European Commission DG for Energy (ENER/D2)

European Commission
DG for Energy
(ENER/D2)
How to Improve Safety in
Regulated Industries
What Could We Learn From
Each Other
Background material
Annex A
ENCO FR-(12)-44
July 2012
Specific Contract No. ENER/ 2011/NUCL/SI2.599383
in
How
w to Im
mprove Safety
S
ndustrie
es
Regullated In
What Could We Lea
arn From
Eac
ch Othe
er
Backgground Material
Fina
al Repo
ort
Annex A
EN
NCO FR
R-(12)-4
44
July
J
201
12
Und
der the Fram
mework Serrvice Contract
forr Technical Assistance TREN/R1/3
350-2008 Lo
ot 3
Specific Contract No. ENER
R/ 2011/NUC
CL/SI2.5993
383
Prepared by:
b
Prep
pared fo
or:
E
European Commissiion
DG
G for Enerrgy
(ENE
ER/D2 Nuclear Energ
gy)
DISCLAIM
MER
The con
ntent of thiis report is the sole ressponsibility
y of the Con
ntractor andd can in no way be tak
ken
e views of the Europea
an Union.
to reflect the
Annex A. Overview Fukushima Dai-ichi Accident
TABLE OF CONTENTS
1.
INTRODUCTION ..................................................................................... 1
2.
BACKGROUND – PLANT CHARACTERISTICS .................................................... 2
2.1.
2.2.
3.
SEISMIC AND TSUNAMI DESIGN BASIS ........................................................... 6
3.1.
3.2.
4.
EARTHQUAKE ................................................................................................ 7
TSUNAMI .................................................................................................... 8
PLANT CAPABILITIES AND RESPONSE .......................................................... 9
5.1.
5.2.
5.3.
5.4.
5.5.
5.6.
5.7.
5.8.
6.
SEISMIC...................................................................................................... 6
TSUNAMI .................................................................................................... 6
MARCH 11 EARTHQUAKE AND TSUNAMI ....................................................... 7
4.1.
4.2.
5.
GENERAL ARRANGEMENTS OF FUKUSHIMA DAI-CHI PLANT ..................................................... 2
DESIGN CHARACTERISTICS OF THE UNITS ..................................................................... 2
PLANT STATUS BEFORE THE EVENT ........................................................................... 9
LOSS OF POWER ............................................................................................ 11
CORE COOLING ............................................................................................. 13
HYDROGEN EXPLOSIONS..................................................................................... 14
CONTAINMENT PRESSURE CONTROL ......................................................................... 15
SPENT FUEL POOLS AND DRY CASK STORAGE ............................................................... 16
ALTERNATIVE INJECTION SOURCES .......................................................................... 16
RADIOLOGICAL CONSEQUENCES ............................................................................. 17
CAUSAL ANALYSIS................................................................................. 17
6.1.
EXISTING STUDIES .......................................................................................... 18
EPRI analysis .................................................................................................. 18
NRC recommendations ...................................................................................... 19
6.2.
CAUSE MAPPING ............................................................................................ 21
Step 1 - Definition of the problem ........................................................................ 21
Step 2 – Analysis of causes (Causal Map) ................................................................. 24
Step 3. Analysis of solutions ............................................................................... 30
6.3.
SUMMARY CONCLUSIONS .................................................................................... 31
7.
REFERENCES FOR ANEX .......................................................................... 33
1. Introduction
On March 11, 2011 at 14:46 Japan standard time, the Fukushima Dai-ichi nuclear power
plant experienced a seismic event and subsequent tsunami [A-1]. The accident and the
ensuing mitigation and recovery activities occurred over several days, involved a number
of incidents, and might provide several opportunities for lessons learned.
The initiating seismic event involved multiple ruptures of seismic sources over an area of
about 400 km x 200 km. The earthquake was very significant (magnitude ~9 on the Richter
Scale), considered the fourth largest in recorded world history. Although the earthquake
did not cause significant structural or operational damage to Fukushima Dai-ichi NPP, the
event did cause major infrastructure damage to areas around the plant. The offsite
damages led to loss of offsite power.
The earthquake caused a series of tsunamis, the largest of which arrived at Fukushima Daiichi approximately 41 minutes after the earthquake, reaching a wave height of
approximately 15 m. The associated volume of water – and the related hydrodynamic
forces – caused extensive and deep flooding in and around all major structures of operating
Units 1 - 3.
The design basis seismic definitions were – in magnitude and frequency content – not
significantly different than the actual seismic event. However, the nature of the seismic
event (that is, occurring across a large area and involving multiple ruptures of seismic fault
segments) was not incorporated into the design basis tsunami definition.
The earthquake and tsunami produced widespread devastation across northeastern Japan,
resulting in approximately 25,000 people dead or missing, displacing many tens of
thousands of people, and significantly impacting the infrastructure and industry in the
northeastern coastal areas of Japan.
The combination of the massive earthquake and devastating tsunami at Fukushima were
well in excess of external events considered in the plant design. The Fukushima accident
also challenged the plant’s mitigation capabilities and emergency preparedness.
Evaluation of the Fukushima accident presented in this Annex addresses the essential
elements of the regulatory framework that play a role in providing protection from designbasis events, as well as events as severe and complex as the Fukushima accident. Those
elements include protection against seismic and flooding events (considered as designbasis events), protection for loss of all AC power (considered as a beyond-design-basis
event), and mitigation of severe accidents (addressing beyond-design-basis topics of core
damage and subsequent containment performance), as well as emergency preparedness.
The Fukushima accident highlights the full spectrum of considerations necessary for a
comprehensive and coherent regulatory framework.
It worth noting that similar issues were raised by the TMI accident and that many beyonddesign-basis requirements, programs, and practices were derived from that experience.
This Annex presents evaluations that address specific elements of protection, mitigation,
and preparedness and evaluate their current capabilities, limitations, and potential
enhancements.
How to Improve Safety in Regulated Industries
What Could We Learn From Each Other
Background Material, ENCO FR-(12)-44
© ENCO
Page
1
2. Background – Plant characteristics
2.1. General arrangements of Fukushima Dai-chi plant
Fukushima Dai-ichi site is located along the northeast coast of Japan, bordering the
western edge of the Pacific Ocean. Initial siting occurred between 1967 and 1973, with
each of the six units coming on-line between 1971 and 1979. Operational startup dates,
power output, and general design information are shown in Table 1.
Table 2-1. Comparison of Units 1 to 6
*)
Unit
Startup
MWe Output
Reactor Type, Containment
High Pressure Cooling*
1
1971
460
BWR-3, Mark I
IC, HPCI
2
1974
784
BWR-4, Mark I
RCIC, HPCI
3
1976
784
BWR-4, Mark I
RCIC, HPCI
4
1978
784
BWR-4, Mark I
RCIC, HPCI
5
1978
784
BWR-4, Mark I
RCIC, HPCI
6
1979
1100
BWR-5, Mark II
RCIC, HPCS
IC - Isolation Condenser; HPCI - High Pressure Coolant Injection, RCIC - Reactor Core Isolation
Cooling , HPCS - High Pressure Core Spray
Fukushima Dai-ichi Units 1 through 4 are located in the southern part of the station; Unit 1
is the northernmost and Unit 4 is the southernmost. Fukushima Dai-ichi Units 5 and 6 are
located farther north and at a somewhat higher elevation than the Unit 1–4 cluster, and
Unit 6 is located to the north of Unit 5.
The grouped units share some common facilities and structures, such as control rooms, and
vent stacks. This commonality applies to Units 1 and 2, Units 3 and 4, and Units 5 and 6. In
addition to individual unit spent fuel pools, the plant also has a shared spent fuel pool and
a shared dry cask storage facility. The shared pool and the dry cask storage are for all six
units. The shared spent fuel pool is located on the inland side (west) of Unit 4. The dry
cask storage facility is located between Units 1 and 5 along the coast. The general
arrangement of the units prior to the earthquake and the tsunami is shown in Figures 2-1
and 2-2.
2.2. Design characteristics of the units
The main design features of Units 1-6 are presented in Table 2. List of core cooling systems
that can be used in emergency conditions is provided in Table 3. The latter information is
limited to Units 1 to 3 - units that experienced the most severe problems during the event.
Configuration of the primary and secondary containment systems and the reactor vessel is
shown on Fig. 2-3.
How to Improve Safety in Regulated Industries
What Could We Learn From Each Other
Background Material, ENCO FR-(12)-44
© ENCO
Page
2
Table 2-2. Unit-specific design characteristics
Design Parameter
General
Unit 1
Unit 2
Unit 3
Unit 4
Unit 5
Unit 6
460
784
784
784
784
1100
Mar-71
Jul-74
Mar-76
Oct-78
Apr-78
Oct-79
Reactor type
BWR-3
BWR-4
BWR-4
BWR-4
BWR-4
BWR-5
Containment type
Mark I
Mark I
Mark I
Mark I
Mark I
Mark II
Main Contractor
GE
GE/Toshi
ba
Toshiba
Hitachi
Toshiba
GE/Toshi
ba
Heat output, MW
1380
2381
2381
2381
2381
3293
No of fuel assemblies
(FA)
400
548
548
548
548
764
Full length of FA, m
4.35
4.47
4.47
4.47
4.47
4.47
Number of control rods
97
137
137
137
137
185
RPV inner diameter, m
4.8
5.6
5.6
5.6
5.6
6.4
RPV hight, m
20
22
22
22
22
23
Design pressure, MPa
8.24
8.24
8.24
8.24
8.24
8.62
PC Vessel height, m
32
33
33
34
34
48
Diameter
part), m
(cylindrical
10
11
11
11
11
10
Diameter
part), m
(spherical
18
20
20
20
20
25
1750
2980
2980
2980
2980
3200
0.43
0.38
0.38
0.38
0.38
0.28
Design temperature, C
140
140
140
140
138
171 (DW)
105 (SC)
Steam temperature, oC
282
282
282
282
282
282
Steam pressure, MPa
6.68
6.68
6.68
6.68
6.68
6.68
Type
UO2
UO2
UO2 (MOX)
UO2
UO2
UO2
69
94
94
94
94
132
2
1/1*
2
1/1*
2
2/1*
Electric output, MW
Start
of
operation
Reactor
Primary
Cont.
(PC)
Water in
Pool, t
commercial
Suppression
Design pressure, MPa
o
Turbine
Fuel
Core inventory, t
AC
Distributi
on
EDGs (*
cooled)
indicates
air
Electrical grid, # of lines
How to Improve Safety in Regulated Industries
What Could We Learn From Each Other
Background Material, ENCO FR-(12)-44
4 (275 kV) 4 (275 kV) 4 (275 kV) 4 (275 kV) 2 (500 kV) 2 (500 kV)
© ENCO
Page
3
FIG. 2-1. Fukush
hima Daiichi Units beforee earthquake and tsunam
mi (Source: EEPRI report [A1].
[
Table 2--3. Unit-speccific Emergency Core Co oling System
ms
Emergency Core Coo
oling system
ms (1)
Oth
her potentia
al cooling syystems
High
Pressure
Low Pressures
Hig
gh Pressure
Low Prressure
1
HPCI, IC
CS, CCS,
C
CCSW, SSHC
SLC
C, CRD
MUWC,, SFP, FP
2
HPCI, RCIIC
LPCI,, CCS, CCSW ,
SLC
C, CRD
MUWC,, SFP, FP
SLC
C, CRD
MUWC,, SFP, FP
Unit
(1) (2)
RHR, RHRW
3
HPCI, RCIIC
LPCI,, CCS, CCSW ,
RHR, RHRW
1) Syste
em Abbrevia
ations:
IC - Isolation Condenser system; HPCI - High Presssure Coolantt Injection; R
RCIC - Reac
ctor Core
Isola
ation Cooling
g system; SLC – Standby Liquid Conttrol system; CRD – Contrrol Rod Drive
e system;
CS - Core Spray; CCS – Con
ntainment C
Cooling syste
em; CCSW – Close Coolinng Sea Wate
er; LPCI Low Pressure Co
oolant Injecttion; SHC – SShutdown Co
ooling system
m; RHR - Ressidual Heat Removal;
RHRSS - Residual Heat Removal Seawate
er; MUWC - Makeup
M
Watter system; FP - Fire Protection
syste
em, SFP - Sp
pent Fuel Poo
ol Cooling syystem;
2) Syste
ems that can
n be used in emergency (based on a special line--up)
How to Improve Safety in Regulated Indu
ustries
What Could We Learn Fro
om Each Other
Background Material, ENC
CO FR-(12)-44
© ENCO
Page
4
F
FIG. 2-2. Gen
neral arrange
ements of Fu
ukushima Da
ai-ichi NPP (S
Source: INPO
O Report [A2]]).
FIG. 2-3. Generic cro
oss-section oof a BWR4 wiith a Mark I containmentt similar to
Unit 1-5 (SSource: INPO report [A2]))
How to Improve Safety in Regulated Indu
ustries
What Could We Learn Fro
om Each Other
Background Material, ENC
CO FR-(12)-44
© ENCO
Page
5
3. Seismic and Tsunami Design Basis
Because seismic and tsunami events were important factors in the accident at Fukushima
Dai-ichi the following sections summarize the methods used to define the design bases for
seismic and tsunami hazards at the plant. Criteria, methods, guidance, standards and
regulations referred to in this report are those used in Japan.
The design bases discussions take into consideration evaluations that have been performed
since the original design and any associated upgrades that have been implemented. It is
worth noting that treatment of the design bases differed for earthquake and tsunami.
Changes in the original seismic design basis resulted from a revised Japanese Regulatory
Guide issued in 2006 [A-3]. A tsunami assessment method document issued by the Japan
Society of Civil Engineers (JSCE) in 2002 [A-4] did not result in changes to the original
design bases, but TEPCO did perform a voluntary reassessment of the tsunami design bases
and implemented some plant design modifications.
3.1. Seismic
Japanese regulators first issued general seismic design guidance in 1978, with subsequent
revisions in 1981, 2001, and 2006 based in part on significant seismic events that occurred
after 1978 [A-3]. All plants, including Fukushima Daiichi, were required to be reviewed
(and upgraded structurally if necessary) for conformance with this guidance.
The 2006 requirements for seismic event definition and qualification are specified in the
Japan Nuclear Safety Commission (NSC) Regulatory Guide NSCRG: L-DS-I.02, entitled
“Regulatory Guide for Reviewing Seismic Design of Nuclear Power Reactor Facilities” [A-3].
The earthquakes that were defined for Fukushima Daiichi were upgraded after the issuance
of NSCRG: L-DS-I.02. The new earthquakes taken into account were much stronger than
the original earthquakes in the region, with frequencies less than 5 Hz. The zero period
accelerations are approximately 500 cm/s2, which is approximately 0.5 g. The probability
of occurrence of these earthquakes, an approach to quantifying “residual risks”, has also
been reported. The annual probability of exceedance for the response spectra was
reported to be 10-4 to 10-6 [A-5].
3.2. Tsunami
When the original licenses were issued for Fukushima Dai-ichi, Japanese regulatory
guidance only stated that “(the effect of the) tsunami should be considered in design” [A3]. There were no specific tsunami assessment numeric simulation methods available. For
design purposes, therefore, the tsunami height was set at 3.1 meters above sea level,
based on the observed wave height at Onahama port from the Chilean earthquake and
tsunami of May 24, 1960.
How to Improve Safety in Regulated Industries
What Could We Learn From Each Other
Background Material, ENCO FR-(12)-44
© ENCO
Page
6
Current Japanese regulatory guidance (in NSCRG L-DS-I.02 [A-4]) states that the “Safety
functions of the Facilities shall not be significantly impaired by tsunami which could be
reasonably postulated to hit in a very low probability in the service period of the
Facilities.” More detailed technical guidance for nuclear power plants in Japan was
provided in 2002 by the Japan Society of Civil Engineers (JSCE). An important aspect of the
2002 JSCE guidance was that it did not consider as credible that a tsunami could be caused
by ruptures across several fault segments in the vicinity of the plant. The JSCE guidance
stated that combined fault segments did not need to be considered for faults along the
Japan Trench (that encompasses the region of Fukushima) [A-4, A-6]. The March 11
earthquake occurred across numerous of the geological fault segments within the Japan
Trench, resulting in a larger-than-expected tsunami.
TEPCO applied the methods described in JSCE document [A-4] considering tsunamis
generated from eight different near-field sources off the coast of Japan. From this, it was
determined that the wave height could reach 5.7m at Fukushima Daiichi. As these changes
in criteria were voluntary, the licensing basis was not changed, although TEPCO made
changes to assure that all vital seawater pump motors were installed higher than 5.7 m.
In conjunction with the revised Japan seismic Regulatory Guide [A-3] issued in 2006 TEPCO
conducted a tsunami reevaluation using the methods in [A-4]. From this reevaluation, that
incorporated updated submarine topography and tide level data, it was determined that
the wave height could reach 6.1 m at Fukushima Daiichi and additional plant actions such
as sealing of pump motors were taken [A-6].
In 2008, calculations by TEPCO to characterize a potential tsunami source without an
established wave source model resulted in an estimated tsunami wave height of up to 10.2
m and a resulting flood inundation height of over 15 m at Fukushima Daiichi. Another
method that applied a wave source model of the Jogan tsunami in 869 A.D. resulted in an
estimated wave height exceeding 9 m. Neither of these estimates was applied to update
the design basis.
In 2009, the Nuclear and Industrial Safety Agency (NISA) asked that operators take into
account the Jogan earthquake for evaluating tsunami height “when new knowledge on the
tsunami of the Jogan earthquake is obtained”.
The new TEPCO survey results were reported in January 2011 and were inconsistent with
the estimate using the Jogan tsunami wave source model used in earlier calculations;
therefore TEPCO considered that it was necessary to conduct further investigation to
determine the Jogan tsunami wave source [A-6].
In the event of March 11, 2011 the actual tsunami maximum height of approximately
15°m, was about 5 m above plant grade. Based on the operational responses, this height
differential – along with the impact forces of the water (hydrodynamic effects) and debris
– was the dominant cause for eventual loss of all practical cooling paths, damage to the
reactor cores and uncontrolled release of radioactive materials to the environment.
4. March 11 earthquake and tsunami
4.1. Earthquake
The earthquake that occurred on March 11, 2011 at 14:46, was of magnitude 9.0 in Richter
scale. The epi-center of the earthquake was 180 km from the Fukushima Daiichi site and
How to Improve Safety in Regulated Industries
What Could We Learn From Each Other
Background Material, ENCO FR-(12)-44
© ENCO
Page
7
the hypocenter was 24 km under the Pacific Ocean. The earthquake lasted approximately
three minutes and resulted in the Japanese coastline subsiding an average of 0.8 meters.
The peak ground acceleration in the horizontal direction was 0.561 g and in the vertical
direction was 0.308 g, as measured at Unit 2. This exceeded the design basis acceleration
of 0.447 g in the horizontal direction. The design basis maximum acceleration was also
exceeded in units 3 and 5. Ratio of Measured to Design Basis acceleration was in the range
of 1.15 (Unit 3 – 1.26 (Unit 2). The design basis maximum acceleration in the vertical
direction was not exceeded in any of the units. According to the government of Japan, the
probability for exceeding the design basis acceleration was in the range of 10-4 to 10-6 per
reactor-year.
The response spectra at Units 2, 3, and 5 had the largest spectral discrepancy (exceedance
of the actual acceleration over the design acceleration) of the six units. It needs to be
noted that the exceedance of the actual spectra over the design spectra occurs primarily
at frequencies between 2.5 and 5 Hz. These are considered low frequencies and only a
small amount of safety-related equipment that has natural frequencies in this range.
The ground motion exceeded the reactor protection system setpoints, causing automatic
scrams. Control rods were inserted as expected. The power lines connecting the site to the
transmission grid were damaged during the earthquake, resulting in a loss of all off-site
power. The emergency diesel generators started and loaded as expected in response to the
loss of off-site power to supply electrical power, with the exception of one emergency
diesel generator on Unit 4, which was out of service for planned maintenance. Feedwater
and condensate pumps, which are powered by nonvital AC sources, were not available
because of the loss of AC power.
As the shaking from the earthquake subsided, the operators began their scram response.
All normal operator actions were taken to respond to the automatic plant shutdown.
Reactor pressure, reactor water level, and containment pressure indications for units 1, 2,
and 3 appeared as expected following a scram and did not indicate any potential breach of
the reactor coolant system (RCS) from the earthquake. However, no detailed walkdowns or
further investigation has been performed.
TEPCO activated its Headquarters for Major Disaster Countermeasures (Corporate
Emergency Response Center) in Tokyo to assess damage from the earthquake and to
support recovery efforts. The Station Emergency Response Center was activated on site to
respond to the event.
4.2. Tsunami
The earthquake generated a series of seven tsunamis that arrived at the site starting at
15:27, 41 minutes after the earthquake. The first wave was approximately 4 meters high.
The height of this wave did not exceed the site design basis tsunami of 5.7 meters and was
mitigated by the breakwater. A second wave arrived at 15:35; however, the wave height is
unknown, because the tide gauge failed (maximum indicated level of the gauge is 7.5
meters). At least one of the waves that arrived at the station measured approximately 14
to 15 meters high (based on water level indications on the buildings).
The tsunami inundated the area surrounding units 1-4 to a depth of 4 to 5 meters above
grade, causing extensive damage to site buildings and flooding of the turbine and reactor
How to Improve Safety in Regulated Industries
What Could We Learn From Each Other
Background Material, ENCO FR-(12)-44
© ENCO
Page
8
buildings. Figure 5-1 shows the general elevations (typical for units 1-4) and the
approximate inundation level. The grade level of units 1-4 is 10 meters and is 13 meters at
units 5 and 6 above mean sea level (commonly referred to as OP, for the level in the
Onahama Port). The intake structures were at an elevation of 4 meters for all units.
FIG. 5-1. General elevations and inundation level; Source EPRI Report [A-1]
The seawater intake structure was severely damaged and was rendered non-functional.
Intake structures at all six units were unavailable because the tsunamis and debris heavily
damaged the pumps, strainers, and equipment, and the flooding caused electrical faults.
The damage resulted in a loss of the ultimate heat sink for all units.
The diesel generators operated for a short time; but by 15:41, the combination of a loss of
cooling water, flooding of electrical switchgear, and flooding of some of the diesel
generator rooms (located in the basement of the turbine buildings and not designed to
withstand flooding) caused a loss of all AC power on site for units 1-5.
5. Plant capabilities and response
5.1. Plant status before the event
On March 11, 2011, Units 1, 2, and 3 were in operation at rated power output before the
event. Unit 1 had been in operation since September 27, 2010, Unit 2 – since September
23, 2010 and Unit 3 – since November 18, 2010.
In Units 1 – 3 all safety systems and both emergency diesel generators were operable. All
high pressure coolant injection systems (HPCI and both isolation condensers in Unit 1 and
HPCI and RCIC in Units 2 and 3) were available and in standby. Reactor water level and
pressure were normal for power operations. In Unit 3 the startup transformer was out of
service for planned modification work.
How to Improve Safety in Regulated Industries
What Could We Learn From Each Other
Background Material, ENCO FR-(12)-44
© ENCO
Page
9
FIG. 5-2. Plan view
w of the site (Units 1 to 4 only) show
wing flooded regions
follow
wing the tsun
nami; Source
e EPRI Reporrt [A-1]
ueling and maintenanc
m
ce activitiess. Unit 4
Units 4,, 5, and 6, were shut down for rooutine refu
had bee
en in an outtage since November
N
330, 2010, Unit
U
5 – sinc
ce January 33, 2011, an
nd Unit 6
– since A
August 14, 2010.
The Uniit 4 reactor fuel was off-loaded to the Uniit 4 spent fuel
f
pool too facilitate reactor
pressure
e vessel shroud work.. The cavitty gate wass installed, isolating tthe spent fuel pool
from th
he upper po
ools. The 4A
A emergenccy diesel generator was
w out of sservice for planned
mainten
nance, with
h the 4B em
mergency die
esel genera
ator operable and in sttandby.
In Unit 5 fuel had
d been loa
aded into tthe reactorr and the reactor
r
preessure vesse
el (RPV)
reassem
mbled. Reacctor water level
l
was h igh, reacto
or coolant sy
ystem tempperature wa
as 89°C,
and rea
actor pressu
ure was 7.15 MPa gaugge to suppo
ort RPV leak
k testing. D
Decay heat removal
was seccured at 07::44 in prepa
aration for the leak te
esting. Both
h emergenccy diesel generators
were op
perable.
In Unit 6 fuel had been loaded into the
e reactor and the RPV
V reassemblled. Reacto
or water
level w
was normal,, and reacttor coolantt system te
emperature
e was 26°C
C with the reactor
coolant system de
epressurized
d. Residuall heat removal (RHR) system B was being used as
needed for decay heat
h
remov
val. All thre
ee emergenc
cy diesel ge
enerators w
were operab
ble.
How to Improve Safety in Regulated Indu
ustries
What Could We Learn Fro
om Each Other
Background Material, ENC
CO FR-(12)-44
© ENCO
Page
10
5.2. Loss of power
All off-site AC power was lost as a result of the earthquake. The emergency diesel
generators started at all six units providing alternating current (AC) electrical power to
critical systems at each unit, and the facility response to the seismic event appears to
have been normal.
The tsunami resulted in extensive damage to site facilities and a complete loss of AC
electrical power at Units 1 through 5, a condition known as station blackout (SBO). Unit 6
retained the function of one of the emergency diesel generators (air-cooled). The
operators were able to successfully cross-tie the single operating Unit 6 air-cooled diesel
generator to provide sufficient AC electrical power for Units 5 and 6 to place and maintain
those units in a safe shutdown condition, eventually achieving and maintaining cold
shutdown. All DC power was lost on units 1 and 2, while some DC power from batteries
remained available on Unit 3.
The loss of on-site AC power was caused by the submergence of the emergency diesel
generators and electrical distribution system equipment inside the plants. Water
penetrated to the reactor building through DG louvres, doors, hatch, trenches and ducts.
Loss of DC power in Units 1 and 2 was caused by submergence of electrical distribution
system equipment. The loss of DC power in Unit 3 to some systems was caused by
submergence of electrical distribution system equipment and then eventually by full
discharge of the batteries. Table 6-1 shows causes of power supply problems for Units 1-3
in more detail.
Table 6-1. Causes for Unavailability of Power Source Following the event, [A-2]
Unit
Unit 1
Power
source
Off-site
power
EDG
6.9 kV AC
480V AC
Unit 2
125V DC
Off-site
power
EDG
6.9 kV AC
Description of unavailability cause
The receiving circuit breaker of the Ookuma Line 1L in the Unit 1 / 2
switchyard was damaged by the earthquake
Submergence of both emergency diesel generators due to the tsunami
Submergence of the 6.9 kV high voltage AC power supply panels due
to the tsunami
Submergence of the 6.9 kV high voltage AC power supply panels due
to the tsunami
Submergence of the 480V low voltage AC power supply panels due to
the tsunami
Submergence of the 125V DC power supply panels due to the tsunami
The receiving circuit breaker of the Ookuma Line 2L in the Unit 1 / 2
switchyard was damaged by the earthquake
The circuit breaker for the Ookuma Line 2L in the New Fukushima
substation was damaged by the earthquake
One emergency diesel generator was submerged in water due to the
tsunami and the power source panels for another, air cooled
emergency diesel generator, was submerged due to the tsunami
Submergence of the 6.9 kV high voltage AC power supply panels due
How to Improve Safety in Regulated Industries
What Could We Learn From Each Other
Background Material, ENCO FR-(12)-44
© ENCO
Page
11
Unit
Power
source
480V AC
Unit 3
125V DC
Off-site
power
EDG
6.9 kV AC
480V AC
125V DC
Description of unavailability cause
to the tsunami
Submergence of the 6.9 kV high voltage AC power supply panels due
to the tsunami
Partial submergence of the 480V low voltage AC power supply panels
due to the tsunami
Submergence of the 125V DC power supply panels due to the tsunami
Ookuma Line 4L was damaged by the earthquake between the plant
switchyard and the off-site substation, Ookuma Line 3L was out of
service for planned renovation work
Submergence of both emergency diesel generators due to the tsunami
Submergence of the 6.9 kV high voltage AC power supply panels due
to the tsunami
Submergence of the 6.9 kV high voltage AC power supply panels due
to the tsunami
Submergence of the 480V low voltage AC power supply panels due to
the tsunami
The DC power supply batteries were exhausted
Three air-cooled emergency diesel generators (EDGs) had previously been installed at the
station as a modification (2B, 4B, and 6B EDGs). These EDGs had independent fuel systems
and were capable of providing power to vital AC systems following a complete loss of the
seawater ultimate heat sink. The air-cooled EDGs were located above grade, and some of
them survived the tsunami. The distribution systems for the Unit 2 and the Unit 4 aircooled EDGs flooded and failed during the tsunami. The Unit 6 air-cooled EDG and portions
of the electrical distribution system survived the tsunami and were used to reestablish cold
shutdown on units 5 and 6.
When all AC power was lost, TEPCO was able to secure some mobile generators from the
Tohoku Electric Power Company. These generators, along with some TEPCO generators,
began to arrive at the site late in the evening of March 11 and continued to arrive into the
next morning.
The portable generators were limited in their effectiveness because they could not be
connected to the station electrical distribution system as a result of the extensive damage
the tsunami and flooding caused. Workers checked motors and switchgear in an attempt to
find usable equipment to support cooling the reactors. The testing revealed that the Unit 2
standby liquid control (SLC) pumps were not flooded or damaged.
Based on the inspection results, the first mobile generator was placed adjacent to Unit 2,
and workers began to lay temporary cables from the generator to the associated
distribution panel for the SLC pumps. The temporary power cables were approximately 10
cm in diameter and 200 meters long and weighed more than 1 ton. Aftershocks and
subsequent tsunami warnings further slowed progress. In spite of the challenges, the
workers completed the task on Unit 2 and terminated the temporary cable to the
associated power panel on March 12 at 15:30.
At 15:36, an explosion occurred in the Unit 1 reactor building. The explosion injured
five workers, and debris from the explosion struck and damaged the cables and mobile
How to Improve Safety in Regulated Industries
What Could We Learn From Each Other
Background Material, ENCO FR-(12)-44
© ENCO
Page
12
generator that had been installed to provide power to the standby liquid control
pumps. The debris also damaged the hoses that had been staged to inject seawater
into Unit 1 and Unit 2. Fieldwork had to be suspended. The explosion significantly
altered the response to the event and contributed to complications in stabilizing the
units.
5.3. Core cooling
With all off-site AC power lost as a result of the earthquake and loss of on-site power
caused by submergence of electrical distribution system equipment following tsunami.
Without AC power, the operators were relying on batteries and turbine-driven and dieseldriven pumps. Steam-driven injection pumps were used to provide cooling water to the
reactors on Units 2 and 3, but these pumps eventually stopped working; and all cooling
water to the reactors was lost until fire engines were used to restore water injection.
Operators were trying to maintain core cooling functions well beyond the normal capacity
of the station batteries. Without the response of offsite assistance, which appears to have
been hampered by the devastation in the area, among other factors, each unit eventually
lost the capability to further extend cooling of the reactor cores.
Cooling was lost to the fuel in the Unit 1 reactor after ~11 hours, the Unit 2 reactor after
about 71 hours, and the Unit 3 reactor after about 36 hours, resulting in damage to the
nuclear fuel shortly after the loss of cooling. Core cooling was eventually established when
a fire engine was used to inject seawater.
With no core cooling to remove decay heat, core damage begun on Unit 1 on the day of the
event. As a result of inadequate core cooling, fuel damage also occurred in units 2 and 3.
Inadequate core cooling resulted in subsequent fuel damage. Conservative calculations
indicate that some of the fuel may have relocated to the bottom head of the reactor
vessel, although this has not been confirmed.
Sequence and timing of events is presented in Table 6-2 [A-2]. This overview of the events
as they occurred is limited to Units 1–3 and shows only those items considered of
significance.
Table 6-2. Timeline of Key Cooling Systems Failures
Unit
All
All
All
All
All
1
1
1
1
1
System*
AC power
DC power
HPCI
SLC, CRD
CCS, CCSW, MUWC, SFP
How to Improve Safety in Regulated Industries
What Could We Learn From Each Other
Background Material, ENCO FR-(12)-44
Hrs
0.00
0.02
0.02-0.03
0.68
0.82
0.85
0.85
0.85
0.85
0.85
Action
Earthquake
Scram signal initiated, control rods inserted
EDGs start on loss of offsite power
First tsunami wave
Second tsunami wave
Unit 1 AC lost
Unit 1 DC lost
Non-functional due to loss of AC power
Non-functional due to loss of AC power
Non-functional due to loss of AC power
© ENCO
Page
13
Unit
1
1
2
2
2
2
2
2
2
3
3
3
3
3
3
3
System*
IC
AC power
DC power
CS, RHR, RHRS
SLC, CRD
HPCI
RCIC
AC power
DC power
CS, RHR, RHRS
SLC, CRD
RCIC
HPCI
-
Hrs
11.03
24.83
0.92
0.92
0.92
0.92
0.92
70.65
87.23
0.87
0.87
0.87
0.87
35.93
35.93
68.25
Action
Unit 1 IC lost
Unit 1 reactor building hydrogen explosion
Unit 2 AC lost
Unit 2 DC lost
Non-functional due to loss of AC power
Non-functional due to loss of AC power
Non-functional due to loss of AC power
Unit 2 RCIC lost
Unit 2 loss of primary containment
Unit 3 AC lost
Unit 3 DC lost
Non-functional due to loss of AC power
Non-functional due to loss of AC power
Unit 3 RCIC lost
Unit 3 HPCI lost
Unit 3 reactor building hydrogen explosion
*) See Table 6-1 for information on system abbreviations
The overview of the events provides the timeline on a per Unit basis. This sequence of
events provides only the level of detail necessary for generating input to the analysis of
accident causes and consideration of potential solutions.
The time of the first seismic ground motion is considered the baseline and all subsequent
events are identified in terms of differential time from this baseline. The timelines are not
inclusive, but focus on systems that could provide cooling functions (as presented in Table
2-3). Information on the unavailability of emergency core cooling systems refers also to
those systems that could provide alternative core cooling sources based on special line-ups
(such as SLC, CRD, etc.). The timelines continue until the safety systems become
unavailable. Information on timing of hydrogen explosions is also included.
5.4. Hydrogen explosions
Hydrogen generated from the damaged fuel in the reactors accumulated in the reactor
buildings - either during venting operations or from other leaks - and ignited, producing
explosions in the Unit 1 and Unit 3 reactor buildings and further damaging the facilities and
primary and secondary containment structures. The Unit 1, 2, and 3 explosions were
caused by the buildup of hydrogen gas within primary containment produced during fuel
damage in the reactor and subsequent movement of that hydrogen gas from the drywell
into the secondary containment. The source of the explosive gases causing the Unit 4
explosion remains unclear.
The most widely accepted theory is associated with the backflow of gases from Unit 3
during venting. The containment vent exhaust piping from Unit 3 is connected to the Unit 4
exhaust piping. The dampers on the Unit 4 standby gas treatment system (SGTS) are airHow to Improve Safety in Regulated Industries
What Could We Learn From Each Other
Background Material, ENCO FR-(12)-44
© ENCO
Page
14
operated and fail open on a loss of power or air (except the cross-connect between SGTS
filter trains). Additionally, the system does not have a backflow damper installed in the
piping that connects to Unit 3. With no power or air, and no fans in service to direct the
gases from Unit 3 up the exhaust stack, the exhaust gases from Unit 3 would be directly
aligned to the Unit 4 SGTS filters. This piping arrangement may have allowed gases from
the Unit 3 containment to be vented to the Unit 4 reactor building via reverse flow through
the Unit 4 standby gas treatment system.
Hydrogen explosions significantly complicated the response to the accident. In addition,
the operators were unable to monitor the condition of and restore normal cooling flow to
the Unit 1, 2, 3, and 4 spent fuel pools.
5.5. Containment pressure control
Without heat removal systems containment pressure and temperature started to increase
as energy from the reactor is transferred to the containment via safety relief valves or
systems such as RCIC and HPCI.
The TEPCO severe accident procedures allow venting when containment pressure reaches
the maximum operating pressure if core damage has not occurred. If core damage has
occurred, venting the containment will result in a radioactive release, so containment is
not vented until pressure approaches twice the maximum operating pressure. In this case,
the Emergency Response Center personnel could not verify the integrity of the core and
this guidance was applied. The decision to vent Unit 1 was made by the site
superintendent with concurrence from government agencies. This was planned for March
12 after evacuation that was scheduled to be completed at 9:00.
The first indication of increasing containment pressure was not available until 23:50 on the
night of the event, when workers connected the temporary generator – which was being
used to provide some control room lighting – to the containment pressure instrument. The
indication read 600 kPa. By this point, access to the reactor building had already been
restricted because of high dose rates. The lack of available containment pressure
indications early in the event may have prevented the operators from recognizing the
increasing pressure trend and taking action earlier in the event.
Unit 1 containment was not vented successfully until approximately 14:30 on March 12.
Additional challenges occurred because of high dose rates and a lack of contingency
procedures for operating the vent system without power, as well as the lack of equipment,
such as an engine-driven air compressor.
The decision to complete evacuations before venting containment, and the subsequent
equipment and radiological challenges encountered as operators attempted to establish a
vent path, delayed injection of water into the Unit 1 reactor. At approximately 02:30 on
March 12, as Unit 1 depressurized, pressure in the reactor and in containment equalized at
approximately 0.84 MPa abs. This pressure is above the discharge pressure of the station
fire pumps and fire engines. Once pressure had equalized, further reductions in reactor
pressure were not possible until containment pressure had lowered. As a result, little to no
injection was achieved until after the containment was vented successfully, which
occurred at approximately 14:30 on March 12.
How to Improve Safety in Regulated Industries
What Could We Learn From Each Other
Background Material, ENCO FR-(12)-44
© ENCO
Page
15
High containment pressures in Unit 1 contributed to the amount of time Unit 1 did not
have adequate core cooling. In units 1, 2, and 3, the extended duration of high
temperature and pressure conditions inside containment may have damaged the drywell
head seals, contributing to hydrogen leaks and the subsequent explosions. Containment
leakage also contributed to ground-level radiation releases from units 1, 2, and 3.
5.6. Spent Fuel Pools and Dry Cask Storage
Fukushima Daiichi had spent fuel stored in pools at each unit, in a common spent fuel pool,
and in on-site dry cask storage. Spent fuel pool cooling flow was lost for all spent fuel
pools following the loss of off-site power and was not immediately restored when the
emergency diesel generators started.
The explosion in the Unit 4 reactor building caused structural damage to the Unit 4 spent
fuel pool, but it is not clear if the integrity of the pool liner was compromised. Subsequent
analysis and inspections performed by TEPCO personnel determined that the spent fuel
pool water levels did not drop below the top of fuel in any spent fuel pool and that no
significant fuel damage had occurred. Current investigation results indicate that any
potential fuel damage was likely caused by debris from the reactor building explosions.
The dry cask storage building was damaged by the tsunami, and some of the casks were
wetted. An inspection confirmed that the casks were not damaged by the event.
5.7. Alternative injection sources
Fukushima Daiichi had three fire engines available that had been added to improve firefighting capabilities following the 2007 Niigata-Chuetsu-oki earthquake. These fire engines
could also be used as an alternative low-pressure water source for injecting into the
reactors. However, one was damaged by the tsunami and a second could not reach units 14 because of earthquake damage to the road. Only one fire engine was immediately
available to support the emergency response on units 1-4.
Using this fire engine was complicated because the fire engine did not have sufficient
discharge pressure to overcome the elevation differences and reactor pressure. To
compensate for this, the truck loaded water at the fire protection tank, then drove to the
Unit 1 reactor building to inject into the fire protection system. This operation was slowed
by debris on the road. Finally, a suction hose was installed to provide connection from the
fire protection tank to the track, and then discharging to the fire protection system piping
and into the reactor via an installed modification to the low pressure coolant injection
system.
The fire protection tank, however, only had one hose connection. As a result, injection
into the reactor had to be stopped each time the tank needed to be refilled so another fire
engine, now available, could attach a hose and fill the tank. Seawater injection was
eventually switched to a flooded pit, then to the harbour itself.
How to Improve Safety in Regulated Industries
What Could We Learn From Each Other
Background Material, ENCO FR-(12)-44
© ENCO
Page
16
5.8. Radiological consequences
The loss of primary and secondary containment integrity resulted in ground-level releases
of radioactive material. Following the explosion in Unit 4 and the abnormal indications on
Unit 2 on the fourth day of the event, the site superintendent directed that all
nonessential personnel temporarily evacuate, leaving approximately 70 people on site to
manage the event.
During releases, dose rates as high as 1,193 millirem per hour (mrem/hr) (11.93 mSv/hr)
were measured at the site boundary, approximately 1 km from units 1 - 4. The windows for
the emergency response center had to be covered with lead shielding to reduce dose rates
in the center. Organized off-site radiation surveys began on March 16. Radiation levels off
site at that time ranged from 0.1 mrem/hr (1 μSv/hr) to 20 mrem/hr (200 μSv/hr). 60 km
northwest of the station, the dose rate was 0.8 mrem/hr (8 μSv/hr). Water and soil
samples indicated the presence of strontium, iodine, and cesium. Food and water
restrictions were implemented in some areas as a result of radioactivity.
People within the 20 km surrounding the station were evacuated, and those living up to 30
km away were directed to shelter inside their homes as the releases of radioactive gases
and materials increased as the event progressed and more fuel damage occurred.
Potassium iodide tablets and powder were distributed to local governments beginning
March 21. Because the evacuations had already been completed, however, the potassium
iodide was not issued to the population.
Radiation surveys of the on-site areas surrounding units 1 - 3 showed dose rates as high as
13 rem/hr (0.13 Sv/hr) in areas around Units 2 and 3. More detailed surveys performed
over the following weeks discovered localized dose rates greater than 1000 rem/hr (10
Sv/hr) around equipment and debris outside units 1 and 3.
Some personnel who responded to the event received high doses of radiation. Two control
room operators received the highest doses a calculated internal and external dose of 67.8
rem (0.678 Sv) and 64.3 rem (0.643 Sv). The majority of dose received by these workers
was internal (85-87 percent). Potassium iodide was provided to some station personnel on
March 13. As of the end of March, approximately 100 workers had received doses eceeding
10 rem (0.1 Sv).
The Fukushima event was rated as a level 7 event on the International Nuclear and
Radiological Event (INES) scale. The Nuclear Safety Commission of Japan estimated
approximately 17 million curies (6.3 E17 Bq) of iodine-131 equivalent radioactive material
was released into the air and 0.127 million curies (4.7 E15 Bq) into the sea between March
11 and April 5.
6. Causal analysis
Generally, the causal analysis is performed as an important step to finding effective
solutions to identified problems in order to prevent similar problems from recurring.
Analysis of this type is well known as Root Cause Analysis (RCA). RCA is intended to identify
specific causes and the associated solutions through which the problem owner may have
control of these causes in order to eliminate problem or reduce its consequences.
How to Improve Safety in Regulated Industries
What Could We Learn From Each Other
Background Material, ENCO FR-(12)-44
© ENCO
Page
17
Appropriate application of Root Cause Analysis techniques can yield significant
organizational and individual benefits in every human endeavour.
RCA conducted in this project is intended to provide lessons that could be of value for the
nuclear sector, but also those that could be shared across the other regulated industries.
From this point of view, the analysis should be broad enough to identify causes (and
potential solutions) related to organizational aspects, including general issues of human
performance and safety culture. Due attention will also be put on the regulatory concept
of dealing with low probability high consequence events. From technical point of view,
focus is on specific elements of protection, mitigation, and preparedness and evaluation of
current capabilities, limitations, and potential enhancements.
6.1. Existing studies
EPRI analysis
The technical analysis performed by EPRI [A-1] traced the cause for the eventual loss of all
practical cooling paths for the reactors to the tsunami’s flooding of the plant protection.
Specifically, the analysis identified the significant difference between the design basis
tsunami height and the actual tsunami height, as well as the limitations of beyond-designbasis tsunami protection or mitigation that could address the effects of the actual event.
From a causal analysis perspective, these were caused by a methodology that specified
that the rupture of combinations of geological fault segments in the vicinity of the plant
need not be considered in establishing the design basis tsunami height. The tsunami that
occurred was caused by a combined rupture of multiple offshore fault segments.
The analysis identified other causes of condition type that were important from the point
of view of the accident severity, mainly "elevations of critical SSCs … below the actual
tsunami level", limited historical records for tsunami, and "limited regulatory guidance for
beyond design basis accidents".
The Cause and Event Chart shown in Fig. 6-1 displays the underlying technical causal
factors.
Technical analysis of the Fukushima Daiichi March 2011 accident conducted by EPRI [A-1]
was intended to determine the fundamental cause for the loss of substantial systems
needed to maintain reactor cooling. The loss of these systems resulted in core damage to
reactors at the site and uncontrolled release of radioactive materials to the environment
from the site.
From this information and review of the capabilities needed to provide core cooling, it is
clear that essentially all plant equipment needed to support core cooling was damaged by
the initial effects of the tsunami event.
Other factors outside of the initial effects of the tsunami may have contributed to the
extreme challenges encountered in attempts to sustain and/or reestablish cooling.
However, the focus of this analysis was on the cause of the loss of the safety systems that
would normally be used to maintain the integrity of the core. The loss of those safety
systems was a result of a tsunami that exceeded the design basis of the plants.
How to Improve Safety in Regulated Industries
What Could We Learn From Each Other
Background Material, ENCO FR-(12)-44
© ENCO
Page
18
F
FIG.6-1.
Cau
usal analysis chart (the source:
s
EPRI report [A-1]])
unami prote
ection strategy for th
he plant co
onsisted of locating ccritical equ
uipment,
The tsu
such as vital seaw
water pump motors, ab
bove the ellevation of the assesseed tsunamii height.
The asssessment off the tsunam
mi height w
was based on
o the 2002
2 tsunami aassessment method
for Nucclear Powerr Plants in Japan, the
e accepted methodolo
ogy by the Japanese industry.
Followin
ng the guid
dance of thiis methodollogy, offsho
ore fault segments we re not combined in
the tsunami assessment. Durring the eaarthquake event
e
nume
erous fault segments acted
a
in
combina
ation, and thus the ac
ctual tsunam
mi caused by
b the earth
hquake signnificantly ex
xceeded
the tsun
nami assesssment for th
he plant.
Fundam
mentally, fo
ollowing the
e accepted
d tsunami assessment
a
technical guidance in Japan
resulted
d in under--prediction of the sizze of the tsunami.
t
Ass a result, the plant tsunami
protection strategy was not adequaate and beyond-design-basis tssunami protection
adequatte to mitiga
ate the effe
ects of the tsunami tha
at occurred
d was not avvailable.
NRC re
ecommend
dations
The stu
udy was prepared
p
by
b the Neaar-Term Task
T
Force establisheed in response to
Commisssion directtion to con
nduct a syystematic and
a
method
dical review
w of U.S. Nuclear
Regulattory Commission proce
esses and re
egulations to
t determin
ne whetherr the agency
y should
make additional im
mprovemen
nts to its re
egulatory sy
ystem and to make reecommenda
ations to
the Com
mmission fo
or its policy
y direction,, in light off the accide
ent at the Fukushima Dai-ichi
Nuclearr Power Pla
ant [A-9]. The study was condu
ucted by te
eam of in-hhouse expe
erts who
collectively had ovver 130 years of reactoor regulatorry experience.
How to Improve Safety in Regulated Indu
ustries
What Could We Learn Fro
om Each Other
Background Material, ENC
CO FR-(12)-44
© ENCO
Page
19
In examining the Fukushima Dai-ichi accident for insights for reactors in the United States,
the Task Force addressed protecting against accidents resulting from natural phenomena,
mitigating the consequences of such accidents, and ensuring emergency preparedness.
As part of its undertaking, the Task Force studied the manner in which the NRC has
historically required protection from natural phenomena and how the NRC has addressed
events that exceed the current design basis for plants in the United States.
In general, the Task Force found that the current NRC regulatory approach includes:
-
Requirements for design-basis events with protection and mitigation features
controlled through specific regulations or the general design criteria (10 CFR-50))
Requirements for some “beyond-design-basis” events through specific regulations
(e.g., station blackout, large fires, and explosions)
Voluntary industry initiatives to address severe accident features, strategies, and
guidelines for operating reactors
This regulatory approach, has been established and supplemented piece-by-piece over the
decades, addressed many safety concerns and issues, using the best information and
techniques available at the time. The result is a patchwork of regulatory requirements and
other safety initiatives, all important, but not all given equivalent consideration and
treatment by licensees or during NRC technical review and inspection. Consistent with the
NRC’s organizational value of excellence, the Task Force believes that improving the NRC’s
regulatory framework is an appropriate, realistic, and achievable goal.
The current regulatory approach, and more importantly, the resultant plant capabilities
allow the Task Force to conclude that a sequence of events like the Fukushima accident is
unlikely to occur in the United States and some appropriate mitigation measures have been
implemented, reducing the likelihood of core damage and radiological releases. Therefore,
continued operation and continued licensing activities do not pose an imminent risk to
public health and safety.
However, the Task Force also concludes that a more balanced application of the
Commission’s defense-in-depth philosophy using risk insights would provide an enhanced
regulatory framework that is logical, systematic, coherent, and better understood. Such a
framework would support appropriate requirements for increased capability to address
events of low likelihood and high consequence, thus significantly enhancing safety.
Excellence in regulation demands that the Task Force provide the Commission with its best
insights and vision for an improved regulatory framework.
The report, among other things, recommends:
-
Requiring plants to reevaluate and upgrade as necessary their design-basis seismic and
flooding protection of structures, systems and components for each operating reactor
and reconfirm that design basis every 10 years;
-
Strengthening Station Black Out (SBO) mitigation capability for existing and new
reactors for design-basis and beyond-design-basis natural events – such as floods,
hurricanes, earthquakes, tornadoes or tsunamis – with a rule to set minimum coping
time without offsite or onsite AC power at 8 hours; establishing equipment,
procedures and training to keep the core and spent fuel pool cool at least 72 hours;
and preplanning and pre-staging offsite resources to be delivered to the site to
support uninterrupted core and pool cooling and coolant system and containment
integrity as needed;
How to Improve Safety in Regulated Industries
What Could We Learn From Each Other
Background Material, ENCO FR-(12)-44
© ENCO
Page
20
-
Requiring that facility emergency plans address prolonged station blackouts and
events involving multiple reactors;
-
Requiring additional instrumentation and seismically protected systems to provide
additional cooling water to spent fuel pools if necessary; and requiring at least one
system of electrical power to operate spent fuel pool instrumentation and pumps at
all times. The Task Force noted it will take some time for a full understanding of the
sequence of events and condition of the spent fuel pools. The report said based on
information available to date the two most cogent insights related to the availability
of pool instrumentation and the plant’s capability for cooling and water inventory
management;
-
Requiring reliable hardened vent designs in boiling water reactors (BWRs) with Mark I
and Mark II containments;
-
Strengthening and integrating onsite emergency response capabilities such as
emergency operating procedures, severe accident management guidelines and
extensive damage mitigation guidelines.
-
6.2. Cause Mapping
This section of the report provides the results of causal analysis performed specially for
this project. The results are presented in the form of Cause Map (CM). It displays the
whole structure of causes in a graphical form. This form of presentation is believed to
facilitate effective communication and documentation of causes of the problem (accident)
[A-10]. It is worth noting that communication of findings to experts from different
industries and of different professions is an important aspect in this project.
The CM for the Fukushima accident was developed and presented in MS Excel using the
worksheet / template prepared by "ThinkReliability" Consulting Company available at web
site page http://www.thinkreliability.com [A-11].
The CM was prepared for the accident at Fukushima Unit 3. Although there are some minor
differences among the units the causal map prepared for this unit represents very well the
situation at other units. A similar analysis could be put together for all of the units
affected by the earthquake, tsunami and resulting events. Parts of this cause map could be
reused as many of the issues affecting the other plants and units are similar to the analysis
shown here.
Step 1 - Definition of the problem
The first step of the Cause Mapping approach is to define the problem by asking the four
questions: What is the problem? When did it happen? Where did it happen? And how did it
impact the goals? Answer to these questions are provided in Table 6-1.
How to Improve Safety in Regulated Industries
What Could We Learn From Each Other
Background Material, ENCO FR-(12)-44
© ENCO
Page
21
Table 6-1. Definition of the problem
What
Problem(s)
Fukushima/Daiichi tsunami
When
Date
March 11, 2011
Time
Where
Different, unusual, unique
Large earthquake/ tsunami;
State, city
Fukushima, Japan
Facility, site
Daiichi nuclear power plant
Unit, area, equipment
Unit 3
Task being performed
Operating at full power
Impact to the Goals
Safety
11 workers injured
Public Safety
Potential for health impacts
Environmental
Release of radiation to the environment
Cust. Service
Evacuation of public within 20 km
Rolling blackouts
Production-Schedule
Loss of electrical production capacity
Property, Equip, Mtls
Catastrophic damage to plant
Labor, Time
Massive efforts to cool reactor
Frequency
Very rare
The impact to goals needs to be determined prior to building a Cause Map. As a
direct result of the events at Unit 3, 11 workers were injured. This is an impact to
the worker safety goal. There is the potential for health effects to the population,
which is an impact to the public safety goal. The environmental goal was impacted
due to the release of radioactivity into the environment. The customer service goal
was impacted due to evacuations and rolling blackouts, caused by the loss of
electrical production capacity, which is an impact to the production goal. The loss
of capacity was caused by catastrophic damage to the plant, which is an impact to
the property goal. Additionally, the massive effort to cool the reactor is an impact
to the labor goal.
The issues surrounding Unit 3 are extremely complex. In events such as these,
where many events contribute to the issues, it can be helpful to make a timeline of
events. A timeline of the events is shown in Fig. 6-2.
How to Improve Safety in Regulated Industries
What Could We Learn From Each Other
Background Material, ENCO FR-(12)-44
© ENCO
Page
22
Table 6-2. Timing of events (Unit 3)
Diff.
(hrs)
Date
Time
March 11,
2011
14:46
0.00
Earthquake of magnitude 9
14:47
14:47
14:48
15:05
0.02
0.02
0.03
0.32
15:25
0.65
15:27
15:35
15:38
15:38
15:38
15:38
16:03
20:15.
21:58
0.68
0.82
0.87
0.87
0.87
0.87
1.28
5.32
6.20
22:00
07:00
6.23
16.23
Scram signal initiated, control rods inserted
Off-site power at Fukushima Daiichi lost
Automatic startup of emergency diesel generators (EDG)
Operators initiated RCIC to maintain reactor pressure and water
level
RCIC automatically shut down because of a high reactor water
level
The first of a series of tsunamis
The largest tsunami hits
Flooding in the turbine building basement
Loss of all AC power
Loss of all DC power
Partial loss of the control board instrumentation and controls
RCIC manually restarted
Emergency declared at Daiichi power plant
A small portable generator used to restore lighting in the units 3-4
MCR
Evacuation of local residents within 3 km radius
Evacuation of local residents within 10 km radius
11:36
12:35
15:36
17:00
19:11
20:36
21:00
22:35
02:42
20.83
21.8
24.83
26.23
28.42
29.83
30.2
31.82
35.9
02:42
35.93
04:15
04:50
37.48
38.1
05:00
38.23
05:10
07:35
38.39
40.8
8:41 9:20
09:10
09:25
42.4
42.65
11:17
44.51
March 12,
2011
March 13,
2011
Description
RCIC shut down unexpectedly and could not be restarted
HPCI automatically started on a low-low reactor water level signal
Hydrogen explosion at the Dai-ichi Unit 1 reactor building
Reactor pressure indicated 2.9 MPa gauge and lowering.
Evacuation expands to 20 km around the Daiichi plant
Reactor water level indication lost
Operators started a review of the vent procedures
Iodine tablets distributed
HPCI system tripped, DC power was failing and RP was low (0.58
MPa gauge),
HPCI could not be restarted due to depleted batteries, failure to
restart RCIC locally
The reactor core started to uncover
Unsuccessful attempt to open the large AOV to vent suppression
chamber
Reactor pressure > 7.38 MPa gauge, reactor water level 2,000 mm
below TAF and lowering, and containment (CT) pressure - 0.36 MPa
abs.
Unable to confirm level of water injection to the reactor by RCIC
Reactor water level had lowered to the bottom of the fuel zone,
the core uncovered
Both CT vent valves open; SRV manually opened to depressurize
the reactor
The maximum indicated containment pressure - 0.637 MPa abs
SRV open; the RP decreased sufficiently to start borated fresh
water injection;
The suppression chamber vent valve (AO-205) was found closed.
How to Improve Safety in Regulated Industries
What Could We Learn From Each Other
Background Material, ENCO FR-(12)-44
© ENCO
Page
23
Date
March 14,
2011
March 16,
2011
March 17,
2011
Time
Diff.
(hrs)
13:12
06:50
46.43
64.06
Injection of sea water and boric acid into reactor vessel
Pressure in the reactor containment vessel increased to 0.53 MPa
11:01.
08:30
68.25
113.7
6
135.3
8
139.0
3
Hydrogen explosion at the Dai-ichi Unit 3 reactor building
Reports of steam coming from the reactor building
06:15
09:48
March 18,
2011
19 20:09
14 14:45
16:00
169.2
3
Description
Increase in pressure of the suppression chamber
Water discharge by Self-Defense Force's helicopters
Water discharge by HP water cannon trucks and Self-Defense
Force's fire engines
Water discharge by Self-Defense Force's fire engines and US army's
fire engines
New electrical transmission line connected
Step 2 – Analysis of causes (Causal Map)
Catastrophic damage to the plant was caused by the hydrogen explosion and severe core
damage. Release of radioactive material to the environment was caused by venting of the
containment and in the later phase of the accident by the loss of containment boundary
due to hydrogen explosion in the reactor building.
Venting of the containment was undertaken in order to decrease the containment pressure
that was too high. Buildup of the containment pressure was caused by the lack of
containment cooling and heating of the containment. Without heat removal systems (no AC
power and a loss of ultimate heat sink), containment pressure and temperature increases
as energy from the reactor is transferred to the containment via safety relief valves (SRV)
or systems such as RCIC and HPCI.
How to Improve Safety in Regulated Industries
What Could We Learn From Each Other
Background Material, ENCO FR-(12)-44
© ENCO
Page
24
Why?
Effect
Possible Solutions:
Cause
Cause
Evidence:
Start with the Goals (in red) that have been impacted.
Read the map to the right by asking Why questions.
Step 2. Cause Map - Page 1
Customer
Service Goal
Impacted
Rolling
blackout
Workers
Safety Goal
Impacted
11 workers
injured
Hydrogen
explosion
Loss of
electrical
power capacity
Public Safety
Goal
Impacted
Production
Goal
Impacted
Potential for
health impact
Customer
Service Goal
Impacted
Labor
Goal
Impacted
Evacuation of
people
within 20 km
Significant
efforts to cool
the reactor
Catastrophic
damage to the
plant
Property
Goal
Impacted
Release of
radioactivity to
environment
Environmental
Goal
Impacted
Severe core
damage
Loss of
containment
boundary
AND / OR
Venting
radioactive
steam
CT pressure
too high
Following loss of RCIC and HPCI, the release of steam from the reactor system via SRV was
performed by the personnel in an attempt to depressurize the system. Reactor pressure
was too high and had to be reduced to allow injection using a fire pump – at this moment
the only available means to maintain the reactor vessel water inventory and to prevent
uncovering of the reactor core. Depressurization of the reactor system was achieved by
releasing steam through the relief valve that was open manually by the personnel. This
action was difficult to achieve due to a high radiation level and the lack of lighting in the
plant compartments.
Opening of vent line required electric power to energize the valve solenoid for the large
air-operated suppression chamber vent valve that was done using a small portable
generator. Completion of this work required also replacing the temporary air bottle for the
AOV vent. These actions took about 4 hrs and contributed to the delay in providing water
injection to the reactor system.
Severe core damage occurred because there was no cooling of the core for a long time.
TEPCO estimates that following the loss of high pressure coolant injection (approximately
36 hours after reactor trip) there was no injection into the reactor for 6 hours and 43
minutes. This led to severe overheating and partial melt of the fuel. Residual heat was at a
relatively high level as the Unit 3 was under operation at the onset of the event.
How to Improve Safety in Regulated Industries
What Could We Learn From Each Other
Background Material, ENCO FR-(12)-44
© ENCO
Page
25
Step 2. Cause Map - Page 2
Difficult working
environment
Relief Valve
stack open?
Hydrogen
penetrated
to CT
Hydrogen
explosion
AND
AND
Leakage
through
penetrations
and seals
Extended
duration of high
temperature and
pressure in CT
H generated
due to
chemical
reaction
Consequence
of normal
operation
Residual heat
Release of
steam
from RV
Connection of
portable power
generator
Replacement
of depleted AOV
air bottle
Reactor core
severly
overheated
Severe core
damage
CT pressure
too high
CT venting
delayed
RP too high for
backup cooling
pumps
AND
AND
Unit was
operating at
power
AND
Loss of reactor
cooling
Loss of CT
cooling
Significant
efforts to cool
the reactor
Hydrogen explosion in the reactor building was caused by the formation of explosive
mixture of hydrogen and air in the reactor building. The lack of core cooling to
compensate for decay heat resulted in excessive fuel temperatures and oxidation of the
zirconium cladding. The oxidation of zirconium in a steam environment creates significant
additional heat from the exothermic reaction and large quantities of hydrogen. This
hydrogen contributed to the increases in containment pressure and to the subsequent
hydrogen explosion. The extended duration of high temperature and pressure conditions
inside containment may have damaged the drywell head seals, leading to hydrogen leaks
and the subsequent explosions.
Venting of containment was delayed because of difficulties with providing power and
compressed air for opening of vent valves (MOV and AOV). In addition, all work had to be
conducted in a difficult working environment. The torus room was very hot because of the
previous use of RCIC, HPCI, and SRVs and the room was completely dark. Increased
radiation level also contributed to these difficulties.
Loss of reactor cooling was caused by the loss of high pressure emergency core cooling
systems HPCI and RCIC. RCIC shut down unexpectedly and could not be restarted. HPCI
tripped and could not be restarted due to depleted batteries. Other potential high
How to Improve Safety in Regulated Industries
What Could We Learn From Each Other
Background Material, ENCO FR-(12)-44
© ENCO
Page
26
pressure cooling systems (SLC, CRD) as well as low pressure systems (CS, RHR, RHRS) were
non-functional due to loss of AC power as well as the loss of heat sink sources.
Step 2. Cause Map - Page 3
Damage to
seawater intake
structures
Impact of
debris and
wawes
Alternative
cooling systems
unavailable
On-site power
unavailable
AND
HPCI tripped
Loss of reactor
cooling
HPCI could not
be restarted
Depleted
batteries
AND
RCIC tripped
RCIC could not
be restarted
AND
Cooling using
portable (fire)
pumps delayed
Lack of plant
controls &
indications
On-site power
unavailable
AND
No lighting
at the plant
locations
AND
Non-routine
connections
difficult
Connections
made in difficult
environment
AND
Routes for
transport of
portable units
blocked
Impact of
debris and
wawes
AND
Need to
depressurize
reactor and CT
Pressure head
of the portable
pump too low
There were also problems with using portable fire engines. Out of the three fire engines
Fukushima Daiichi had available, one was damaged by the tsunami and a second could not
How to Improve Safety in Regulated Industries
What Could We Learn From Each Other
Background Material, ENCO FR-(12)-44
© ENCO
Page
27
reach units 1-4 because of earthquake damage to the road. Only one fire engine was
immediately available to support the emergency response on units 1-4.
Use of this fire engine required non-routine connections that had to be established
in a difficult environment (increased temperature, radiation and in darkness).
Additional problem, caused by the loss of on-site power, was the lack of indications
and controls of valves involved in the implementation of the required lineup.
Step 2. Cause Map - Page 4
Off-site
power supply
unavailable
Damage of lines
and off-site
substation
Tsunami of the
hight ~15 m
AND
On-site AC
distribution
degraded
On-site power
unavailable
Earthquake
AND
Submergence
of critical
SSCs
AND
DB tsunami
underpredicted
On-site DC
distribution
degraded
AND
Emergency
DGs lost
The fire engine did not have sufficient discharge pressure to overcome the elevation
differences and reactor pressure. Personnel actions associated with depressurization of
reactor and venting of the containment, necessary for reducing reactor pressure to the
level acceptable for the use of fire engine, were not accomplished in time and failed to
prevent core damage.
How to Improve Safety in Regulated Industries
What Could We Learn From Each Other
Background Material, ENCO FR-(12)-44
© ENCO
Page
28
On-site power was lost because of the loss of off-site power and the loss of emergency
diesel generators (EDGs). Off-site power was lost because of earthquake which damaged
breakers and distribution towers. Following the earthquake, the on-site power relayed on
EDGs which started on loss of offsite power. This source was lost because of tsunami.
Tsunami resulted in flooding of both the EDGs and the on-site electrical distribution system
(switchgear rooms). Submergence of critical SSCs in the power supply system resulted in
the complete loss of AC power and led to a partial loss of DC power. All DC power was lost
on units 1 and 2, while some DC power from batteries remained available on Unit 3.
Step 2. Cause Map - Page 5
Limited
historical
data?
DB tsunami
underpredicted
Multiple seismic
faults not
considered
AND
Limited
regulatory
guidance?
Regulatory
framework in
Japan
AND
O-V-R
interfaces and
communication
Organization of
regulatory
system in Japan
Extensive damage of the electrical power supply system at the site was caused by the
tsunami impacting the site that exceeded the design basis of the plant. The maximum
tsunami height was estimated to be 14 to 15 meters as compared to the design basis
tsunami height of 5.7 meters. This was above the site grade levels of 10 meters at units 14. The seawater intake structure was also severely damaged and was rendered
nonfunctional.
The tsunami protection strategy for the plant consisted of locating critical equipment,
such as vital seawater pump motors, above the elevation of the assessed tsunami height.
The assessment of the tsunami height was based on the 2002 Tsunami Assessment Method
for Nuclear Power Plants in Japan, the accepted methodology by the Japanese industry.
Following the guidance of this methodology, offshore fault segments were not combined in
the tsunami assessment. During the earthquake event numerous fault segments acted in
combination, and thus the actual tsunami caused by the earthquake significantly exceeded
the tsunami assessment for the plant.
Fundamentally, following the accepted tsunami assessment technical guidance in Japan
resulted in under-prediction of the size of the tsunami. As a result, the plant tsunami
protection strategy was not adequate and beyond-design-basis tsunami protection
adequate to mitigate the effects of the tsunami that occurred was not available.
The issue of specifying appropriate tsunami design basis is not straightforward. This issue
should be considered in the light of existing historical data consistently with the risk
How to Improve Safety in Regulated Industries
What Could We Learn From Each Other
Background Material, ENCO FR-(12)-44
© ENCO
Page
29
considerations. It seems that the regulatory framework that could be used by the nuclear
industry in Japan for adequate protection of NPPs against tsunami was lacking clarity.
Appropriate regulatory strategy and framework for protection of NPPs against natural
hazards, which would appropriately balance defense-in-depth and risk considerations, was
lacking. One of the reasons may be the structure of regulatory authority that is composed
of several organizations. This issue requires further analysis.
Step 3. Analysis of solutions
The Cause Map is used to identify all the possible solutions for the problem so that the best
solutions can be selected. Potential solutions correspond to those causes which can be
controlled by the problem owner (Operator, Vendor, TSO organization, Regulator) so that
the problem is prevented from recurring.
The following causes, which can be subject of interest in this context, can be identified on
the Cause Map for the March 11 Fukushima tsunami accident (as developed in Step 2):
Design Basis tsunami under-predicted;
Submergence of critical SSCs due to flooding;
Limited regulatory guidance on the seismic and flooding protection of structures,
systems, and components for operating plants;
4. Regulatory framework for adequate protection;
5. Organization of regulatory system in Japan;
6. Depleted batteries;
7. Cooling using portable pumps delayed;
8. Containment venting delayed;
9. Limited number of portable generators available at the site;
10. Non-routine connections of portable cooling pumps difficult and not realized in
time.
1.
2.
3.
These causes are related to various elements of the defense-in-depth protection of safety
of nuclear power plants. The potential solutions are briefly discussed below.
Clarifying the Regulatory Framework

A logical, systematic, and coherent regulatory framework for adequate protection
against external events that appropriately balances defence-in-depth and risk
considerations should be established.
Such framework should clearly specify the requirements that allow the industry to
determine the protections covered within the design basis and those to be considered as
beyond-design-basis (i.e. part of the emergency preparedness plan). In particular,
appropriate regulatory endorsed guidance should be available to specify the design basis
tsunami. Such guidance should provide a clear basis for answering the question "What
should be the design basis tsunami given the existing historical data and plant specific
seismic information?"
How to Improve Safety in Regulated Industries
What Could We Learn From Each Other
Background Material, ENCO FR-(12)-44
© ENCO
Page
30
Ensuring Protection

Given the design basis tsunami the licensees should ensure that the critical SSCs are
adequately protected against seismically induced flooding.
Licensees need to re-evaluate and upgrade as necessary the design-basis seismic and
flooding protection of structures, systems, and components for each operating plant. The
regulator should enforce appropriate corrective actions and ensure adequate oversight of
their implementation.
Enhancing Mitigation
The licensees should strengthen station blackout mitigation capability at operating and
new plant for design-basis and beyond-design-basis accidents induced by external events,
including:

Increasing the capability of batteries;

Ensuring availability of portable generators and enhancement of their use during
prolonged station blackout conditions;

Enhancement of methods to reduce reactor pressure and feed cooling water to the
reactor using portable cooling means /pumps;

Ensuring additional sources of coolant water for the reactor;

Enhancement of methods for non-routine connections of portable cooling means and
ensure plant features to facilitate their realization during prolonged station blackout
conditions;

Enhancement of the containment venting system so that it is independent of AC power
and operates with limited operator actions from the control room.
6.3. Summary conclusions
Some of these issues mentioned above can be of general interest to different industries to
be discussed during the workshop.
One of such issues is the concept of protection based on the combination of appropriately
balanced defence-in-depth and the risk considerations. General problem of broader
interest is the treatment of accident scenarios with low probability and high
consequences.
Based on Fukushima accident it seems that accident scenario initiated by a tsunami of this
severity level was underestimated with regard to its frequency and potential
consequences, and further levels of protection against severe consequences shown to be
ineffective.
The lower and higher mean values of the Bayesian analyses show that accident scenarios
initiated by a tsunami > 8m and an earthquake > Shindo 6 may be equal to, or greater
than, regulatory limits for CDF and LER, especially when some support and backup systems
are guaranteed to fail after such events [A-11].
How to Improve Safety in Regulated Industries
What Could We Learn From Each Other
Background Material, ENCO FR-(12)-44
© ENCO
Page
31
Important leasson learned from Fucushima accident is that defense in depth must move
from strong to stronger. Important role in achieving this goal is through the use of risk
analysis (PSA). It is worth noting that PSA has difficulties with rare events (very large
models, calculation cutoffs, screening-out), PSA models must present uncertainty to
decision makers, PSA must be used as a “living tool” not only for showing regulators that
safety goals have been attained, PSA professionals must be willing to ask and to begin to
answer the difficult questions to themselves, the regulators, and the public [A-11].
Another important issue of general interest is related to the role of individual actors of the
"Safety Net" (Operator, Vendor, TSOs and Regulator) in protecting the plants against severe
hazard events of low likelihood and high consequences. Coordination of efforts and
communication between the actors in this context is one of the important aspects to be
discussed.
How to Improve Safety in Regulated Industries
What Could We Learn From Each Other
Background Material, ENCO FR-(12)-44
© ENCO
Page
32
7. References for Anex
[A-1]
"Fukushima Daiichi Accident – Technical Causal Factor Analysis", EPRI Report
#°1024946, Final Report, March 2012.
[A-2]
"Special Report on the Nuclear Accident at the Fukushima Daiichi Nuclear Power
Station", No. INPO 11-005, Revision 0, November 2011.
[A-3]
"Regulatory Guide for Reviewing Seismic Design of Nuclear Power Reactor
Facilities", Nuclear Safety Commission (NSC, Japan), Document No. NSCRG: L-DSI.02, September, 2006.
[A-4]
"Tsunami Assessment Method for Nuclear Power Plants in Japan", The Tsunami
Evaluation Subcommittee, The Nuclear Civil Engineering Committee, Japan
Society of Civil Engineers (JSCE), 2002 and 2006.
[A5]
"Report of the Japanese Government to the IAEA Ministerial Conference on
Nuclear Safety - The Accident at TEPCO’s Fukushima Nuclear Power Stations",
Nuclear Emergency Response Headquarters, Government of Japan, June, 2011.
[A-6]
"International Conference on Advances in Nuclear Power Plants- Fukushima
Accident: An Overview", Akira Omoto, University of Tokyo, May 3, 2011.
[A-7]
"Fukushima Nuclear Accident Analysis Report (Interim Report)", Tokyo Electric
Power Company, December 2011.
[A-8]
"Fukushima Analysis 11 03 2011 – In-depth Analysis of the Accident at Fukushima
on 11 March 2011 With Special Consideration of Human and Organisational
Factors", Swiss Federal Nuclear Safety Inspectorate (ENSI).
[A-9]
"Recommendations for Enhancing Reactor Safety in the 21st Century", The NearTerm Task Force Review of Insights from the Fukushima Dai-ichi Accident, U.S.
Nuclear Regulatory Commission, July 12, 2011.
[A-11]
Gano,D.L., "Apollo Root Cause Analysis – A New Way of Thinking", Apolonian
Publications, Yakima, Washington, 2003.
[A-10]
[A-11]
ThinkReliability,
daiichiunit3.pdf
http://www.thinkreliability.com/InstructorBlogs/blog-
Epstein, W., "A PRA Practioner Looks at the Fukushima Daiichi Accident", Visiting
Scholar, Ninokata Lab, presentation at Tokyo Institute of Technology, March 20,
2012.
How to Improve Safety in Regulated Industries
What Could We Learn From Each Other
Background Material, ENCO FR-(12)-44
© ENCO
Page
33