Progression of the Fukushima Accident

Size: px

Start display at page:

Download "Progression of the Fukushima Accident"

Penelope Allison
6 years ago
Views:

1 Workshop on Advances in Understanding the Progression of Severe Accidents in Boiling Water Reactors Progression of the Fukushima Accident ~ Accident Progression at Unit 2 ~ July 17-21, 2017 Shinichi KAWAMURA, Ph.D. Shinya Mizokami, Ph.D. Takeshi HONDA Tokyo Electric Power Company

2 Overview of the 10-Unit Simultaneous Accidents Date (2011) 3/11 3/12 3/13 Fukushima Dai-ichi (1F) Fukushima Daini (2F) OPERATING OUTAGE OPERATING 15:27 1 st Tsunami, 15:35 2 nd Tsunami 15:22~ Tsunamis Station Blackout Water Injection: NO Heat Removal: NO 15:36 Unit 1 Explosion 11:01 Unit 3 Explosion 6:00-6:10 Unit 4 Explosion 3/12 8:13 D/G-6B Water Injection: YES Heat Removal: NO 1:24 RHR 17:00 7:13 RHR 3/12 12:15 Loss of Ultimate Heat Sink 1 15:42 RHR 7:15 3/ /19 5:00 RHR 3/19 22:14 RHR 3/20 3/20 15:46 P/C-2C 3/20 15:46 P/C-2C 3/22 10:36 P/C-4D 3/22 10:35 P/C-4D 3/20 14:30 3/20 19:27 Water Injection: YES Heat Removal: YES Cold Shutdown

Plant Status after Tsunami (1F2 : BWR-4/Mk-I) 2 Stack SRV Generator Off-site power D/W vent valve S/C vent valve SLC Reactor pressure vessel Primary containment vessel MSIV FDW Feedwater pump LPCP

3 Plant Status after Tsunami (1F2 : BWR-4/Mk-I) 2 Stack SRV Generator Off-site power D/W vent valve S/C vent valve SLC Reactor pressure vessel Primary containment vessel MSIV FDW Feedwater pump LPCP Turbine Condenser Sea Operational Circulating Water pump Main equipment inundated Seawater cooling lost Emergency AC power lost DC power lost HPCP Normal AC power lost Temporarily operational Batteries Core spray Heat exchanger RHR CRD RCIC HPCI MUWC CST Sea Power panels DDFP Filtered water RHRS Diesel generators Turbine-driven RCIC cooled the core for about 3 days Sea

RPV 原子炉水位 water level (m) (m) 1F2 RPV Water Level (March 11-15 ) 8 6 RPV water level

3 4 2 0 RCIC was manually started 2 minutes before AC and DC power failures.

Around 2 Mar 14 th Injection Started using Fire Trucks.

4 RPV 原子炉水位 water level (m) (m) 1F2 RPV Water Level (March ) 8 6 RPV water level decreased after RCIC function failure RCIC was manually started 2 minutes before AC and DC power failures. RPV water level was maintained by RCIC for more than 2 days. Around 2 Mar 14 th Injection Started using Fire Trucks. TAF -2 燃料域水位計 Fuel range water (A) level (A) (measured) -4 燃料域水位計 Fuel range water (A)( level 補正 (A) ) (Corrected) 燃料域水位計 Fuel range water (B) level (B) (measured) -6 3/11 3/12 3/12 3/13 3/13

Depressurization by SRVs operation 7 6 5 4 3

operated in order to switch to injection by

5 RPV 原子炉圧力 Pressure (MPa[abs]) 1F2 RPV Pressure Trend (March ) Rapid Depressurization by SRVs operation 原子炉圧力 RPV Pressure (A) (A) SRVs were operated in order to switch to injection by fire trucks after RCIC function failure 0 3/11 3/12 3/12 3/13 3/13

PCV 格納容器圧力 Pressure (MPa[abs]) 1F2 PCV Pressure Trend (March 11-15 ) 5 0.8 0.7 0.

6 PCV 格納容器圧力 Pressure (MPa[abs]) 1F2 PCV Pressure Trend (March ) D/W Dry Well 圧力 Suppression S/C Chamber 圧力 PCV pressure increased by H 2 gas production with Zr-water reaction /11 3/12 3/12 3/13 3/13

7 1F2 Steam Discharging through the Blowout Panel 6 (Shot at 8:58 AM on March 15 th )

8 7 Analysis of Intermittent Reactor Pressure Spikes

Water level in RPV (m) RPV, D/W(x10) pressure (MPaa) Background 8 Observed 3-time intermittent

Core relocation followed by steam generation (c) SRV behavior (It is not clear whether SRV actually

Non-condensable) to reproduce RPV and Dry Well pressure increases by GOTHIC Core melt progression

9 Water level in RPV (m) RPV, D/W(x10) pressure (MPaa) Background 8 Observed 3-time intermittent re-pressurization can be related to (a) Water injection by fire trucks and induced Zr-water reaction (b) Core relocation followed by steam generation (c) SRV behavior (It is not clear whether SRV actually opened or closed) The amounts of steam and hydrogen generation were analyzed (Condensable or Non-condensable) to reproduce RPV and Dry Well pressure increases by GOTHIC Core melt progression scenario was studied RCIC stopped 6:00 TAF BAF SRV opened Water injection started by fire trucks? 6:

10 Postulated Process leading to RPV Pressurization On RPV depressurization, flashing caused the water level to drop steeply, completely exposing the core. Water level was restored by water injections from fire trucks. Steam generation caused Zr-water reaction, generating hydrogen and large amounts of heat, and causing RPV pressure to rise. Water injections from fire trucks were suspended due to the rise in RPV pressure. Steam generation decreased and let RPV pressure drop. Generated steam and H 2 could cause RPV pressure increase even if SRV was kept open

11 Injection 注水量 Rate (m 3 /h) (m 3 /h) Causes of Insufficient Core Cooling by Fire Trucks 10 The amount of water pumped by fire trucks was adequate to cool the cores of units 1 to Water injection using RCIC The injection rate needed to remove the decay heat DHRIR Estimated using the fire truck pump discharge pressure gauge and flow meter Pumping by fire trucks 海水 ( 減圧前 ) 海水 ( 減圧後 ) 必要注水量 Seawater (before pressure loss) Seawater (after pressure loss) Injected amount necessary 0 3/11 3/12 3/12 3/13 3/13 3/16 3/16 The amount of water from fire trucks and MDRIR (1F2 example) Assessed how much amount of water injected into the reactors and how those effects influenced accident progressions

12 Causes of Insufficient Core Cooling by Fire Trucks 11 Confirmed that there were paths that generated bypass flows to the main condenser and condensate storage tank -- Background to the identification of possible bypass flows -- At 1F2, core damaged after the start of water injection by fire trucks Late March 2011 Found out water accumulated in main condenser Suggested possible bypass flows existed Main paths generating bypass flows (example of Unit 1) Reactor building Valve seal Reactor Pressure vessel Condensate transfer pump Flow of water into the reactor Evaporator Condensate storage tank Condensate pump Bypass flow path Turbine building Main condenser Confirmed from piping drawings that there were paths that could generate bypass flows to the main condenser and the condensate storage tank during the accident Fire-extinguishing pump Fire truck pump discharge pressure gauge and flow meter Filtrate tank

13 RPV pressure (MPaa) D/W pressure (MPaa) Correlation between RPV and Dry Well Pressure Depressurization by manual SRV opening. RPV PCV RPV PCV RPV PCV :00 3:00 The 1st re-pressurization could be caused by SRV closure and re-open The 2nd re-pressurization could occur while SRV was kept open

14 Pressure (MPa[abs]) Simulation results Pressure Behavior RPV Press. (Measured) D/W Press. (Measured) RPV Press. (GOTHIC) D/W Press. (GOTHIC) S/C Press. (GOTHIC) :00 3:00 By tuning the steam and hydrogen generation rates, RPV and PCV pressure behaviors were reproduced by GOTHIC code.

Steam generation (kg/s) H2 generation (kg/s) Evaluation of H 2 and Steam Generation 14 10 8 Steam Hydrogen

1 0 21:00 From the relationship between H 2 generation and energy release by Zrwater reaction, it is

15 Steam generation (kg/s) H2 generation (kg/s) Evaluation of H 2 and Steam Generation Steam Hydrogen 2nd 3rd st re-press :00 From the relationship between H 2 generation and energy release by Zrwater reaction, it is estimated that majority of the fuel melted during the second spike. Calculation result shows that Zr in the core was almost consumed by the end of the second spike. 3:00 0.0

16 15 Analysis of SRV Behavior and Cause of Inoperability

1F2 Condition of SRVs after the Tsunami 16 格納容器内

for excitation of solenoid valves failed due to the

HPIN system to supply N 2 gas from outside the PCV

17 1F2 Condition of SRVs after the Tsunami 16 格納容器内格納容器外 Operations to open SRVs were tried consecutively to depressurize the RPV. ADS accumulators RV accumulators N2 Bottles DC power for excitation of solenoid valves failed due to the Tsunami, so used car batteries. HPIN system to supply N 2 gas from outside the PCV also failed due to the Tsunami, so N 2 accumulators were used. There were two kind of tanks; RV accumulators and ADS accumulators. N2 gas SOLV Steam Flow Not Available Due to Tsunami DC Power Portable Batteries

18 1F2 SRV Basic Information 17 No. SV function (Set Point) SRV Set Points at 1F2 RV function (Set Point) ADS function A 7.81 MPa[abs] 7.61 MPa[abs] Yes B 7.88 MPa[abs] 7.68 MPa[abs] Yes C 7.88 MPa[abs] 7.68 MPa[abs] Yes D 7.81 MPa[abs] 7.61 MPa[abs] No E 7.74 MPa[abs] 7.61 MPa[abs] Yes F 7.74 MPa[abs] 7.54 MPa[abs] No G 7.81 MPa[abs] 7.68 MPa[abs] Yes H 7.88 MPa[abs] 7.68 MPa[abs] Yes Set Points of RV for F and SVs for E, F are relatively low. SRVs other than D, F have ADS function.

19 原子炉圧力 (MPa[abs]) RPV Pressure (MPa[abs]) 1F2 RPV Pressure Trend and SRV Operations (1) Around 16:34 A, B, C, G(RV, ADS) (2) Around 18:02 E, D, F(RV) 原子炉圧力 RPV Pressure ( 実測値 (Measured) ) :00 14:00 15:00 16:00 日時 17:00 19: :00

whole circuits (2)18:02 Mar 14 th Power supplied only to solenoid valves Failed Successful Failures of RPV

20 1F2 SRV Operations around 16:34 and 18:02 19 Switch Solenoid valve Portable Batteries 12V 10 Solenoid valve Portable Batteries 12V 10 SRV Panel Lamp Red/Green Lamp Red/Green (1)16:34 Mar 14 th Power supplied to the whole circuits (2)18:02 Mar 14 th Power supplied only to solenoid valves Failed Successful Failures of RPV depressurization at 16:34 was highly likely to be caused by the shortage of the power supply voltage to the solenoid valves.

21 原子炉圧力 (MPa[abs]) RPV Pressure (MPa[abs]) 1F2 RPV Pressure Trend and SRVs Operations 原子炉圧力 RPV Pressure ( 実測値 (Measured) ) (3)Around 21:00 and 21:20 A, B (RV) (5) After A, B, E, G, H(ADS) (6) Around 1:00 C(ADS) : :00 22:00 日時 23:00 (4) After C, E, G (RV) 1:00 2:00 3:00

原子炉圧力 RPV Pressure (MPa[abs]) D/W D/W Pressure 圧力 (MPa[abs]) 1F2 SRV A, B Operations around 21:20 March 14 21 4.0 3.5 3.0 2.5 2.0 1.5 1.

22 原子炉圧力 RPV Pressure (MPa[abs]) D/W D/W Pressure 圧力 (MPa[abs]) 1F2 SRV A, B Operations around 21:20 March Around 2 Start Injection (3) RPV D/W Discharge Pressure of Fire Trucks: 1MPa : :00 22:00 23:00 原子炉圧力 RPV Pressure 1:00 D/W 圧力 Pressure :00 3:00 Close Open Dry Well pressure rose with RPV pressure decrease around 21:20. It is likely to be caused by SRV opening and by hydrogen gas transfer from the RPV to the PCV, which was generated during core damage.

原子炉圧力 RPV Pressure (MPa[abs]) D/W D/W Pressure 圧力 (MPa[abs]) 1F2 RPV Pressure 2 nd Spike after Depressurization 22 4.0 3.5 3.0 RPV D/W 1 0.9 0.8 2.5 2.0 1.5 1.

23 原子炉圧力 RPV Pressure (MPa[abs]) D/W D/W Pressure 圧力 (MPa[abs]) 1F2 RPV Pressure 2 nd Spike after Depressurization RPV D/W Around 2 Start Injection Discharge Pressure of Fire Trucks: 1MPa : :00 22:00 23:00 原子炉圧力 RPV Pressure 1:00 D/W 圧力 Pressure :00 3:00 Open Open When Pressure When Pressure Increase Decrease Both RPV pressure and Dry Well pressure increased. It is possible that SRVs were being kept open during the 2 nd spike since 21:20. Pressure increase was likely caused by vast amount of H 2 gas generation in the RPV. (Intense Zr-water reaction could have happened during the 2 nd spike.)

24 原子炉圧力 (MPa[abs]) RPV Pressure (MPa[abs]) 1F2 SRV Operations after March 15 (Point 4) 原子炉圧力 RPV Pressure ( 実測値 (Measured) ) (5) After A, B, E, G, H(ADS) (6) Around 1:00 C(ADS) (3)Around 21:00 and 21:20 A, B (RV) : :00 22:00 日時 23:00 (4) After C, E, G (RV) 1:00 2:00 SRV (C), (G) operations were halted because sparks were ignited when batteries were connected to the solenoid valves. 3:00

PN+PR, PP+PA 原子炉圧力 RPV Pressure (MPa[abs]) D/W D/W Pressure 圧力 (MPa[abs]) 1F2 SRV Operations after March 15 (Point 4) 24 4.0 1 3.5 0.9 3.0 0.8 2.5 2.0 1.5 1.

25 PN+PR, PP+PA 原子炉圧力 RPV Pressure (MPa[abs]) D/W D/W Pressure 圧力 (MPa[abs]) 1F2 SRV Operations after March 15 (Point 4) Around 2 Start Injection (4) Discharge Pressure of Fire Trucks: 1MPa : :00 22:00 23:00 原子炉圧力 RPV Pressure 1:00 D/W 圧力 Pressure :00 3:00 Close SRV (E) (RV) could not be opened because of the increased PCV pressure :00 Orange:P N (ADS function) +P R Red :P N (RV function) +P R Blue :P P +P A Moment to close Moment to close was bigger 2 21:00 22:00 23:00 (4) 1:00 Moment to open 2:00 3:00

26 原子炉圧力 (MPa[abs]) RPV Pressure (MPa[abs]) 1F2 SRV Operations after March 15 (Points 5, 6) 原子炉圧力 RPV Pressure ( 実測値 (Measured) ) (5) After A, B, E, G, H(ADS) (6) Around 1:00 C(ADS) (3)Around 21:00 and 21:20 A, B (RV) : :00 22:00 日時 23:00 (4) After C, E, G (RV) 1:00 2:00 3:00 Point 5: SRVs were tried to open consecutively but RPV pressure decrease was not confirmed. Point 6: RPV pressure decrease was confirmed when SRV (C) was operated.

原子炉圧力 RPV Pressure (MPa[abs]) D/W D/W Pressure 圧力 (MPa[abs]) 1F2 SRV Operations after March 15 (Points 5, 6) 26 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 19:00? Close? Open?

27 原子炉圧力 RPV Pressure (MPa[abs]) D/W D/W Pressure 圧力 (MPa[abs]) 1F2 SRV Operations after March 15 (Points 5, 6) :00? Close? Open? Around 2 Start Injection 2 21:00 22:00 23:00 (5) (6) 原子炉圧力 RPV Pressure 1:00 Discharge Pressure of Fire Trucks: 1MPa D/W 圧力 Pressure :00 3:00 SRV operations at (5),(6) using ADS accumulators Those operations should have been successful considering the pressure of ADS accumulators and Dry Well pressure. However, SRV open/close is still not clear because of complicated situations during core melt progression. Steam and H 2 generation could cause RPV pressurization. Possibility of leakage from the RPV to the Dry Well. Possibility of leakage from the Dry Well to the Reactor Bldg.

28 27 Interpretation of CAMS Data to understand Accident Progression

29 CAMS (Sv/h) Pressure (MPa[abs]) Background 28 Core melt progression scenario was studied using CAMS data trend, which could provide some insights to know the FP transport in the PCV 1E+3 A B C D 9 1E E+1 RPV Press. 1E E-1 1E-2 CAMS(D/W) D/W Press. CAMS(S/C) S/C Press E-3 6:00 0 3/16

30 FP Gas Transport 3 1 RPV to S/C via SRV S/C to D/W via Vacuum Breaker 29 FDW MS 2 RPV leakage to D/W D/W to S/C via vent tube down comer 3 D/W leakage to Reactor Bldg. Vacuum Breaker

31 CAMS (Sv/h) Pressure (MPa[abs]) Accident Progression inferred from CAMS data 30 A: Core damage started, dose rates in D/W and S/C started to increase. FP flowed through the SRV from RPV into S/C, then into D/W through VB. 1E+3 1E+2 1E+1 1E+0 A E-1 1E-2 CAMS(D/W) D/W Press. CAMS(S/C) S/C RPV Press E-3 6:00 0 3/16

32 CAMS (Sv/h) Pressure (MPa[abs]) Accident Progression inferred from CAMS data 31 A : Steep dose rate increase might be related to 2nd RPV re-pressurization. (i.e. large Zr-water reaction with SRV kept open) 1E+3 1E+2 1E+1 1E+0 A E-1 1E-2 CAMS(D/W) D/W Press. CAMS(S/C) S/C RPV Press E-3 6:00 0 3/16

33 CAMS (Sv/h) Pressure (MPa[abs]) Accident Progression inferred from CAMS data 32 B: S/C dose rate decreased while D/W dose rate increased, which indicates the possibility that FP was leaking directly from the RPV to the D/W. 1E+3 1E+2 1E+1 1E+0 B E-1 1E-2 CAMS(D/W) D/W Press. CAMS(S/C) S/C RPV Press E-3 6:00 0 3/16

34 CAMS (Sv/h) Pressure (MPa[abs]) Accident Progression inferred from CAMS data 33 D: No large dose rate change was observed in D/W and S/C, although measurement was interrupted in this period. 1E+3 1E+2 C E+1 6 1E E-1 1E-2 CAMS(D/W) D/W Press. CAMS(S/C) S/C RPV Press E-3 6:00 0 3/16

35 CAMS (Sv/h) Pressure (MPa[abs]) Accident Progression inferred from CAMS data 34 E: D/W dose rates largely increased, which might be attributed to RPV rupture. 1E+3 1E+2 D Max 138 Sv/h E+1 6 1E E-1 1E-2 CAMS(D/W) D/W Press. CAMS(S/C) S/C RPV Press E-3 6:00 0 3/16

36 35 Thank you for your attention.

Seismic Damage Information (the 120th Release) (As of 15:00 April 29, 2011)

Extract April 29, 2011 Nuclear and Industrial Safety Agency Seismic Damage Information (the 120th Release) (As of 15:00 April 29, 2011) Nuclear and Industrial Safety Agency (NISA) confirmed the current