Overview of
Fukushima-Accident Analysis
ERMSAR 2012, Cologne (Germany)
March 21 – 23, 2012
JNES
Masanori FUKASAWA
0
Contents
1. 1F※1 Accident Analyses (Plant behavior) at JNES
2. Plant Behavior Analysis using MELCOR
A) Results at IAEA Ministerial Conference (June 2012) and Problems
B) Revised Analysis
3. Primary System Behavior during IC operation
4. Hydrogen Mixing and Explosion in Reactor Building
(R/B)
5. Conclusions
※1F: Fukushima Daiichi
1
1. 1F Accident Analyses (Plant behavior) at JNES
Published Analyses and Evaluations
 ”Report of Japanese Government to the IAEA Ministerial Conference on
Nuclear Safety,” June 2011. (June Report), JNES-RE-2011-0002.
 Documents of Hearings at Nuclear Safety Commission (NSC) and Nuclear
and Industrial Safety Agency (NISA).
 Analyses and Evaluations submitted to NISA (published on JNES web).
Accident Analysis
Plant Behavior
• Getting chronology
• MELCOR analysis
information together
• Primary system
• Event tree analysis of
behavior during IC
the accident
operation
• Possibility of recriticality • Hydrogen mixing and
• Reactivity constraint by
explosion
sea water
• MCCI in case water
• Time before fuel damage injection stops
in SFP
• Possibility of PCV failure
• Salt precipitation
by H2 deflagration
• H2, O2 concentration
FP Release
• FP release and dose
evaluation
• Influence in case water
injection stops
• FP release and EPZ
• Estimation of FP release
and dose based on
monitoring data
• FP release in case of
venting
2
2. Plant Behavior Analysis using MELCOR
Objectives: To figure out plant behaviors of 1F1 – 3 and
enhance safety measures.
6
Code: MELCOR1.8.5
Analytical configuration:
Primary boundary
Secondary boundary
Environment
Blowout panel
D/W
PCV
FHB
ADS
Core
3F
Bypass
SGTS
W/W vent
Upper plenum
4F
S/RV
Downcomer
Leakage
Steam
dome
蒸気ドーム
Vent pipe
R/B
Lower plenum
下部プレナム
2F
RPV
RPV failure
Pedestal
Vacuum
breaker
1F
W/W
volumes of primary system
 4 volumes of containment
 5 volumes of reactor building
to simulate FP transfer.
 Junctions of S/R valves,
vacuum breaker, PCV leak,
W/W vent
 Further (not depicted),
activated cooling systems
and assumed leak to simulate
transports of steam, coolant
and FP.
1F1：IC
1F2：RCIC
1F3：RCIC, HPCI
3
2.A) Result of June Report and Problems (1F1)
1F1: Lower coolant injection case
Analytical results
Water injection (fire pump, F/P) by 3/15 is 88 m3
実時刻
Date
①
②
③
⑥
⑤
④
7.0
炉圧 (MPa)
(MPa)
P/S pressure
3/13
3/14
6000
1F1
(○,△)実測値
○,△：Measured
data
4000
炉圧
P/S
Pressure
6.0
2000
5.0
W/W vent
4.0
TAF
3.0
2.0
-2000
RPV水位
-4000
0
12
IC actuation is limited and water
level decreases at an early stage.
Core melts before alternate
water injection.
RPV failure is calculated at 5 hrs.
Most core is calculated to melt
and slump to PCV floor.
Water level
1.0
0.0
0
level (mm)
Water
水位 (mm)
8.0
3/12
24
36
48
60
経過時間
Time
(hr) (h)
72
84
96
P/S pressure and water level
①IC stop、②water injection、③W/W vent open、
④W/W vent close、⑤Sea water injection、
⑥Increase of PCV leakage
Problems
RPV failure timing (MELCOR
default model calculates early
failure.)
Actuation of W/W vent at 1st
attempt (3/12 10:17).
4
2.A) Result of June Report and Problems (1F2)
Assumption of analysis in June Report
1F2: In case PCV confinement maintains
Date
1.0
D/W pressure increases due
to temperature rise of S/P.
①RCIC manual actuation
②SBO
③ Change of RCIC water
source from CST to S/P
④RCIC stop
⑤Sea water injection
⑥S/R valve1 open
⑦S/R valve open
⑧explosion
3/13
0:00
3/14
0:00
3/15
0:00
④ ⑥⑦
⑤
⑧
2Pd
pressure
D/W圧力
（MPa）(MPa)
Water source of RCIC is
switched from CST to S/P.
①
②
3/12
0:00
③
実時刻
0.8
D/W圧力実測値(○)
0.6
D/W圧力
(破損無し)
1Pd
0.4
○：Measured data
0.2
0.0
0
12
24
36
48
60
経過時間
Time (hr)(h)
72
84
96
D/W pressure
D/W pressure could not be reproduced in case PCV confinement maintained.
→PCV leakage was assumed.
5
2.A) Result of June Report and Problems (1F2)
1F2: Lower coolant injection case with PCV leakage
Water injection (F/P) by 3/15 is 213 m3
1.0
pressure
D/W圧力
（MPa） (MPa)
①
②
3/12
0:00
③
3/13
0:00
Date
実時刻
3/14
0:00
④ ⑥
⑦
⑤
D/W圧力実測値(○)
Analytical results
3/15
0:00
⑧
2Pd
D/W pressure
0.8
○：Measured data
0.6
1Pd
0.4
RPV failure is calculated
at 80 hrs because water
injection by F/P is not
enough.
D/W圧力
D/W Pressure
0.2
0.0
0
12
24
D/W pressure is well
simulated on assumption
of PCV leakage (50 cm2)
at an early stage.
36
48
60
経過時間
(h)
Time (hr)
72
84
96
①RCIC manual actuation, ②SBO,
③ Change of RCIC water source from CST to S/P,
④RCIC stop, ⑤Sea water injection, ⑥S/R valve1 open,
⑦S/R valve open, ⑧explosion
Higher FP release due to
assumed early PCV
leakage
6
2.A) Result of June Report and Problems (1F2)
1F2: Lower coolant injection case with PCV leakage
Water injection (F/P) by 3/15 is 213 m3
1.0
pressure
D/W圧力
（MPa） (MPa)
①
②
3/12
0:00
③
3/13
0:00
Date
実時刻
3/14
0:00
④ ⑥
⑦
⑤
D/W圧力実測値(○)
Problems
3/15
0:00
⑧
2Pd
D/W pressure
0.8
○：Measured data
0.6
1Pd
Measured D/W pressure does
not increase even S/R valve
opened.
On the other hand, pressure
increases in calculation.
0.4
D/W圧力
D/W Pressure
D/W pressure at this stage is
not simulated due to assumed
PCV leakage.
0.2
0.0
0
12
24
36
48
60
経過時間
(h)
Time (hr)
72
84
96
①RCIC manual actuation, ②SBO,
③ Change of RCIC water source from CST to S/P,
④RCIC stop, ⑤Sea water injection, ⑥S/R valve1 open,
⑦S/R valve open, ⑧explosion
Some heat removal (instead
of PCV leakage) possibly
occurred by then.
7
2.A) Result of June Report and Problems (1F3)
1F3: Lower coolant injection case
Water injection (F/P) by 3/17 is 369 m3
②
①
炉圧 (MPa)
(MPa)
P/S pressure
7.0
③
④
3/15
⑩ ⑪⑫⑬
P/S
炉圧 Pressure
6.0
3/17
3/18
6000
⑭
S/R Valve open
4000
○,△：Measured data
RPV水位
5.0
3/16
4.0
TAF
2000
0
3.0
-2000
2.0
1F3
1.0
0.0
水位 (mm)
level (mm)
Water
3/13 ⑤～⑨ 3/14
3/12
8.0
Analytical results
実時刻
Date
L-2とL-8の間で推移
Water level
-4000
(○、△)実測値
0
24
48
96
72
(h)
経過時間
Time (hr)
120
144
P/S pressure and water level
①RCIC actuation、②RCIC stop, ③HPCI actuation, ④HPCI stop,
⑤S/R valve open, W/W vent open, water injection,
⑥W/W vent close, ⑦W/W vent open, ⑧Sea water injection,
⑨W/W vent close, water injection、⑩～⑭W/W vent open⇔close
 Much H2 is produced
due to water level
decrease when S/R
valve opens.
 RPV failure is calculated
at 79 hrs because sea
water injection is not
enough.
 FP is released to
environment through
W/W vent.
8
2.A) Result of June Report and Problems (1F3)
Date
実時刻
3/12
12:00
3/12
0:00
1.0
②
①
D/W圧力実測値(○)
3/13
0:00
④
③
3/13
12:00
⑤ ⑥～⑨
2Pd
0.8
(MPa)
D/W pressure
圧力 （MPa）
○：Measured data
0.6
D/W Pressure
D/W圧力
1Pd
0.4
0.2
0.0
0
12
24
36
Problems
48
 During RCIC operation, D/W
pressure is underestimated.
(inverse trend to 1F2)
 Measured pressure drops when
HPCI actuates.
 During HPCI operation, water
level is not clear.
 Amount of produced H2
Explosions of 1F3 and 1F4
are possibly attributed to H2
produced in 1F3
経過時間
(h)
Time (hr)
D/W pressure
9
2.B) Revised Analysis (1F3: Analytical Conditions)
 Major
problems; D/W pressure
underestimation during RCIC operation
and pressure drops after HCPI actuation.
information and examination
 RCIC continuous operation using
return line to CST.
→Assume S/P thermal stratification by
RCIC exhaust steam. (see Figs.)
RCIC steam
exhaust pipe
 Latest
Hot water flow
• Simulated
by upper/lower S/P nodes and
RCIC exhaust to the upper.
• HPCI initiated, steam exhausted to lower
node assuming lower temp. of the water
near HPCI exhaust pipe.
 PCV
spray during HPCI
 Similar pressure transition between
P/S and S/C after 42.4 hrs.
→Assume RPV failure at this time.
Hot water
10
2.B) Revised Analysis (1F3: Analytical Result)
3/12
0:00
3/13
0:00
3/14
0:00
1.0
(MPa)
P/S pressure
原子炉圧力
(MPa)
0.8
6
0.6
S/C Spray
4
D/W
pressure
D/W圧力
0.4
→
2
0.2
0.0
0
0
10
20
30
40
50
経過時間 (h)
Time
(hr)
P/S and D/W pressures
60
(MPa)
(MPa)
Pressure
D/WPCV圧力
○：Measured data
原子炉圧力
P/S
Pressure
8
Improved matching with
measured data
D/W pressure increases
during RCIC operation
Depressurization due to S/C
spray （No depressurization if
thermal stratification not
considered because of lower
spray flow rate)
 Remained problems
Modeling of S/P thermal
stratification; investigation
using CFD
Further investigation is
needed for PCV leakage and
W/W vent, which have large
influence on FP release.

Date
実時刻
11
2.B) Revised Analysis (1F2: Analytical Conditions)


Major problems; D/W pressure after S/R valve open. (Some heat
removal instead of PCV leakage)
Latest information and examination
 Tsunami water flooded at a depth of boots length in RCIC room
(same level as S/P torus room) at 1:00, 3/12 and increased at
2:00.*
→Assume S/P heat removal by flooding water
• 60% heat of RCIC exhaust steam is removed.
 Early PCV leakage is not assumed. Instead;
• Small leakage at 70 hrs because measured D/W pressure
slightly decreases.
• Enlargement of leakage at 90 hrs when large pressure drop
is measured.
 RCIC injection rate is adjusted to simulate time when water
level comes down to TAF.
 Assume S/P thermal stratification (Same as 1F3).
*TEPCO, “Report regarding factual information related to the investigation results of the accident situation
at Fukushima Daiichi Nuclear Power Plant,” Dec. 22, 2011.
12
2.B) Revised Analysis (1F2: Analytical Result)
9.0
3/12
A,B 0:00
3/13
0:00
3/14
0:00
C
8.0
ダウンカマ水位
Water level
E
[2]
D
[1]
F
[3] 3/15
0:00
G
[5]
6000
H
[10]
(MPa)
P/S pressure
炉圧 (MPa)
7.0
6.0
2000
5.0
炉圧
P/S Pressure
TAF
4.0
0
3.0
-2000
2.0
1.0
BAF
(○,△)実測値
○△：Measured
0
20
 P/S
4000
コラプスト水位
(mm)
level (mm)
Water
Date
実時刻
pressure is also
simulated by adjusting
RCIC injection rate
-4000
data
40
60
経過時間
Time
(hr) (h)
80
100
P/S pressure and water level
A:RCIC manual actuation, B:SBO, C:Change of RCIC water source from CST to S/P,
D[1]:RCIC stop, F[3]Sea water injection, E[2]:S/R valve1 open,
G[5]:S/R valve open, H[10]explosion
13
2.B) Revised Analysis (1F2: Analytical Result)
Enlargement of leakage (32 cm2)
Date
実時刻
3/12
0:00
1.0
3/13
0:00
3/14
0:00
3/15
0:00
 High
1.0
(○,△)実測値
○△：Measured
data
炉圧
P/S Pressure
0.8
0.6
0.6
0.4
0.4
D/W圧力
D/W pressure
0.2
0.0
0.2
0
20
40
60
Time (hr)
経過時間
(h)
80
D/W圧力
(MPa)
level (mm)
Water
炉圧 (MPa)
(MPa)
D/W pressure
0.8
0.0
100
0.6 cm2 of leakage
D/W pressure is
reproduced.
 Calculated pressure
increase becomes lower
and consistent with
measured data.
Steam through S/R
valve flows to lower
level of S/P, whose
temperature is lower
due to thermal
stratification.
P/S and D/W pressures
14
3. Primary System Behavior during IC operation

Isolation Condenser (IC) is a unique system for reactor cooling in unit-1, and
worked at the initial stage of the accident.
(ceased due to AC/DC valve power loss by the tsunami)

RELAP5/mod3 analyses were performed to investigate the IC behavior.
 IC functioned properly to the original design. (Not impaired by the earthquake)
 Sensitivity analysis shows that the core uncovery could have been avoided by
continued operation of IC after the tsunami.
Earthquake
SR/V
8.0 SCRAM
8000
7.0
6000
4000
IC start
IC manual
operation
4.0
IC stop
3.0
2.0
Above TAF (mm）
Rx Press(MPa,g）
6.0
5.0
Earthquake
SCRAM
Reactor pressure
1.0
2000
Assuming IC continued
operation after tsunami
Feed seconary
coolant
0
-2000
RELAP5
Recorder
Chart
Water level above TAF
-4000
-6000
0.0
3/11 14:45
3/11 15:00
3/11 15:15
3/11 15:30
3/11 15:45
3/11 16:00
DAY
IC system (Unit-1)
Initial stage of the accident
(after earthquake)
3/11 16:15
3/11
14:30
3/11
15:00
3/11
15:30
3/11
16:00
3/11
16:30
3/11
17:00
DAY
3/11
17:30
3/11
18:00
3/11
18:30
3/11
19:00
3/11
19:30
Assuming IC
continued operation
15
4. Hydrogen Mixing and Explosion in R/B

JNES is conducting analyses of hydrogen gas mixing and detonation in Reactor
Buildings (R/Bs) for investing explosion phenomena during Fukushima accident.
 MELCOR for hydrogen source evaluation , FLUENT(CFD code) for
hydrogen gas transport and mixing, and AUTDYN for structural analysis of
detonation
 The objectives are to better understand the phenomena that took place in
Unit 1 and Unit 3, and assess and improve the methods and tools.
CFD model of Reactor Building
Mixture gas velocity
Hydrogen gas concentration
16
4. Hydrogen Mixing and Explosion in R/B

Some results from the analysis
of detonation at unit 3
Debris Velocity
R/B Pressure
 The amount of hydrogen
gas leaked into R/B is
estimated to be
approximately 1 ton.
 If it is assumed that the
leakage took place at S/C
or D/W’s penetration,
overall detonation
behavior is well
reproduced.
 With initial velocity 70 m/s
debris is supposed to
reach at about 250 from
the top of R/B at 7.1
seconds.
This photo is quoted from Fukushima-chuo TV
This photo is quoted from TEPCO web
17
5. Conclusions
• JNES has been conducting various analyses of the
Fukushima accident.
• Plant behavior analysis using MELCOR improved by
assuming S/P thermal stratification and latest information
for 1F2 and 1F3.
• P/S behavior analysis of 1F1 using RELAP5/mod3 shows
IC functioned properly to the original design. (Not impaired
by the earthquake)
• Detonation analysis with the assumption that leakage took
place at S/C or D/W’s penetration well estimates overall
R/B behavior of 1F3.
18