
of the release during the heat up phase at low H2
concentration
0.37 mole of CO2
24
Gas release Comparison with the code simulations
experiment
1 mole
1 mole
CO (0.64 mole)
B4C 7H2O(g) ? 2B2O3 CO(g) 7H2(g)
H2
1 mole
1 mole
CO2 (0.37 mole)
B4C 8H2O(g) ? 2B2O3 CO2(g) 8H2(g)
Taking into account the most probable reaction
1 mole of Carbon produced during the experiment
should come from 1 mole of oxidised B4C
At the end of the test, carbon release by codes
(0,95 to 1.11 mole) is consistent with the test
result (roughly 1 mole) Nevertheless, kinetics
have to be improved
experiment
MELCOR produces only 0.2 g of CO (part of the
reacted B4C lead to elemental carbon not account
for in the curve equilibrium chemistry model in
SOLGASMIX model
MELCOR
25
Bundle degradation
Clad fuel
Early bundle degradation due to fuel rods /
B,C,SS interaction is suspected Not accounted
for by S.A codes
ICARE2
ATHLET-CD (GRS) predicted bundle degradation due
to higher temperature at the end of the transient
ATHLET-CD
1.6 kg
26
Thermal results temperatures evolution
Slight temperature over estimation during the
heat up phase, in the ATHLET-CD (GRS) case
leading to fuel degradation at 2600 K
C
0.6 m elevation
Code overestimation
bundle
Test
Shroud
27
Conclusion

• FPT3 post test calculations have been performed
  using 3 different codes (ICARE2 , ATHLET-CD and
  MELCOR)
• Overall thermal behaviour well reproduced.
• Cladding temperature during oxidation are
  overestimated.

• Total amount of hydrogen production is correct
  but not the kinetics.
• Starvation duration not well reproduced by both
  S.A codes.

• Kinetics of solid B4C oxidation with a suitable
  correlation in the codes (ICARE2 and ATHLET-CD)
  and gas release are rather well predicted
• Modelling effort is foreseen in MELCOR (only
  adapted for BWR)

• No significant fuel degradation is predicted by
  those S.A codes
• Early bundle degradation is suspected due to
  interaction between CR materials and neighbouring
  fuel rods
• Modelling effort on this area has to be done in
  both S.A codes

28
Thank you for your attention ..
Regarding ATHLET-CD analyses performed by RUB and
GRS, the authors give specific acknowledgement to
the German Federal Ministry of Economics and
Technology (BMWi ) for their sponsoring.