Jiří Duspiva Nuclear Research Institute Řež, plc. Nuclear Power and Safety Division Dept. of Reactor Technology 11 th International QUENCH Workshop Karlsruhe,

Jiří DuspivaNuclear Research Institute Řež, plc.Nuclear Power and Safety Division

Dept. of Reactor Technology

11th International QUENCH WorkshopKarlsruhe, Germany, October 25-27, 2005

Conclusions from Conclusions from

Quench-03 Test Analyses with Quench-03 Test Analyses with

ICARE2 and MELCOR CodesICARE2 and MELCOR Codes

J. Duspiva 11th International QUENCH Workshop Karlsruhe, Germany, October 25-27, 2005

2

Background of Quench analysis in NRI Řež

Quench-03 test analysis with ICARE2(Short summary of 10th QWS presentation)

•Comparison to MELCOR 1.8.5 Improved model of Quench-03

•Comparison to MELCOR 1.8.5

•Regressive application to Quench-01 Summary and conclusions

Outline


3

Backgroundof NRI Quench Analyses

First NRI – Quench activities were performed with MELCOR Code Own MELCOR input model was developed for Quench-01 [1],

successfully also applied to Quench-06 [ISP-45 Blind phase] calculations Quench-03 Calculation was last with MELCOR 1.8.5 [2]

• Temperature at onset of reflooding higher than Q-01 and Q-06

• Temperature escalation was not predicted correctly

• Strong underestimation of Hydrogen production Under EU Project SARNET in TPA1/JPA1 (WP9: Early Phase of Core

Degradation ST-1 Hydrogen Generation during Core Reflooding) started application of ICARE2 Code with following schedule [3] and [4]

• Step 1 – Preparation of model and calculation of Quench-01 test

• Step 2 - Sensitivity Study on the change of important parameters

• Step 3 – Calculation of Quench-03 test

[1] J. Duspiva: Quench-01 Test Calculation with MELCOR Code, CSARP Meeting, May 7-9, 2001, Bethesda, Maryland [2] J. Duspiva: Quench-03 Test Calculation with MELCOR Code, 8 th International Quench Workshop, Karlsruhe, Germany,

October 29-31, 2002[3] J. Duspiva: Quench Test Calculations with ICARE2 Code and Comparison with MELCOR Code Results, 10 th International

Quench Workshop, Karlsruhe, Germany, October 26-28, 2004[4] J. Duspiva: Quench Test Calculations with ICARE2 Code and Comparison with MELCOR Code Results (Quench-01 and

Quench-03 Test Analyses), Report NRI Řež, UJV-12204-T, March 2005


4

ICARE2 Input ModelQuench Facility Nodalization

Original model, received from L. Belovsky (ALIAS CZ), was prepared by G. Bandini (ENEA)

Macrocomponents• Unheated rod

• 8 heated rods of inner ring

• 12 heated rods of outer ring

• 3 Corner rods (withdrawal of one corner rod neglected)

• Grid spacers and

• Shroud with gap above heated zone

• One TH channel – FLUID2 type 41 axial levels

• 4 below heated part

• 21 in heated part

• 16 upper plenum First test on Quench-01

• Sensitivity Matrix Improvements

1.5

1.4

1.3

1.2

1.1

1.0

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0.0

-0.2

-0.3

-0.4

-0.47

CLAD1

PZO1

PZO2

MO2B

I E2B

CLAD2

TUNG2

MO2A

I E2A

CU2A

MO3A

I E3A

CU3A

TUNG3

CLAD3

I E3B

MO3B

PZO3

BZI RR

BZI R

BSS

BZF5

BZF4

BZF3

BZF2

BZF1

Central Unheated Rod

I nner Heated Ring Rod

Outer Heated Ring Rod

Shroud Wall

Ax.L. 41

Ax.L. 40

Ax.L. 39 Ax.L. 38

Ax.L. 37 Ax.L. 36

Ax.L. 35 Ax.L. 34

Ax.L. 33 Ax.L. 32

Ax.L. 31 Ax.L. 30

Ax.L. 29 Ax.L. 28

Ax.L. 27 Ax.L. 26

Ax.L. 25 Ax.L. 24

Ax.L. 23 Ax.L. 22

Ax.L. 21 Ax.L. 20

Ax.L. 19 Ax.L. 18

Ax.L. 17 Ax.L. 16

Ax.L. 15

Ax.L. 14

Ax.L. 13

Ax.L. 12

Ax.L. 11

Ax.L. 10

Ax.L. 9

Ax.L. 8

Ax.L. 7

Ax.L. 6

Ax.L. 5

Ax.L. 4

Ax.L. 3

Ax.L. 2

Ax.L. 1

Bundle, Shroud & Central Rod Axis I nner Ring Rod Axis Outer Ring Rod Axis

Elevation [m]

-0.1

GRI DU1

GRI DU2

GRI DM1

GRI DM2

GRI DM3

GRI DL1

GRI DL2

GRI DL3

GRI DU3

GRI DB1 GRI DL3

GRI DL2


5

Quench-01Results of Reference Calculation Total Hydrogen

production predicted correctly tuned up by external resistivity

Temperature profiles also predicted well until the beginning of reflooding

Heat Balance was checked, based on methodology from ISP-45


6

Quench-03Input File Changes

Changes done in comparison with Quench-01 calculation (identical approach as in MELCOR analyses of Quench-01 and Quench-03 tests)

Redefinition of (initial and boundary conditions) Power per ring Inlet temperatures and mass flow rates

(Ar, steam and water) Initial temperatures in bundle Time of reflooding beginning Timestep definition


7

Quench-03Hydrogen Production

First step in the input modifications was done in tuning up of external resistivity temperature instability in bottom part of bundle

Tuned up external resistivity 2.1 m/rod

Quench-01 like external resistivity 3.09 m/rod


8

Quench-03Temperature Profile at 2600 s

Quench-01 like external resistivity 3.09 m/rod

Tuned up external resistivity 2.1 m/rod

Water injection onset at 2600 s


9

ICARE2 to MELCOR Comparison

Similar behaviour of both analyses for H2 production when all settings from Q-01 are applied in Q-03 (change of initial and boundary conditions)

Also temperature evolutions had a lot of similarities• Relatively good agreement of temperature prediction in lower and middle part of heated zone

• No temperature escalation results in underestimation of H2 production

[2] J. Duspiva: Quench-03 Test Calculation with MELCOR Code, 8th International QUENCH Workshop, Karslruhe October 29-31, 2002

End of 10th QWS 2004 NRI Contribution Summary


10

Identification of Critical Point

Application of ATLAS Postprocessor

At 2600 s Significant

underprediction of temperatures in hot zone

From analyst point of view identified as overestimation of heat losses through shroud in area of hot zone

From Q-03 test point of view identified as “loss of heat removal”

GRS postprocessor ATLAS

ICARE2 results reprocessing by MELCOR


11

Modelling Approach Used

More possible ways exist Definition of material properties was separated

for• BZF2-5

• BZF1 Conductivity of solid material was modified for

temperature above 550 K Final values were defined iteratively based on

temperature profiles at onset

1.5

1.4

1.3

1.2

1.1

1.0

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0.0

-0.2

-0.3

-0.4

-0.47

CLAD1

PZO1

PZO2

MO2B

I E2B

CLAD2

TUNG2

MO2A

I E2A

CU2A

MO3A

I E3A

CU3A

TUNG3

CLAD3

I E3B

MO3B

PZO3

BZI RR

BZI R

BSS

BZF5

BZF4

BZF3

BZF2

BZF1

Central Unheated Rod

I nner Heated Ring Rod

Outer Heated Ring Rod

Shroud Wall

Ax.L. 41

Ax.L. 40

Ax.L. 39 Ax.L. 38

Ax.L. 37 Ax.L. 36

Ax.L. 35 Ax.L. 34

Ax.L. 33 Ax.L. 32

Ax.L. 31 Ax.L. 30

Ax.L. 29 Ax.L. 28

Ax.L. 27 Ax.L. 26

Ax.L. 25 Ax.L. 24

Ax.L. 23 Ax.L. 22

Ax.L. 21 Ax.L. 20

Ax.L. 19 Ax.L. 18

Ax.L. 17 Ax.L. 16

Ax.L. 15

Ax.L. 14

Ax.L. 13

Ax.L. 12

Ax.L. 11

Ax.L. 10

Ax.L. 9

Ax.L. 8

Ax.L. 7

Ax.L. 6

Ax.L. 5

Ax.L. 4

Ax.L. 3

Ax.L. 2

Ax.L. 1

Bundle, Shroud & Central Rod Axis Inner Ring Rod Axis Outer Ring Rod Axis

Elevation [m]

-0.1

GRI DU1

GRI DU2

GRI DM1

GRI DM2

GRI DM3

GRI DL1

GRI DL2

GRI DL3

GRI DU3

GRI DB1 GRI DL3

GRI DL2

950 K

550 K

610 K Q- 01

Q- 01 and Q- 03

600 K Q- 03

Modification of Zirconia Fibers CDS

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0 500 1000 1500 2000 2500

Temperature [K]

Co

nd

uct

ivit

y [W

/m*K

]

Original

BZF1

BZF2-5


12

Improved PredictionTemperature Profile at ______

Temperature profiles are in agreement with measured values

Significantly lower temperature of Zirconia fiber layers

All three types of rod are degraded, shroud remains intact

Visualization

2500 s 2520 s 2540 s 2560 s 2580 s 2590 s 2600 s 2610 s 2620 s 2630 s 2640 s 2650 s 2660 s 2670 s 2680 s 2690 s 2699 s

Reflooding phase observations

• Overprediction of temperatures in bottom part

• Oscillations of swollen water level

• Quench front level remains at bottom

Continuation of calculation is problematic due to unconvergency and too small timestep (< 10-4 s)

Calculation of whole Q-03 was not finished and is not planned

2710 s


13

Application of New ZF Properties

to MELCOR Analysis“Identical” figure

of temperature profiles with modified ZrO2 Fiber Conductivity

At 2600 s Significant reduction

of ZF temperatures as in ICARE2 run

Slightly underpredicted temperatures of hot zone

H2 production underpedicted 22 g

•Reason trick in shroud modelling to allow its oxidation


14

Bundle Refloodingin MELCOR Analysis

At 2725 s Reflooding phase

observations•Correct prediction

of water level

•Agreement in temperature drop due to quenching

Rod and shroud degradations are not predicted so intensive as in ICARE2 run

Code stability – no termination of run


15

Summary and Conclusions

Application of new postprocessing (using GRS ATLAS tool via MELCOR code) made a possible to identified cause of modelling troubles• Three screens prepared – one of them for both of codes used direct

comparison

• Working term of phenomenon identified: “loss of heat removal through shroud in hot zone”

• Modelling of this feature was done by changing of Zirconia fiber conductivity

Improved modelling resulted in very good agreement in temperature profiles at the time of water injection onset in ICARE2 analysis • In MELCOR analysis - agreement was not found, but temperature profiles

were improved too trick in shroud oxidation modelling with fixed heat transfer coefficient (COR00011 input row)

ICARE2 calculation of reflooding phase results in unconvergency and timestep reduction, oscillation in swollen water level occurred • MELCOR run was more stable during reflooding phase


16

Summary and Conclusions

Application of ICARE2 code allowed to identify cause, which was not possible to identify with integral code due to its modelling simplifications, but conclusions are relevant to both of codes (new features of MELCOR1.8.6 will allow more direct validation on Quench tests)

Regressive application of Zirconia fiber properties to Quench-01 analysis with ICARE2 code resulted in strong overpediction of temperatures and hydrogen production• Phenomenon occurred in Quench-03 test only, not in Quench-01, so

it is not possible to use the same model for analysis of both tests

• Final identification and description of phenomenon, which occurred in Quench-03, should be done by experimenters from FZK, only one of possible analytical approaches to model this test was presented

• Specificity of Quench-03 results from shroud behaviour • Not objective of Quench program

• Unimportant for plant applications no another analysis are needed

Output from effort is available in Report UJV-12204-T

(2)

Jiří Duspiva Nuclear Research Institute Řež, plc. Nuclear Power and Safety Division Dept. of Reactor Technology 11 th International QUENCH Workshop Karlsruhe,

Documents

calculation of quench

quench test calculations

melcor code results

nri e quench

improved model of quench

nri quench activities

background of quench

test analyses