Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
WIR SCHAFFEN WISSEN – HEUTE FÜR MORGEN
MELCOR source term modeling for BWR "Fukushima-
like" scenarios with containment venting
Adolf Rýdl, Leticia Fernandez, Terttaliisa Lind :: Paul Scherrer Institut
EMUG meeting, Imperial College London, April 2016
Outline
Page 2
• BWR "Fukushima-like" scenarios with containment venting
• plant model and initial stage of sequence progression
• sequences with different times of ADS and containment venting
• sequences with changing times of the second period of venting
• different approaches to Cs and I modeling
• conclusions and outlook
Page 3
BWR "Fukushima-like" scenarios with containment venting
• SBO in generic Mark-I BWR• Prolonged operation of relevant core cooling systems =>
"Fukushima-like" (FU2, FU3)
• Role of some mitigative safety systems can be even more important here than for a fast progressing accident
-example: CONTAINMENT VENTING (containment protection)
• Current studies look at the impact of various containment venting strategies, different timings as well as link to RPV depressurization
• Whole-plant integral calculations with MELCOR_2.1, including source-term analyses (emphasis on Cs) for both
− hardened, non-filtered system− and hypothesized filtered venting (with DF =1000 for aerosols)
Page 4
Plant model and "Fukushima-like" sequence
• BWR Mark-I plant model based on Peach Bottom NPP deck (Sandia SOARCA), with
enormous amount of modifications through
some years allowing to simulate Fukushima
accidents
• sequence based on extensive PSI analyses of FU3 accident for the OECD-BSAF project (*)
(*) L.Fernandez-Moguel, J.Birchley, Annals of Nuclear Energy, 83 (2015) 193–215
������� �� ���
����������������
��� �������� ���
�� ��
WETWELL (TORUS)
VENTING
RPV
PCV
TSUNAMI
AWI
Page 5
Containment venting studies and cases considered
• containment venting closely linked to RPV depressurization (for FU3, ADS at ~42h)• roughly at 40h into the accident, things starting to "go definitively wrong" :
− RPV water level below BAF (Bottom of Active Fuel)− first hydrogen− beginning of FP (Fission Product) release
• for containment venting studies (venting from WW), intention was to− shift ADS by 1hr steps (42h, 43h, 44h, …), followed always –after 25min- by venting
or
− activate venting on reaching 5atm in the PCV• and analyze the impact on TH (and possible accident recovery) and on FP release to
environment, mostly for noble gases (NG), Cs, and I
• simpler (boundary) conditions than for FU3 used to see the impact clearly: fixed venting period time, AWI always at full delivery when RPV pressure allows, no DW head flange leak, …
• first set of calculations− scenario c01: ADS at 42h, followed by venting− scenario c02: ADS at 43h, followed by venting− scenario c03: reached 5atm in containment at 43h 35min, WW venting initiated without
prior depressurization, ADS shifted to the latest time where recovery still possible (~48h)
• for every case, always 2 full-length sequences calculated, both for unfiltered venting (DF=1 for aerosols) and filtered (DF=1000)
Page 6
Containment pressure and reactor waterlevel for different times of ADS andcontainment venting
-8
-6
-4
-2
0
2
4
6
8
10
10 20 30 40 50
DC
wat
er le
vel
time (hours)
TAF
BAF
AD
S
VE
NT
c01c02c03
0
1
2
3
4
5
6
7
8
9
0 10 20 30 40 50
cont
ainm
ent p
ress
ure
(at
m)
time (hours)
AD
S
VE
NT
c01c02c03
Page 7
RPV pressure and hydrogen generation in-vessel
0
100
200
300
400
500
600
700
800
900
1000
1100
35 40 45 50 55 60
H2
gene
ratio
n in
cor
e (
kg)
time (hours)
ADS
VE
NT
c03c02c01
0
1
2
3
4
5
6
7
8
9
10 20 30 40 50 60
RP
V p
ressu
re (M
Pa
)
time (hours)
AD
S
c01c02c03
Page 8
in-vessel accident progression with the base-case scenario
video for scenario c03, ca. 20h - 51h
Page 9
Source Term: Cs_class radionuclides behavior
10-5
10-4
10-3
10-2
10-1
100
101
102
103
35 40 45 50
Cs_
clas
s re
leas
es (
kg)
time (hours)
Cs release_from_fuel
Cs release_to_ENV
Cs release_to_ENV
(DFaerosols = 1000)
c03c02c01
~4% of the
initial Cs
inventory
major assumption here: Cs behaving mostly like CsOH
Page 10
noble gases and iodine (CsI)
10-3
10-2
10-1
100
101
102
103
35 40 45 50
NG
_cla
ss r
ele
ase
s
(kg
)
time (hours)
NG release_from_fuel
NG release_to_ENV
c03c02c01
10-5
10-4
10-3
10-2
10-1
100
101
102
103
35 40 45 50
CsI
_cl
ass
re
lea
ses
(kg
)
time (hours)
CsI release_from_fuel
CsI release_to_ENV
CsI release_to_ENV
(DFaerosols = 1000)
c03c02c01
• calculated aerosol (CsI) scrubbing efficiency in WW (SPARC code default) very low ?!− code calculates boiling in WW in the decisive periods of time –can be one of the reasons,
but the whole thing rather unclear: MELCOR thermohydraulic predictions for WW very
uncertain (distinct stratification there ?), scrubbing itself uncertain, particle sizes and
distribution are rough estimates, etc, etc -needs more work
− for CsOH (also on aerosols) this has been masked with its (irreversible) trapping on steel surfaces inside RPV and in the steam lines -will be discussed with the Cs behavior models
Page 11
differences in progression of very similarscenarios –MELCOR related
0
1
2
3
4
5
6
7
8
9
0 10 20 30 40 50 60
cont
ainm
ent p
ress
ure
(at
m)
time (hours)V
EN
T
c03 (DF=1) c03 (DF=1000)
c03 scenario without the second period of venting
–base line calculations for further restarts
Page 12
Scenarios with changing the second period of venting
• what if venting right after ADS is unsuccessful (attempted venting in Fukushima accidents not always success)
• second set of scenarios− scenario c03_00
− original c03 (with second period of vent, 25min after ADS at 43h) − scenario c03_01
− second period of vent shifted to the time of significant containment pressurization
− scenario c03_02− without second period of vent => accident recovery unlikely (!)
• again, all of them calculated for both filtered and unfiltered venting
• no real difference found for aerosol-borne FP release before anticipated Vessel Failure at c03_02 scenario, only for NG
Page 13
containment pressure and RPV pressure
1
2
3
4
5
6
7
8
9
25 30 35 40 45 50 55 0
1
2
3
4
5
6
7
8
cont
ainm
ent p
ress
ure
(at
m)
RP
V p
ress
ure
(M
Pa)
time (hours)
HPCI
AD
S
VE
NT
c03_00 (cont_p)c03_01 (cont_p)c03_02 (cont_p)c03_02 (RPV_p)
Page 14
RPV water levels and FP release
-8
-6
-4
-2
0
2
4
6
8
10
25 30 35 40 45 50 55 60
DC
wa
ter
leve
l
time (hours)
TAF
BAF
AD
S
VE
NT
c03_00c03_01c03_02
0
100
200
300
400
35 40 45 50 55 60N
G_
cla
ss r
ele
ase
s
(kg
)
time (hours)
NG release_from_fuel
NG release_to_ENV
c03_00c03_01c03_02
Page 15
Different approaches to Cs modeling
• important aspect of FP modeling: uncertainties with Cs behavior/modeling (dating back to PHEBUS FP test interpretation)
• one approach: Cs behaving mostly as CsOH (equivalent part of Cs also in CsI) - our "c03" sequence
− CsOH expected to be adsorbed on stainless steel. But how much?− fast, irreversible chemisorption of CsOH being default treatment in M2.1 (partially based
on AEA VICTORIA code modeling) versus almost no sorption as an original approach to
Cs modeling, also in M1.8.6 - impact is huge, deeper insight needed
(in the model, mass transfer through boundary layer assumed to be much faster than the
chemisorption itself)
, = 0.139 and = 5.96x107 (R = 8314 J/kg.K)
• Cs2MoO4 as the primary Cs species: rather new approach, at least from the modeling point of view, still not firmly established
• the same questions (what is the Cs speciation, how the sorption would look like, …) asked right now with the analyses of Fukushima accidents
Page 16
Impact of different approaches to Cs modeling
0
50
100
150
200
35 40 45 50Cs b
ehavio
r (C
sO
H n
o_
chem
isorp
tion)
[kg]
time (hours)
from_fuel
RPV
RPV-up
core
S/C
STM-LIN
0
50
100
150
200
35 40 45 50Cs b
eh
avio
r (C
sO
H +
ch
em
iso
rptio
n)
[kg
]
time (hours)
from_fuel
RPV
RPV-up
core
S/C
STM-LIN
• Cs mostly in CsOH, default M2.1 treatment of CsOHsorption - our "c03" sequence
• Cs mostly in CsOH, sorption negligible (default in M1.8.6)
• yet another way, Cs2MoO4 as the primary Cs species (small fraction still remains in CsOH)
–this simulation not analyzed in detail so far
• CsOH cases versus Cs2MoO4 cases would also use different models for release from fuel (CORSOR
versus Booth model) 0
50
100
150
200
35 40 45 50
Cs (
Cs2
Mo
O4
) d
istr
ibu
tio
n
[kg
]
time (hours)
from_fuel
RPV
RPV-up
core
S/C
STM-LIN
Conclusions and outlook
Page 17
• BWR "Fukushima-like" scenarios with containment venting and FP behavior − whole-plant calculations performed for different strategies/timing of
containment venting (filtered, non-filtered), linked also to RPV depressurization
− if fire pumps injection reaches RPV at its full capacity –and DW/WW pressure is sufficiently low after successful venting(!)- recovery at this late stage of a
Fukushima-like accident is possible
− uncertainties as for vessel failure/penetration failure still rather high− early ADS followed by venting helps to keep FP releases reasonable− venting based on high DW/WW pressure led to significat source term for
volatile FPs in our simulations
− filtered venting very effective for aerosol-borne activity in relevant cases− uncertainties related to Cs speciation and behavior at an accident also high, can
have a huge impact on source term prediction; crucial also for Fukushima
modeling
− CsOH versus Cs2MoO4 as a principal Cs species, chemisorption of CsOH, …
Conclusions and outlook (contd)
Page 18
• in general− presented analyses might be rather difficult to perform without an integral
code => MELCOR can be useful in some instances : )
− in whole-plant analyses, MELCOR is still very sensitive "beast", as shown clearly with our BSAF FU3 calculations -e.g., tiny differences in the amount
of injected water, AWI, would drive the sequence ex-vessel (as compared to
recovered one which was our best estimate case)
− one needs to exercise caution
• outlook− looking in more detail into WW retention/scrubbing of FPs –we've started− thermohydraulic behavior of WW (S/C), with particular regard to impact on
aerosol retention
− issues with Cs behavior and its modeling, linked also to iodine− all in close connection with on-going Fukushima analyses
Page 19
Wir schaffen Wissen – heute für morgen
thank you
for your atttention
Page 20
containment pressure and RPV pressure
Reactor Core Isolation Cooling (RCIC)
CST
RPV
Reactor Pressure Vessel
Test
lin
e
Injection line
Maintain the water level
WW
SRV
Safety
Relief
Valves
Maintain the pressure
Condensate Storage Tank
The extracted steam is discharged to the wetwell(WW)
WW pressure increase when the pool is saturated
Steam is extracted
operators recirculated water to the CST to avoid the RCIC to stop (due to high water level in RPV)
RCIC Turbine
Page 22
in-vessel accident progression