ASTEC V2 severe accident integral code main features, current V2.0 modelling status, perspectives

Abstract

The severe accident integral code ASTEC, jointly developed since almost 20 years by IRSN and GRS, simulates the behaviour of a whole nuclear power plant under severe accident conditions, including severe accident management by engineering systems and procedures. Since 2004, the ASTEC code is progressively becoming the reference European severe accident integral code through in particular the intensification of research activities carried out in the frame of the SARNET European network of excellence. The first version of the new series ASTEC V2 was released in 2009 to about 30 organizations worldwide and in particular to SARNET partners. With respect to the previous V1 series, this new V2 series includes advanced core degradation models (issued from the ICARE2 IRSN mechanistic code) and necessary extensions to be applicable to Gen. III reactor designs, notably a description of the core catcher component to simulate severe accidents transients applied to the EPR reactor. Besides these two key-evolutions, most of the other physical modules have also been improved and ASTEC V2 is now coupled to the SUNSET statistical tool to make easier the uncertainty and sensitivity analyses. The ASTEC models are today at the state of the art (in particular fission product models with respect to source term evaluation), except for quenching of a severely damage core. Beyond the need to develop an adequate model for the reflooding of a degraded core, the main other mean-term objectives are to further progress on the on-going extension of the scope of application to BWR and CANDU reactors, to spent fuel pool accidents as well as to accidents in both the ITER Fusion facility and Gen. IV reactors (in priority on sodium-cooled fast reactors) while making ASTEC evolving towards a severe accident simulator constitutes the main long-term objective. This paper presents the status of the ASTEC V2 versions, focussing on the description of V2.0 models for water-cooled nuclear plants.

Full text

Nuclear
Engineering
and
Design
272
(2014)
119–135
Contents
lists
available
at
ScienceDirect
Nuclear
Engineering
and
Design
jou
rn
al
hom
ep
age:
www.elsevier.com/locate/nucengdes
ASTEC
V2
severe
accident
integral
code
main
features,
current
V2.0
modelling
status,
perspectives
P.
Chatelarda,∗,
N.
Reinkeb,
S.
Arndtb,
S.
Belona,
L.
Cantrela,
L.
Careninia,
K.
Chevalier-Jabeta,
F.
Cousina,
J.
Eckelb,
F.
Jacqa,
C.
Marchettoa,
C.
Muna,
L.
Piara
aInstitut
de
Radioprotection
et
de
Sûreté
Nucléaire
(IRSN),
PSN-RES,
B.250,
Cadarache
BP3
13115,
Saint-Paul-lez-Durance,
Cedex,
France
bGesellschaft
für
Anlagen-
und
Reaktorsicherheit
(GRS)
mbH,
Schwertnergasse
1,
50677
Köln,
Germany
a
r
t
i
c
l
e
i
n
f
o
Article
history:
Received
4
April
2013
Received
in
revised
form
12
June
2013
Accepted
12
June
2013
a
b
s
t
r
a
c
t
The
severe
accident
integral
code
ASTEC,
jointly
developed
since
almost
20
years
by
IRSN
and
GRS,
simulates
the
behaviour
of
a
whole
nuclear
power
plant
under
severe
accident
conditions,
including
severe
accident
management
by
engineering
systems
and
procedures.
Since
2004,
the
ASTEC
code
is
progressively
becoming
the
reference
European
severe
accident
integral
code
through
in
particular
the
intensification
of
research
activities
carried
out
in
the
frame
of
the
SARNET
European
network
of
excellence.
The
first
version
of
the
new
series
ASTEC
V2
was
released
in
2009
to
about
30
organizations
worldwide
and
in
particular
to
SARNET
partners.
With
respect
to
the
previous
V1
series,
this
new
V2
series
includes
advanced
core
degradation
models
(issued
from
the
ICARE2
IRSN
mechanistic
code)
and
necessary
exten-
sions
to
be
applicable
to
Gen.
III
reactor
designs,
notably
a
description
of
the
core
catcher
component
to
simulate
severe
accidents
transients
applied
to
the
EPR
reactor.
Besides
these
two
key-evolutions,
most
of
the
other
physical
modules
have
also
been
improved
and
ASTEC
V2
is
now
coupled
to
the
SUNSET
statistical
tool
to
make
easier
the
uncertainty
and
sensitivity
analyses.
The
ASTEC
models
are
today
at
the
state
of
the
art
(in
particular
fission
product
models
with
respect
to
source
term
evaluation),
except
for
quenching
of
a
severely
damage
core.
Beyond
the
need
to
develop
an
adequate
model
for
the
reflooding
of
a
degraded
core,
the
main
other
mean-term
objectives
are
to
further
progress
on
the
on-going
extension
of
the
scope
of
application
to
BWR
and
CANDU
reactors,
to
spent
fuel
pool
accidents
as
well
as
to
accidents
in
both
the
ITER
Fusion
facility
and
Gen.
IV
reactors
(in
priority
on
sodium-cooled
fast
reactors)
while
making
ASTEC
evolving
towards
a
severe
accident
simulator
constitutes
the
main
long-term
objective.
This
paper
presents
the
status
of
the
ASTEC
V2
versions,
focussing
on
the
description
of
V2.0
models
for
water-cooled
nuclear
plants.
©
2013
Elsevier
B.V.
All
rights
reserved.
1.
Introduction
The
ASTEC
code
(Accident
Source
Term
Evaluation
Code),
jointly
developed
since
several
years
by
the
French
Institut
de
Radiopro-
tection
et
de
Sûreté
Nucléaire
(IRSN)
and
the
German
Gesellschaft
für
Anlagen
und
Reaktorsicherheit
mbH
(GRS),
aims
at
simulating
an
entire
severe
accident
(SA)
sequence
in
a
nuclear
water-cooled
reactor
from
the
initiating
event
through
the
release
of
radioactive
elements
out
of
the
containment.
The
main
ASTEC
applications
are
therefore
source
term
determination
studies,
level
2
Probabilistic
safety
assessment
(PSA2)
studies
including
the
determination
of
uncertainties,
accident
management
studies
and
physical
analyses
∗Corresponding
author.
E-mail
address:
patrick.chatelard@irsn.fr
(P.
Chatelard).
of
experiments
to
improve
the
understanding
of
the
phenomeno-
logy.
The
first
series
of
V0
versions
was
developed
between
1996
and
2002
while
from
2003
the
V1
series
extended
the
scope
of
ASTEC
applications
to
the
front-end
thermal-hydraulic
phases
(Van
Dorsselaere
et
al.,
2009a).
Since
2004,
ASTEC
is
progressively
becoming
the
reference
European
severe
accident
integral
code
for
water-cooled
reactors
through
the
intensification
of
the
jointly
research
activities
mainly
carried
out
in
the
frame
of
the
SARNET
European
Network
of
Excel-
lence
in
the
6th
and
7th
Research
Framework
Programmes
(FP6
and
7)
of
the
European
Commission
(Micaelli
et
al.,
2005;
Van
Dorsselaere
et
al.,
2010b).
Indeed,
the
ASTEC
code
is
a
key
integrat-
ing
component
in
this
network
since
one
of
the
ultimate
goals
of
the
joint
research
activities
is
to
provide
validated
physical
models
to
be
implemented
in
ASTEC.
So,
most
ASTEC
models
are
today
close
0029-5493/$
–
see
front
matter
©
2013
Elsevier
B.V.
All
rights
reserved.
http://dx.doi.org/10.1016/j.nucengdes.2013.06.040

120
P.
Chatelard
et
al.
/
Nuclear
Engineering
and
Design
272
(2014)
119–135
Nomenclature
AFWSG
Auxiliary
Feedwater
of
Steam
Generator
BWR
Boiling
Water
Reactors
CANDU
CANada
Deuterium
Uranium
(pressurized
heavy
water
reactor)
CESAM
Code
for
European
Severe
Accidents
Management
CIT
Corium
Interactions
and
Thermochemistry
DCH
Direct
Containment
Heating
EPR
European
Pressurized
Reactor
ERMSAR
European
Review
Meetings
on
Severe
Accident
Research
FP5
5th
European
Commission
Research
Framework
Programme
FP6
6th
European
Commission
Research
Framework
Programme
FP7
7th
European
Commission
Research
Framework
Programme
HTR
High
Temperature
Reactors
ICHEMM
Iodine
CHEmistry
and
Mitigation
Mechanisms
ISP
International
Standard
Problems
of
OECD
ISTP
International
Source
Term
Programme
IVR
In-Vessel
melt
Retention
JASMIN
Joint
Advanced
Severe
accidents
Modelling
and
Integration
for
Na-cooled
fast
neutron
reactors
MCCI
Molten-Core–Concrete
Interaction
MOX
Mixed-Oxide
Fuel
NPP
Nuclear
Power
Plants
OECD
Organization
for
Economic
and
Cooperation
Devel-
opment
PAR
Passive
Autocatalytic
Recombiners
PASSAM
PASsive
Severe
Accidents
Management
PBMR
Pebble
Bed
Modular
Reactor
PHWR
Pressurized
Heavy
Water
Reactors
PSA
Probabilistic
Safety
Assessment
PTR
French
acronym
to
identify
the
Refuelling
Water
Storage
Tank
(RWST)
PWR
Pressurized
Water
Reactor
RCS
Reactor
Cooling
System
SA
Severe
Accident
SAM
Severe
Accident
Management
SFR
Sodium-cooled
Fast
neutron
Reactor
SGTR
Steam
Generator
Tube
Rupture
SARNET
Severe
Accident
Research
NETwork
of
Excellence
VVER
Vodo-Vodyanoi
Energetichesky
Reactor
or
Water–Water
Energetic
Reactor
(Russian
PWRs)
to
the
state
of
the
art.
Besides
such
integration
of
new
or
improved
physical
models,
efforts
continuously
address
the
assessment
of
the
ASTEC
code
(both
within
SARNET
and
through
bilateral
agreements
signed
with
IRSN
or
GRS)
through
comparison
either
with
results
of
the
most
important
international
experiments
or
with
results
of
other
severe
accident
simulation
codes.
The
first
version
of
the
new
series
ASTEC
V2
was
released
in
July
2009
to
about
30
organizations
worldwide
and
in
particular
to
SARNET
FP7
partners
involved
in
the
work-package
on
ASTEC
topics.
This
new
series
of
version
provided
users
with
significant
modelling
improvements
with
respect
to
the
former
V1
series.
In
particular,
the
V2.0
version
includes
advanced
core
degradation
models
(issued
from
the
IRSN
ICARE2
mechanistic
code)
and
the
ASTEC
applicability
has
been
extended
to
Gen.
III
reactor
designs,
so
that
the
V2.0
allows
to
simulate
the
EPR
reactor
or
to
launch
investi-
gations
about
the
mitigation
strategy
relying
on
the
In-Vessel
melt
Retention
concept
(IVR).
Besides
these
two
key
evolutions,
other
physical
models
have
also
been
improved
(e.g.
first
account
for
gas
chemistry
kinetics
in
the
Reactor
Cooling
System
(RCS)
or
new
mod-
elling
of
hydrogen
combustion
in
the
containment)
and
ASTEC
V2
is
now
coupled
to
the
SUNSET
IRSN
statistical
tool
to
make
easier
the
uncertainty
and
sensitivity
analyses
(Chevalier-Jabet
et
al.,
2014).
Since
2009,
two
subsequent
revisions
of
the
V2.0
version
have
been
successively
delivered
to
users,
respectively
in
2010
and
2011,
and
a
third
one
is
planned
to
be
frozen
in
the
early
beginning
of
2013
while
the
next
V2.1
major
version
is
to
be
issued
in
2014.
Similarly
to
what
was
done
few
years
ago
in
SARNET
FP6
for
the
former
series
ASTEC
V1
(Van
Dorsselaere
et
al.,
2010a),
the
ASTEC
V2.0
version
and
its
subsequent
revisions
have
been
also
largely
assessed
by
both
IRSN
or
GRS
developers
and
also
by
several
other
partners
in
the
peculiar
frame
of
the
SARNET
FP7
project
(Chatelard
et
al.,
2012).
Both
code-to-experiments
and
code-to-code
assess-
ment
tasks
being
a
continuous
process,
they
are
being
continued
using
now
the
latest
V2.0
revision.
This
paper
presents
the
series
of
ASTEC
V2
versions,
focussing
on
V2.0.
The
main
part
is
devoted
to
a
description
of
the
V2.0
physical
modelling,
along
with
a
short
summary
of
both
the
ASTEC
V2
vali-
dation
strategy
and
the
V2.0
plant
applications
capabilities,
before
concluding
briefly
with
the
main
perspectives
of
development
in
the
next
years.
2.
ASTEC
V2
code
structure
and
main
programming
features
The
ASTEC
scope
of
application
covers
most
of
the
physical
phenomena
involved
in
SA,
except
steam
explosion
and
mechan-
ical
response
of
the
containment.
For
the
latter
phenomena,
the
code
can
yield
initial
and
boundary
conditions
for
a
specific
analy-
sis
using
detailed
codes
such
as,
respectively,
Computational
Fluid
Dynamics
(CFD)
and
Finite
Element
codes.
For
the
hydrogen
risk
in
containment,
ASTEC
provides
a
global
evaluation
of
the
risk,
never-
theless
such
evaluation
needs
to
be
consolidated
by
CFD
approach.
The
ASTEC
code
structure
is
modular,
each
of
its
modules
sim-
ulating
a
reactor
zone
or
a
set
of
physical
phenomena
(see
Fig.
1).
Two
different
running
modes
are
possible
in
ASTEC
V2.0:
-
Stand-alone
mode
for
running
each
ASTEC
module
indepen-
dently,
which
is
useful
for
module
validation
and
calculation
of
separate-effect
tests;
-
Coupled
mode
where
all
(or
a
subset
of)
the
ASTEC
modules
are
run
sequentially
within
a
macro-time
step.
This
mode
allows
explicit
feedback
between
modules.
A
specific
tool
SIGAL-ODESSA
was
specifically
developed
by
IRSN
in
Fortran
(see
Section
5.2)
for
managing
the
database
associ-
ated
to
any
transient
calculation.
The
ASTEC
modules
communicate
with
each
other
through
a
“dynamic”
memory
(see
Fig.
1)
and
data
are
exchanged
between
the
ASTEC
modules
at
macro-time
steps
through
this
dynamic
database,
i.e.
evolving
throughout
the
calcu-
lation
and
mirroring
at
each
time
the
state
of
the
reactor.
The
ASTEC
V2
programming
language
is
mainly
Fortran
(today
mainly
Fortran
95
with
progressive
evolution
towards
Fortran
2003).
The
code
size
is
about
450,000
lines,
distributed
in
more
than
2000
routines
and
the
ASTEC
V2.0
computer
targets
are
PCs
with
both
Linux
and
Windows
operating
systems.
3.
Description
of
the
ASTEC
V2.0
models
3.1.
RCS
thermal-hydraulics
The
CESAR
module
(Bandini
et
al.,
2010)
simulates
the
thermal-
hydraulics
in
the
primary
circuit,
secondary
circuit
and
in
the

P.
Chatelard
et
al.
/
Nuclear
Engineering
and
Design
272
(2014)
119–135
121
Fig.
1.
Schema
of
the
ASTEC
V2.0
modules,
code
structure
and
running
mode.
reactor
vessel
(with
a
simplified
core
modelling)
up
to
the
begin-
ning
of
the
core
degradation
phase,
i.e.
roughly
up
to
the
start
of
core
uncovery,
and
in
any
case
before
the
start
of
Zr
cladding
oxi-
dation
by
steam.
After
the
onset
of
the
core
degradation
phase,
the
CESAR
module
computes
only
the
thermal-hydraulics
in
primary
and
secondary
circuit
as
well
as
in
the
vessel
upper
plenum.
The
thermal-hydraulics
in
the
reactor
vessel
during
core
degradation
is
covered
by
the
ICARE
module
(see
Section
3.2).
The
CESAR
thermal-hydraulics
modelling
is
based
on
a
1D
2-fluid
5-equation
approach.
Up
to
N
non-condensable
gases
(hydrogen,
helium,
nitrogen,
argon
and
oxygen)
are
available.
As
a
result
5
+
N
differential
equations
and
1
algebraic
equation
are
solved:
-
2
+
N
mass
differential
balance
equations
(one
for
the
vapour
phase,
N
for
the
non-condensable
gases
and
one
for
the
liquid
phase);
-
2
energy
differential
balance
equations
(one
for
the
gas
mixture
phase
and
one
for
the
liquid
phase);
-
1
mixture
(liquid
and
gas
phases)
differential
momentum
balance
equation;
-
1
algebraic
equation,
which
models
the
interfacial
drag
between
the
liquid
phase
and
the
gas
phase.
It
must
be
underlined
here
that
the
interfacial
drag
is
a
complex
model,
which
has
been
assessed
on
a
large
number
of
experimental
data.
Thermal
non-equilibrium
is
considered
between
phases,
with
the
possibility
of
sub-cooled
liquid
and
superheated
steam,
and
mechanical
non-equilibrium
between
phases
is
considered
too,
with
the
possibility
of
counter-current
as
well
as
stratified
flows.
Specific
swollen
water
level
volumes
are
used
to
model
large
two-phase
domains
such
as
pressurizer.
It
can
be
noted
that,
while
only
one
average
value
for
pressure,
gas
temperature
and
wall
tem-
perature
was
considered
for
these
peculiar
components
in
ASTEC
V1,
the
physical
modelling
of
the
swollen
water
level
volumes
has
been
significantly
improved
in
ASTEC
V2
since
pressures
and
tem-
peratures
are
now
properly
evaluated
in
each
sub-volume
through
a
clear-cut
distinction
between
the
values
above
and
below
the
water
level.
Special
components
represent:
hydro-accumulators,
pumps
(which
are
described
through
a
0D
approach),
valves,
and
breaks,
for
the
latter
with
correlations
for
critical
and
sub-critical
flow.
In
addition,
the
pressurizer
spray
and
heater
systems
can
be
considered
specifically.
Most
of
the
CESAR
physical
constitutive
laws
are
issued
from
the
correlations,
which
are
included
in
the
French
best-estimate
thermal-hydraulics
CATHARE2
code
(Camous
et
al.,
2005).
In
par-
ticular,
the
critical
break
flow
rate
is
based
on
the
Gros
D’Aillon
correlation
whereas
the
heat
transfer
coefficient
between
the
struc-
ture
and
the
fluid
is
based
on
a
boiling
(Nukiyama)
curve.
Different
heat
transfer
processes
are
modelled:
forced
convection
to
liquid,
nucleate
boiling,
critical
heat
flux,
transition
boiling,
film
boiling,
forced
convection
to
vapour
and
radiative
heat
transfer.
Moreover
a
droplet
projection
model
is
implemented
which
enables
CESAR
to
simulate
the
reflooding
of
intact
or
slightly
degraded
cores
(i.e.
still
in
rod-like
geometries).
Fig.
2
shows
an
example
of
the
RCS
nodalization
of
a
French
1300
Mwe
PWR.
The
numerical
method
follows
the
finite
volume
technique.
The
space
is
discretized
using
a
staggered
grid
with
the
use
of
the
donor
cell
principle.
The
time
integration
is
performed
using
a
Newton’s
method
and
applying
a
fully
implicit
scheme.
The
Jacobian
matrix
inversion
is
based
on
a
highly
optimized
Lower
Upper
algorithm,
which
makes
CESAR
a
fast
running
and
at
the
same
time
stable
mod-
ule.
Moreover,
a
special
coupling,
based
on
a
prediction–correction
method,
is
applied
in
V2.0
between
the
CESAR
and
ICARE
modules
during
the
core
degradation
phase
(see
Section
4.1).
As
regards
the
CESAR
domain
of
application,
the
extension
to
the
transport
in
the
RCS
of
N
non-condensable
gases
allows
now
to
properly
manage
with
ASTEC
V2
the
situations
of
total
oxy-
gen
starvation
that
may
occur
during
a
reactor
accidental
scenario
as
well
as
to
apply
CESAR
to
circuits
with
any
gas
(either
experi-
ments
or
Gen.
IV
reactors),
including
pure
non-condensable
flows.
In
particular,
some
assessment
was
done
on
HTR-specific
helium
thermal-hydraulics
experiments
(Becquin
and
Maas,
2010)
with
good
results.
Moreover,
while
it
was
identified
as
a
shortcoming
of
former
V1
series,
operation
in
reactor
states
at
low
pressure,
e.g.
shutdown
states
or
mid-loop
states,
can
be
now
handled
with
ASTEC
V2.0.

122
P.
Chatelard
et
al.
/
Nuclear
Engineering
and
Design
272
(2014)
119–135
Fig.
2.
ASTEC
V2
nodalization
(before
core
degradation)
of
the
RCS
of
a
French
1300
MWe
PWR.
3.2.
In-vessel
core
degradation
The
ICARE
module
describes
the
in-vessel
core
degradation
phenomena,
both
early
and
late
degradation
phases.
In
the
core
region,
the
ICARE
module
is
directly
derived
from
the
IRSN
mech-
anistic
code
ICARE2
for
core
degradation
(Chatelard
et
al.,
2006),
except
for
the
fluid
dynamics
which
modelling
remains
similar
to
the
one
from
ASTEC
V1.
In
the
lower
head
region,
while
the
general
modelling
approach
remains
somewhat
similar
to
the
one
from
the
simplified
modelling
which
was
available
in
ASTEC
V1.3
(Van
Dorsselaere
et
al.,
2009a),
several
improvements
of
the
phys-
ical
models,
in
particular
as
concerns
the
corium
distribution
and
transient
evolution,
have
been
integrated
in
the
V2.0
version
with
respect
to
former
V1
series;
these
modelling
evolutions
are
sum-
marized
hereafter
as
a
part
of
the
short
descriptive
list
of
the
main
ICARE
module
features.
The
core
degradation
process
is
characterized
by
the
high
complexity
of
phenomena
to
be
considered
and
geometry
to
be
accurately
presented,
with
a
permanent
appearance
and
disappear-
ance
of
a
large
number
of
components
in
each
control
volume
by
e.g.
melting,
failure,
relocation,
and
chemical
reactions.
This
demands
a
dynamic
management
of
these
components.
Besides,
the
geom-
etry
of
a
degraded
core
is
very
complex
and
heterogeneous:
rod
bundles
with
spacer
grids,
fluid
channels
possibly
blocked
with
molten/frozen
mixtures
of
materials,
corium
molten
pool
with
crusts,
debris
beds,
peripheral
and
lower/upper
core
structures
(e.g.
horizontal
plates,
vertical
surrounding
walls
such
as
barrels
or
shrouds),
also
partly
or
totally
molten.
In
ICARE,
the
core
is
discre-
tised
in
cylindrical
rings
and
axial
meshes,
only
one
representative
component
of
the
fuel
and
control
rods
being
considered
in
each
ring,
weighted
by
the
true
number
of
rods.
So,
as
illustrated
on
Fig.
3,
ICARE
allows
to
simulate
the
early
phase
of
core
degradation
with
fuel
rod
heat-up,
ballooning
and
burst,
exothermic
clad
oxidation,
control
rods
behaviour,
fuel
rod
embrittlement
or
melting,
molten
mixture
candling
and
relocation,
etc.
.
.
and
then
the
late
phase
of
core
degradation
with
corium
accumulation
within
the
core
channels
and
formation
of
blockages,
corium
slump
into
the
lower
head
and
corium
behaviour
in
the
lower
head
until
vessel
failure.
The
main
ICARE
models
in
ASTEC
V2.0
are:
-
Thermal-hydraulics:
simplified
modelling
based
on
0D
liquid
water
components
below
(r–z)
gas
flows,
i.e.
quasi
static
swollen
level
of
water
in
a
multi-channel
configuration
(no
equation
of
motion
for
the
liquid
phase,
simply
assuming
pressure
equilib-
rium
at
the
channel
inlet),
and,
in
the
core
region
above,
2D
gaseous
phase
composed
of
steam
and
non
condensable
gases;
in
addition,
a
special
channel
made
of
one
single
mesh
models
the
lower
head
region;
-
Heat
transfer:
axial
and
radial
conduction
between
two
walls,
gap
exchanges
between
rod
and
clad,
convection
between
fluid
and
wall
as
well
as
radiation.
For
the
latter,
a
general
in-core
heat
transfer
model
(based
on
an
equivalent
radiative
conductivity
approach)
allows
to
deal
with
radiative
exchanges
in
a
reactor
core
whatever
the
degradation
level
is
(intact
rods,
moderately
degraded
rods,
severely
damaged
core,
large
cavities,
.
.
.),
thus
managing
in
a
continuous
way
the
heat
transfers
all
along
the
evolution
of
the
core
geometry
degradation.
In
addition,
radia-
tion
from
the
lower
core
structures
to
the
residual
water
in
lower
plenum
is
also
modelled,
which
favours
vaporization
of
water;
-
Power:
either
nuclear
power
generated
by
fission
products
(FP)
or
generated
in
a
given
material,
or
electric
power
generated
in
some
out-of-pile
experiments;
-
Rod
mechanics:
ballooning,
creep
and
burst
of
Zircaloy
fuel
rod
claddings
(including
both
Zry-4
and
Zr1%Nb
alloys),
creep
of
con-
trol
rod
stainless
steel
claddings,
loss
of
integrity
of
fuel
rods
(using
user-criteria);
-
Chemistry:
oxidation
of
Zr
by
steam
(including
correlations
for
both
Zry-4
and
Zr1%Nb
alloys),
oxidation
of
stainless
steel
by
steam,
dissolution
of
UO2by
solid
and
liquid
Zr,
dissolution
of

P.
Chatelard
et
al.
/
Nuclear
Engineering
and
Design
272
(2014)
119–135
123
Fig.
3.
Illustration
of
core
degradation
simulation
with
ASTEC
(status
just
before
first
corium
slumping
into
the
lower
head
for
a
LOCA
2 sequence).
bis:
Illustration
of
core
degradation
simulation
with
ASTEC
(status
just
before
the
vessel
lower
head
failure
for
a
LOCA
2 sequence).
Zr
by
liquid
Silver–Indium–Cadmium
alloy,
dissolution
of
Zr
by
solid
steel,
oxidation
and
degradation
of
B4C
control
rods,
oxi-
dation/dissolution
of
relocating
and
relocated
U–O–Zr
magmas,
oxidation
of
solid
debris
particles,
.
.
.;
-
Reflooding
of
intact
or
slightly
degraded
cores
(i.e.
still
in
rod-
like
geometry),
based
on
a
special
tracking
of
the
quench
front
evolution;
-
Material
melting
and
relocation
(both
in
early
phase
and
late
phase
of
core
degradation):
formation
of
solid
debris
and/or
solid/liquid
magma,
2D
movement
of
magmas
(axial
candling
of
a
mixture
of
molten
and
solid
masses
as
a
film
along
the
rods
or
radial
spreading
in
case
of
downwards
obstacles
due
to
e.g.
reso-
lidified
material
or
horizontal
plates),
vertical
collapse
of
solid
debris,
formation
and
expansion
of
a
molten
pool,
and
last
but
not
least
corium
slump
into
the
lower
plenum;
-
Corium
jet
fragmentation
on
contact
with
water
located
in
the
vessel
lower
head;
-
Corium
behaviour
in
lower
plenum:
2D
meshing
of
the
ves-
sel
lower
head
wall,
combined
with
a
0D
approach
within
the
plenum
volume
accounting
for
3
stratified
liquid
corium
lay-
ers
(light
metallic
layer,
oxide
pool,
heavy
metallic
layer)
and
2
possible
debris
layers.
The
heat
transfers
between
neighbouring
layers,
between
layers
and
vessel
walls
or
residual
water
are
using
established
correlations
from
literature,
depending
on
layer
mean
temperature
and
power.
Moreover,
a
specific
stratification
model
allows
to
manage
both
thermochemical
and
hydrodynamic
phase
separation
processes
in
the
lower
plenum,
along
with
the
possible
inversion
of
metal/oxide
layers
in
case
the
heaviest
layer
is
above
the
lightest
one
(based
on
the
outcomes
of
the
MASCA
experi-
ments:
cf.
Asmolov
and
Tsurikov,
2004);
-
Vessel
lower
head
rupture:
melt-through
or
mechanical
failure
(either
instantaneous
plastic
rupture
or
creep
rupture)
account-
ing
for
the
corium
and
water
loading
on
the
lower
head
wall
and
also
for
the
possible
vessel
wall
partial
melting
based
on
differ-
ent
approaches
considering
e.g.
different
lower
head
geometries;
user-criteria
such
as
temperature,
degradation
rate,
stress
can
also
be
defined;
-
Slumping
(at
vessel
failure
occurrence)
of
corium
into
the
cav-
ity
for
triggering
possible
DCH
(direct
containment
heating)
and
subsequent
MCCI
(molten
core
concrete
interaction)
phenomena.
The
flexible
description
of
geometry
of
vessel
lower
head
allows
simulating
any
type
of
shape
such
as
hemispherical
one
for
PWRs
and
ellipsoidal
one
for
most
of
VVERs.
Indeed,
two
alternative
mod-
els
are
available
for
the
vessel
lower
head
mechanical
failure:
a
general
one
(LOHEY
model)
valid
for
both
hemispherical
and
ellip-
tical
shapes
and
a
more
sophisticated
one,
valid
(OEUF
model)
only
for
hemispherical
shape
which
peculiarity
is
to
assume
the
final
shape
to
look
like
an
“egg”
shell.
As
concerns
numerics,
the
oxidation
reactions
obey
an
implicit
scheme
in
order
to
properly
manage
the
calculation
of
hydrogen
production
while
reducing
computing
time.
Moreover,
as
already
noticed,
there
is
a
special
coupling
at
the
core
boundaries
between
the
CESAR
and
ICARE
modules
(see
Section
4.1).
3.3.
FP
release
from
the
core
The
ELSA
module
(Brillant
et
al.,
2013a)
aims
at
simulating
the
release
of
fission
products
and
structure
materials
from
the
degrad-
ing
core.
ELSA
is
tightly
coupled
with
the
ICARE
module,
which
treats
the
phenomena
of
the
core
degradation.
The
ELSA
modelling
allows
describing
the
release
from
fuel
rods
and
control
rods,
followed
by
the
release
from
debris
beds
(if
any)
and,
then,
the
release
from
the
in-core
molten
pool
(i.e.
from
a
set
of
ICARE
liquid
magma
components).
The
modelling
is
based
on
a
semi-empirical
approach
and
the
physical
phenomena
taken
into
account
are
the
main
limiting
phenomena,
which
govern
the
release.
For
intact
fuel
rods
and
debris
beds,
the
FP
release
is
described
according
to
the
degree
of
fission
product
volatility.
Three
categories
are
distinguished
with
the
following
characteristics:
(a)
Volatile
species
(such
as
Xe,
I,Cs,
or
Te):
-
Release
is
described
by
species
intra-granular
diffusion
through
UO2fuel
grains,
taking
into
account
fuel
oxidation
(UO2+x)
and
a
grain-size
distribution;
-
Te,
Se
and
Sb
can
be
partially
trapped
in
the
cladding
depend-
ing
on
temperature
and
on
the
degree
of
cladding
oxidation;
-
At
fuel
melting
point,
all
the
remaining
species
located
in
the
liquid
part
of
the
fuel
are
supposed
to
be
instantaneously
released.
(b)
Semi-volatiles
species
(such
as
Ba
or
Mo):
-
Release
is
described
by
evaporation
into
inter-granular
porosities
and
mass
transfer
processes.
(c)
Low
volatiles
species
(such
as
U
or
Pd):
-
Release
is
described
by
fuel
volatilization
treated
as
the
vapor-
ization
of
UO3.
This
process
can
therefore
take
place
only
after
a
severe
degradation
of
the
fuel
rods.
The
difference
between
the
configuration
of
fuel
rod
and
debris
bed
is
the
determination
of
the
average
geometrical
ratio
“sur-
face/volume”
used
in
the
calculation
of
the
stoichiometry
deviation.

124
P.
Chatelard
et
al.
/
Nuclear
Engineering
and
Design
272
(2014)
119–135
vapour
aerosol
vapour
aerosol
chemical
reactions
chemical
reactions
sorptionsorption
condensation
evaporation
condensation
evaporation deposition
resuspension
deposition
resuspension
evapor
ationevapor
ation
inlet flowinlet flow nucleationnucleation
agglo
meratio
nagglo
meratio
n
outlet flowoutlet flow
Fig.
4.
ASTEC
modelling
of
aerosol/FP
behaviour
in
the
RCS.
Concerning
the
molten
pool
configuration,
given
the
high-
temperature
conditions,
the
chemical
equilibrium
can
be
assumed
in
the
magma
so
that
release
is
governed
by
mass
transfer
and
evap-
oration
processes
from
the
free
surface
of
the
molten
pool.
Central
to
the
modelling
is
the
calculation
of
the
vapour
pressures
of
the
elements
in
the
molten
pool.
The
assumption
of
non-ideal
solution
chemistry
is
also
used
for
phase
distribution.
Finally,
for
the
structure
materials,
release
of
Ag,
In,
Cd,
Sn,
Fe,
Ni,
and
Cr
is
taken
into
account
in
ELSA
as
follows:
-
Ag,
In,
and
Cd
(SIC
alloy)
are
released
from
degraded
control
rods.
The
same
approach
as
semi-volatile
species
is
used,
i.e.
release
is
described
by
evaporation
and
mass
transfer
processes.
The
SIC
release
happens
at
the
control
rod
failure.
It
is
followed
by
release
from
free
surface
of
the
control
rod
and
from
the
control
rod
molten
alloys
during
their
candling
along
the
rod
external
surface;
-
Fe,
Ni,
and
Cr
are
supposed
to
be
released
during
the
candling
of
steel
materials,
using
the
same
approach
as
for
the
release
of
Ag,
In
and
Cd;
-
Sn
is
supposed
to
be
released
as
a
proportion
of
the
rate
of
ZrO2formation,
as
lessons
drawn
from
Phébus.FP
observations
(Grégoire
and
Haste,
2012).
These
structure
materials
can
also
be
released
from
the
corium
molten
pool.
For
B4C
control
rods,
release
of
boron
and
carbon
is
not
managed
by
ELSA:
it
is
an
output
from
the
simplified
boron
carbide
oxidation
model
which
belongs
to
the
ICARE
module
included
in
the
V2.0
version.
3.4.
FP
and
aerosol
transport
in
the
RCS
The
SOPHAEROS
module
(Cousin
et
al.,
2008)
simulates
trans-
port
of
FP
vapours
and
aerosols
in
the
RCS,
composed
of
a
1D
series
of
control
volumes,
through
gas
flow
to
the
containment,
account-
ing
for
the
chemical
reactions
in
the
vapour
phase.
Using
five
states
(suspended
vapours,
suspended
aerosols,
vapour
condensed
on
walls,
deposited
aerosols,
sorbed
vapours),
SOPHAEROS
uses
either
a
mechanistic
or
a
semi-empirical
approach
to
model
the
main
vapour-phase
and
aerosol
phenomena
(Fig.
4):
(a)
Vapour-phase
phenomena
-
Gas
equilibrium
chemistry.
The
reference
databank
in
MDB
(see
Section
5.1)
contains
about
800
species;
-
Chemisorption
of
vapours
on
walls;
-
Homogeneous
and
heterogeneous
nucleation;
-
Condensation/revaporisation
on/from
aerosols
and
walls;
-
Preliminary
model
for
kinetics
of
gaseous
phase
chemistry
(focusing
first
on
the
Cs–I–O–H
system),
mainly
based
on
the
interpretation
of
the
on-going
IRSN
CHIP
experimental
pro-
gramme
(Grégoire
and
Mutelle,
2012).
(b)
Aerosol
phenomena
-
Agglomeration:
gravitational,
Brownian
diffusion,
turbulent
dif-
fusion;
-
Deposition
mechanisms:
Brownian
diffusion,
turbulent
diffusion,
eddy
impaction,
sedimentation,
thermophoresis,
diffusiophore-
sis,
impaction
in
bends.
Deposit
of
aerosols
in
a
flow
contraction
(either
abrupt
one
with
a
90◦angle
or
conical)
can
be
simulated;
-
Remobilization
of
deposits:
revaporisation
and
mechanical
resus-
pension.
Two
models
are
available
for
aerosol
mechanical
resuspension:
the
“Force
balance”
model
and
the
“rock
and
roll”
one
(based
on
the
JRC
approach);
-
Dedicated
pool
scrubbing
model
to
deal
for
example
with
the
retention
of
aerosols
in
the
secondary
side
of
flooded
steam
gen-
erators
in
case
of
SGTR
scenario.
A
sectional
approach
is
adopted
to
model
the
aerosol
distribu-
tion:
the
same
species
composition
is
considered
in
all
the
aerosol
size
classes.
Unlike
in
containment
where
the
multi-component
size
distribution
is
important
to
evaluate
the
aerosol
behaviour,
the
multi-component
effect
is
not
pronounced
as
much
as
in
RCS
where
the
occurrence
of
coagulation
of
fresh
particles
with
aged
particles
is
not
so
common.
For
SOPHAEROS
applications
to
containment
issues
(see
Sec-
tion
3.7.3),
a
new
“liquid
state”
has
been
specifically
developed.
It
allows
to
compute
FP
transport
also
in
water
phase
(Foucher
et
al.,
2014)
but
no
speciation
is
assumed
except
for
species
directly
com-
puted
by
kinetic
reactions
such
as
in
IODE
module.
The
other
FPs
are
transported
under
global
element
form.
Materials
for
primary
circuits
of
CANDU
reactors
are
available
in
ASTEC
V2.0,
along
with
the
corresponding
chemisorption
corre-
lations.
By
default,
the
RCS
axial
nodalization
(set
of
volumes
connected
by
junctions)
used
by
SOPHAEROS
is
fitted
on
the
CESAR
one
(as
in
former
ASTEC
V1
versions),
but
it
is
now
possible
in
ASTEC
V2
to

P.
Chatelard
et
al.
/
Nuclear
Engineering
and
Design
272
(2014)
119–135
125
allow
through
a
dedicated
user
option
reducing
the
primary
circuit
meshing
to
be
used
by
SOPHAEROS
in
order
to
save
CPU
time.
3.5.
Direct
containment
heating
The
RUPUICUV
module
simulates
the
direct
containment
heat-
ing
(DCH)
which
may
potentially
develop
after
vessel
lower
head
rupture
occurrence
under
relatively
high
pressure;
corium
is
in
that
case
discharged
at
high
temperature
driven
by
primary
pressure
into
the
cavity
(involving
vessel
blow-down
and
cavity
pressuriza-
tion),
where
some
part
of
the
ejected
corium
may
be
potentially
oxidized
and
entrained
into
the
containment,
thus
contributing
to
the
containment
heat-up
and
additional
hydrogen
production.
Two
kinds
of
cavity
geometry
are
considered:
-
“Closed”
cavities
such
as
in
German
KONVOI
or
US
Zion
PWRs,
i.e.
without
direct
connection
to
the
“open
part”
of
the
containment,
but
through
a
series
of
intermediate
compartments;
-
“Open”
cavities
such
as
in
French
P’4
or
US
Calvert
Cliffs
PWRs,
i.e.
with
an
annular
space
around
the
vessel
towards
the
contain-
ment.
A
simplified
model
describes
the
amount
of
entrained
corium
and
the
entrainment
kinetics,
with
the
assumption
of
instanta-
neous
suspension
of
corium
in
cavity.
Droplets
entrainment
by
the
gas
flow
is
the
predominant
mechanism
for
corium
entrainment
from
cavity
into
the
containment.
The
entrainment
efficiency
is
evaluated
from
the
ratio
of
the
relative
velocities
particle/gas
in
the
annular
space
(based
on
gas/particle
friction
depending
on
particle
size).
A
simplified
particle
flow
path
in
the
cavity
is
assumed.
The
particle
size
is
either
given
by
the
user
or
calculated
with
the
Weber
number.
Particle
trapping
in
compartments
between
cavity
and
containment
is
directly
taken
into
account
through
the
global
correlation
of
entrained
corium.
In
the
case
of
“open”
geometry,
this
global
correlation
was
fitted
on
analytical
SURTSEY
(Blanchat
et
al.,
1999)
and
KAERI
experiments
(Kim
et
al.,
1999)
that
were
performed
in
this
geometry.
In
addition
to
this
basic
“open”
cavity
option,
a
new
generalized
correlation,
based
on
the
IRSN
interpre-
tation
of
DISCO
experiments
in
KIT
(Meyer
et
al.,
2004),
has
been
implemented
in
the
V2.0-rev2
version
to
be
specifically
used
for
French
PWRs
(Gen.
II
and
EPR)
safety
analyses.
Heat
transfer
debris/gas
is
assumed
to
be
instantaneous
in
cav-
ity
(complete
gas/debris
thermal
equilibrium).
Oxidation
of
the
entrained
corium
(zirconium
and
steel)
is
modelled
without
any
reaction
kinetics.
The
simple
CORIUM
parametric
module
simulates
the
behaviour
of
corium
droplets
transported
by
hot
gases
into
the
containment
atmosphere
and
sump,
heat
transfer
between
corium
and
gas
being
modelled
in
each
containment
zone.
3.6.
Molten-core–concrete-interaction
The
MEDICIS
module
(Cranga
et
al.,
2005)
simulates
Molten-
Core–Concrete
Interaction
(MCCI)
using
a
lumped-parameter
0D
approach
with
averaged
corium
layers.
Corium
remaining
in
the
cavity
interacts
with
concrete
walls
and
both
bottom
and
lateral
interfaces.
This
module
assumes
either
a
well-mixed
oxide/metal
pool
configuration
or
a
possible
pool
stratification
into
separate
oxide
and
metal
layers,
with
in
both
cases
the
possibility
to
account
for
a
detailed
description
of
the
upper
crust
(Fig.
5).
It
describes
concrete
ablation,
corium
oxidation
and
release
of
non
condensable
gases
(H2,
CO,
CO2)
into
the
containment.
Its
structure
is
flexible
enough
to
allow
an
easy
implemen-
tation
of
new
models
generated
by
R&D
outcomes.
A
robust
numerical
algorithm
for
cavity
erosion
was
developed,
includ-
ing
the
possibility
to
represent
a
multi-layered
concrete
basemat.
The
module
is
interfaced
with
the
general
physico-chemistry
package
MDB
(see
Section
5.1)
for
element
speciation
in
a
mix-
ture,
thermodynamic
data
(i.e.
liquidus
temperatures,
enthalpies),
and
thermo-physical
properties
(density,
viscosity,
etc.).
Moreover,
as
an
option,
MEDICIS
can
use
more
precise
thermo-chemistry
data
generated
outside
ASTEC
with
the
GEMINI2
software
(Cheynet
et
al.,
2002)
and
the
NUCLEA
database
(Bakardjieva
et
al.,
2008)
for
some
thermo-physical
properties
such
as
the
liquid
fractions
for
the
corium/concrete
mixture.
The
main
MEDICIS
models
in
ASTEC
V2.0
are:
-
A
model
of
the
structure
of
the
corium/concrete
interface
tak-
ing
into
account,
from
the
pool
bulk
to
the
concrete
interface,
a
convective
zone,
a
possible
conductive
zone
described
as
a
crust
and
a
slag
layer;
the
boundary
temperature
between
pool
con-
vective
and
conductive
zones,
called
solidification
temperature,
is
a
crucial
parameter;
the
intensive
validation
led
to
recommend
a
solidification
temperature
based
on
a
corium
volumetric
liquid
fraction
of
0.5;
-
Most
convective
heat
transfer
correlations
available
in
literature
for
both
the
corium/concrete
interface
(Kutateladze,
Bali)
and
the
interface
between
corium
layers
(Greene)
are
implemented,
and
a
special
attention
is
continuously
paid
(in
particular
as
concerns
the
heat
transfer
description
to
pool
interfaces
taking
into
account
gas
driven
convection
and
heat
conduction
in
series
across
a
possibly
existing
crust
and
a
slag
layer
located
along
the
con-
crete
interface)
to
on-line
feed-back
from
ASTEC
validation
on
the
ongoing
experiments
in
real
or
simulant
materials
(VULCANO
with
real
materials
in
CEA
(Journeau
et
al.,
2009),
CCI
with
real
materials
in
ANL
(Farmer
et
al.,
2006),
ARTEMIS
with
simulant
materials
in
CEA
(Veteau
et
al.,
2006));
-
Radiative
heat
transfer
in
the
cavity.
In
a
stand-alone
mode,
the
gas
in
the
cavity
is
supposed
to
be
fully
absorbent.
When
MEDI-
CIS
is
coupled
with
CPA,
the
radiative
heat
transfers
between
the
corium,
the
gas
and
the
cavity
walls
can
be
described
(see
Section
4.2).
User’s
parameters
can
be
specified
such
as
the
absorption
length
by
atmosphere
(containing
steam,
carbon
dioxide
or
con-
crete
aerosols)
and
the
gas
emissivity;
-
Models
of
corium
coolability
in
case
of
water
injection
upon
the
corium
pool
surface
(derived
from
simple
models
from
the
USNRC
CORQUENCH
code
version
(Farmer,
2001)),
including
water
ingression
through
the
upper
crust
and
melt
eruption
from
the
corium
pool
towards
the
overlying
water
pool;
-
Models
of
evolution
of
corium
pool
configurations,
depending
on
criteria
using
the
superficial
gas
velocity
and
on
differ-
ences
between
oxide
and
metal
densities
determining
the
switch
between
homogeneous
and
stratified
pools;
-
Use
of
the
MDB
package.
It
allows
to
evaluate
the
corium
layers’
thermo-physical
properties
and
to
treat
the
corium
oxidation:
metals
are
oxidized
instantaneously
in
proportion
to
the
mass
of
available
gases
with
a
priority
rank
starting
from
zirconium
to
chromium,
then
nickel
and
finally
iron;
-
Capability
to
account
for
successive
MCCIs
(useful
feature
to
sim-
ulate
MCCI
in
the
EPR
cavity
and
then
in
the
spreading
chamber),
keeping
in
mind
that,
in
the
V2.0
version,
only
sequential
MCCIs
in
different
volumes
are
possible
up
to
now
(that
means
MCCI
in
the
first
volume
stops
when
MCCI
in
the
second
one
starts);
-
Corium
pouring
kinetics
from
the
cavity
towards
the
core-catcher
(EPR
design)
using
a
simple
model
(based
on
a
combined
use
of
a
Bernoulli
flow
approach
and
of
properties
from
MDB);
-
Evaluation
of
the
melt
spreading
capability
(EPR
design)
thanks
to
a
simple
analytical
model
(combination
of
the
fraction
of
the
occupied
area
and
of
a
spreading
kinetics).
The
spreading
mod-
elling
allows
in
particular
to
evaluate
the
spreading
radius
and,
thus
the
spreading
duration
of
corium,
the
thickness
of
the
spread
corium
and
of
the
spread
solidified
part,
and
finally
the
duration

126
P.
Chatelard
et
al.
/
Nuclear
Engineering
and
Design
272
(2014)
119–135
metal layer
upper hconv
lateral hconv
in oxide
radiation
metal layer
oxide layer
h oxide/me
tal
upp
er oxid
e crust
hco
nv
in me
tal
conc
rete
hslag (θ)
slag layer
liquid/gas)
conc
rete
Oxide or metal c
rust at
corium/co
ncrete interfa
ce
hslag
laye
r
slag laye
r
(liquid/ga
s)
metal layer
upper hconv
lateral hconv
in oxide
radiation
metal layer
oxide layer
h oxide/me
tal
upp
er oxid
e crust
hco
nv
in me
tal
conc
rete
hslag (θ)
slag layer
liquid/gas)
conc
rete
Oxide or metal c
rust at
corium/co
ncrete interfa
ce
hslag
laye
r
slag laye
r
(liquid/ga
s)
Fig.
5.
ASTEC
MCCI
modelling.
for
corium
to
reach
the
trigger
and
the
time
for
the
beginning
of
corium
quenching
by
water
injection
above
corium;
-
Model
to
evaluate
the
FP
release
from
the
ex-vessel
corium
pool
during
MCCI
(this
model
is
derived
from
that
of
ELSA
assuming
the
FP
vapours
are
transported
by
gas
bubbles);
-
Empirical
model
to
evaluate
the
aerosol
production
during
MCCI.
This
model
is
based
on
the
evaluation
of
the
concrete
aerosol
concentration
above
the
pool
using
an
empirical
approach
with
parameters
fitted
on
ACE
experiments
(performed
in
ANL
(Fink
et
al.,
1992))
and
it
determines
the
aerosol
mass
release
rate
in
proportion
to
the
gas
volumetric
rate
escaping
out
of
the
pool.
Several
improvements
are
foreseen
at
short
term
in
the
next
ASTEC
V2.0
revision
concerning
MCCI
as
follows.
As
already
noticed,
in
order
to
answer
specific
EPR
safety
analyses
requirements,
most
of
the
on-going
MEDICIS
modelling
efforts
are
devoted
at
IRSN
to
continuously
improve
the
description
of
the
EPR
core
catcher
com-
ponent.
For
instance,
works
are
currently
underway
to
develop
a
more
adequate
modelling
of
the
steel
structure
behaviour
under
the
sacrificial
concrete
of
the
spreading
chamber
of
the
core
catcher
in
order
to
evaluate
if
the
steel
structure
could
be
damaged
by
the
steel
fusion.
The
modelling
of
the
corium
upper
crust
behaviour
will
be
improved
at
short
term
in
order
to
better
evaluate
the
upper
crust
thickness,
which
was
overestimated
by
the
code
compared
to
some
MCCI
experiment
observations.
Further
improvements
concerning
the
corium
coolability
are
also
planned.
Notably,
a
debris
layer,
formed
from
the
melt
erup-
tion
process
and
interacting
with
the
corium
upper
crust,
will
be
modelled.
The
review
of
the
water
ingression
and
the
melt-eruption
modelling
is
also
scheduled
in
2013.
3.7.
Thermal-hydraulics
and
FP/aerosols
behaviour
in
containment
The
CPA
module
simulates
thermal-hydraulics
(including
hydrogen
combustion)
and
aerosol
behaviour
in
the
containment
(Kljenak
et
al.,
2010).
It
consists
of
two
main
sub-modules,
namely
THY
(for
thermal-hydraulics)
and
AFP
(for
aerosols
and
FPs).
The
discretisation
through
a
“lumped-parameter”
approach
(0D
zones
connected
by
junctions
and
surrounded
by
walls)
simulates
sim-
ple
or
multi-compartment
containments
(tunnels,
pit,
dome)
with
possible
leakages
to
the
environment
or
to
normal
buildings,
with
specified
openings
to
the
environment.
Several
real
compartments
can
either
be
combined
to
become
one
CPA
zone
or
large
compart-
ments
can
be
divided
into
several
zones
to
cover
flow
peculiarities
more
realistically,
e.g.
steam
or
hydrogen
plumes.
Using
the
sources
of
steam,
hydrogen,
FP
gases
and
aerosols
from
RCS
or
from
corium
in
the
cavity
during
MCCI
provided
by
other
modules
of
ASTEC,
CPA
calculates
gas
distribution,
temperature
field,
pressure
build-up,
hydrogen
combustion
and
FP
and
aerosol
distribution
and
deposi-
tion.
As
concerns
the
description
of
the
thermal-hydraulics
models,
general
information
are
provided
in
the
next
sub-section
while
the
modelling
of
hydrogen
combustion
phenomena
will
be
discussed
separately
(see
Section
3.7.2).
3.7.1.
Containment
thermal-hydraulics
The
CPA-THY
models
describe
phenomena
such
as
pressure
and
temperature
build-up
and
history,
local
temperature
and
pressure
distributions,
local
gas
distributions
(steam
and
different
non-
condensable
gases),
local
heat
transfer
to
walls
(free
and
forced
convection,
radiation,
condensation),
1D
heat
conduction
in
struc-
tures
(plate
or
cylinders,
consisting
of
several
material
layers),
as
well
as
gas
(hydrogen
and
carbon
monoxide)
combustion.
The
thermal-hydraulics
state
of
a
zone
can
be
described
accord-
ing
to
two
concurrent
approaches:
-
Either
by
the
equilibrium
model
assuming
water
and
atmosphere
homogeneously
mixed
for
saturated
and
superheated
(no
water)
conditions,
i.e.
water
and
gases
at
the
same
temperature;
-
Or
by
the
non-equilibrium
model
where
deposited
and
airborne
water
are
separately
balanced,
i.e.
with
their
own
energy
(sepa-
rate
temperatures
and
mass
balances
for
atmosphere
and
water
sump)
and
mass
balances.
Mass
transfer
between
zones
is
described
separately
for
gas
and
liquid
flows
by
momentum
equations
(unsteady,
incompress-
ible
or
steady
compressible)
accounting
for
the
height
differences
between
zone
centres.
For
a
realistic
description
of
accidents,
models
are
available
for
the
behaviour
of
engineered
safety
systems
such
as
passive
auto-
catalytic
recombiners
(PAR)
of
different
types,
usual
containment
sprays
(Plumecocq
et
al.,
2005),
other
pressure
suppression
sys-
tems,
etc.
Besides
the
detailed
recombiner
models,
fast
running
correla-
tions
were
added,
faster
and
simpler
to
use.
Two
kinds
of
such
correlations
are
available
for
the
simulation
of
different
kinds
of
Siemens
PAR
type
FR90/1
series.
But
they
do
not
give
any
informa-
tion
on
PAR
temperature
or
outlet
concentration,
on
velocity
and
on
mass
flow
rate.

P.
Chatelard
et
al.
/
Nuclear
Engineering
and
Design
272
(2014)
119–135
127
The
effect
of
spray
systems
on
thermal-hydraulics
is
validated
by
post-calculation
of
various
experiments,
among
them
the
IRSN
separate-effect
test
facility
CARAIDAS
(Ducret
et
al.,
1996).
The
DRASYS
model
simulates
pressure
suppression
systems,
e.g.
the
German
BWR
type
suppression
system
or
the
Bubble
Condenser
system
of
VVER-440/213.
CPA
has
been
used
also
(Fontanet
et
al.,
2008;
Ramlakan
et
al.,
2010)
for
evaluations
of
the
thermal-hydraulic
and
dust
aerosol
behaviour
in
a
HTR
containment
(here
PBMR
South-African
project)
in
case
of
a
postulated
breach
in
the
helium
pressure
boundary.
The
comparison
between
ASTEC
and
CONTAIN
code
calculations
had
shown
very
similar
results.
3.7.2.
Gas
combustion
In
ASTEC,
gas
combustion
occurs
according
to
different
criteria:
user-input
or
crossover
of
flammability
limits
in
the
Shapiro
dia-
gram.
For
the
latter,
4
different
flammability
limits,
determined
at
atmospheric
pressure
and
room
temperature,
are
defined
on
the
ternary
Shapiro
diagram
hydrogen-air-steam.
Few
concurrent
models
are
available
in
ASTEC
V2.0
to
simulate
hydrogen
combustion:
FRONT
and
DECOR
models,
fully
integrated
in
the
CPA
module,
and
COVI
and
PROCO
as
separate
modules.
Since
only
FRONT
and
COVI
are
now
recommended
for
use,
they
will
be
described
below.
The
simplest
one,
namely
the
COVI
module,
computes
the
maximal
value
of
pressure
build-up
under
adiabatic
conditions
of
hydrogen
combustion
(AICC
or
Adiabatic
Isochoric
Complete
Combustion).
It
yields
the
situation
of
the
whole
containment
in
the
Shapiro
diagram,
and
the
temperature
and
pressure
peaks
at
current
time
as
if
H2and/or
CO
were
instantaneously
burnt.
The
available
hydrogen
in
all
CPA
zones
is
summed
up
and
burnt
down
to
a
user-defined
percentage.
The
energy
is
distributed
on
the
total
volume
of
all
zones.
There
is
no
feedback
on
CPA
calculations,
but
only
an
evaluation
of
this
pressure
that
can
be
considered
as
a
max-
imal
envelope
value.
It
is
activated
either
at
a
time
selected
by
the
user
or
when
one
of
the
4
abovementioned
flammability
limits
is
crossed
in
a
given
zone.
Moreover,
it
can
be
noted
that,
thanks
to
its
new
“multi-domain”
functionality
(a
domain
being
a
cluster
of
CPA
zones)
with
respect
to
former
ASTEC
V1
capabilities,
the
para-
metric
COVI
module
is
now
able
to
compute
the
virtual
combustion
in
both
the
inner
containment
and
the
inter
containment
space
for
configurations
like
in
the
French
PWR
1300
MWe.
The
new
flame
FRONT
model
(ASTEC
V2
novelty
with
respect
to
former
V1
versions)
allows
to
account
for
the
flame
front
propaga-
tion
(Brähler
and
Koch,
2011),
thus
providing
a
valuable
extension
to
the
V1
simple
COMB
model
in
case
of
a
multi-compartment
geometry.
Indeed,
this
model
notably
provides
ignition
time
and
burning
duration
for
the
CPA
zones.
The
propagation
of
the
flame
front
takes
place
along
the
1D
network
comprised
of
the
junctions
between
the
containment
zones;
besides
regular
atmospheric
con-
nections
between
the
zones,
also
rupture
discs,
doors
and
pipes
are
taken
into
account
for
the
flame
front
propagation.
The
flame
front
velocity
is
the
sum
of
gas
velocity
and
combustion
velocity.
Therefore
a
correct
treatment
of
the
interaction
of
combustion
and
flow
is
possible.
The
turbulent
flame
velocity
is
calculated
by
the
Flamelet
model
of
Peters.
By
integrating
the
flame
front
velocity,
the
position
of
the
flame
front
is
obtained.
Since
there
is
no
calculation
of
the
turbulent
intensity
in
the
lumped-parameter
code,
an
ade-
quate
correlation
is
selected,
assuming
the
turbulent
intensity
to
be
a
function
of
the
Reynolds
Number.
Moreover,
in
order
to
answer
plant
calculations
requirements
(since
plant
conditions
are
quite
different
from
dedicated
hydrogen
deflagration
experiments
where
the
initial
ignition
is
artificially
triggered
at
a
specific
time),
the
pos-
sibility
of
ignition
in
multiple
zones
at
any
time
was
implemented
in
the
V2.0-rev2
version
while
the
re-ignition
of
already
burnt
zones
can
be
now
also
taken
into
account
in
the
ASTEC
simulations.
Fig.
6.
ASTEC
modelling
of
aerosol/FP
behaviour
in
containment.
3.7.3.
Aerosols
and
FP
behaviour
in
the
containment
The
CPA-AFP
models
for
aerosol
transport
and
depletion
are
mostly
based
on
the
GRS
COCOSYS
code
(Allelein
et
al.,
2008).
They
describe
phenomena
such
as
volume
condensation
and
growth
of
insoluble
and
soluble
aerosol
particles,
behaviour
of
chemically
different
aerosol
components,
and
agglomeration
and
deposition
processes
(Fig.
6).
The
aerosol
calculation
is
based
on
the
poly-
disperse
MAEROS
model
(Gelbard,
1982).
The
aerosol
retention
in
gas-sparged
water
pools
can
be
simu-
lated
with
the
pool-scrubbing
model
SPARC.
SPARC
can
be
linked
with
the
detailed
DRASYS
model
for
pressure
suppression
systems
(PSS)
in
CPA-THY.
However,
since
the
DRASYS
zone
includes
mech-
anistic
hydrodynamic
models
(to
describe
the
processes
in
the
PSS
like
vent
clearing
and
swell,
the
condensation
of
steam
in
the
water
and
the
transport
of
non
condensable
gases
through
the
water),
this
detailed
model
can
consume
considerable
CPU
time
under
certain
conditions.
Therefore,
as
an
option
(insertion
model),
an
alterna-
tive
fast-running
model
neglecting
short
term
processes
has
been
developed
at
GRS
which
concentrates
on
the
simulation
of
quasi-
steady
flow
of
steam/gas
mixtures
in
the
water
pool.
The
FP
transport
model
FIPHOST
calculates
the
transport
and
depletion
of
gaseous
and
particulate
FPs
by
the
mobile
hosts
like
gas,
aerosols,
and
water
or
immobile
hosts
like
walls
in
the
atmo-
sphere
or
in
a
sump,
respectively.
With
the
aerosols
retained
in
water
pools
the
hosted
particulate
FP
are
scrubbed
too
using
the
SPARC
modelling
approach.
The
dry
aerosols
resuspension
model
has
been
improved
in
the
V2.0-rev2
version.
This
model,
which
describes
resuspension
from
a
multi-layer
particle
bed
deposited
on
solid
surfaces,
calculates
a
force
balance
between
adhesive
forces
that
retain
the
particles
on
the
surface
and
aerodynamic
forces
that
lift
the
particles
into
the
atmosphere.
The
acting
forces
are
distributed
according
to
proba-
bility
density
functions.
The
effect
of
spray
systems
on
aerosols/FPs
is
thoroughly
mod-
elled
and
validated
e.g.
on
the
basis
of
CSE
(USA)
experiments:
wash-out
of
aerosol
particles/particulate
FPs
from
the
atmosphere
and
wash-down
of
deposits
from
containment
walls.
For
specific
ASTEC
applications
(such
as
for
instance
simulations
of
accidents
in
the
ITER
fusion
installation),
it
appeared
valuable
to
use
the
yet
available
very
detailed
and
mechanistic
modelling
of
gas
chemistry
for
the
treatment
of
large
volumes.
Thus
the
scope
of
the
SOPHAEROS
module
(see
Section
3.4)
has
been
extended
to
make
it
applicable
outside
the
circuits
in
order
to
be
able
to
also
deal
with
element
speciation
in
the
containment.
For
that
purpose,
three
new
states
have
been
specifically
developed
in
SOPHAEROS
from
V2.0-rev2
in
addition
to
the
five
existing
states
yet
handled
up

128
P.
Chatelard
et
al.
/
Nuclear
Engineering
and
Design
272
(2014)
119–135
Liquid phase
Volati
l
spec
ies are transferred to
the gase
ous phase (I
2
, RI
)
Volati
l
spec
ies are transferred to
the gase
ous phase (I
2
, RI
)
Ag
2
OAg
2
O
AgI ( ↓
↓
)
Ag
(↓
↓
Ag
If Ag is present,
iodides ions can
form insoluble compounds (AgI…)
Ag
2
OAg
2
O
AgI ( ↓
↓
)
Ag
(↓
↓
Ag
Ag
2
OAg
2
OAg
2
OAg
2
O
AgI ( ↓
↓
)
Ag
(↓
↓
Ag
AgI ( ↓
↓
)
Ag
(↓
↓
Ag
AgI ( ↓
↓
)
Ag
(↓
↓
Ag
AgI ( ↓
↓
)
Ag
(↓
↓
Ag
If Ag is present,
iodides ions can
form insoluble compounds (AgI…)
Gaseous
phase
A trapp
ed fraction of I
2
is
converte
d into
RI, th
at are
destructed into IO
x
I
2
reacts
with
surfaces (ads
orption,
desorption)
I
2
reacts
with
surfaces (ads
orption,
desorption)
Iodides ions are oxidided by water
radicals (OH°) and form I
2
that can be
hydrolised,
adsorbe
d on imm
ersed
paint,
or
react with
or
ganics in
solution to form organic iodides
RI
γ
γ
ROH
H
2
O
OH
-
R
RI
γ
γ
ROH
H
2
O
OH
-
R
-
γ
γ
IO
3
I
H
2
O
HOI
γ
γ
3
I
H
2
O
HOI
γ
γ
2
Iodides ions are oxidided by water
radicals (OH°) and form I
2
that can be
hydrolised,
adsorbe
d on imm
ersed
paint,
or
react with
or
ganics in
solution to form organic iodides
Iodides ions are oxidided by water
radicals (OH°) and form I
2
that can be
hydrolised,
adsorbe
d on imm
ersed
paint,
or
react with
or
ganics in
solution to form organic iodides
RI
γ
γ
ROH
H
2
O
OH
-
R
RI
γ
γ
ROH
H
2
O
OH
-
R
-
γ
γ
IO
3
I
H
2
O
HOI
γ
γ
3
I
H
2
O
HOI
γ
γ
2
RI
γ
γ
ROH
H
2
O
OH
-
R
RI
γ
γ
ROH
H
2
O
OH
-
R
RI
γ
γ
ROH
H
2
O
OH
-
R
RI
γ
γ
ROH
H
2
O
OH
-
R
-
γ
γ
IO
3
I
H
2
O
HOI
γ
γ
3
I
H
2
O
HOI
γ
γ
2
-
γ
γ
IO
3
I
H
2
O
HOI
γ
γ
3
I
H
2
O
HOI
γ
γ
-
γ
γ
IO
3
I
H
2
O
HOI
γ
γ
3
I
H
2
O
HOI
γ
γ
γ
γ
IO
3
I
H
2
O
HOI
γ
γ
3
I
H
2
O
HOI
γ
γ
IO
3
I
H
2
O
HOI
γ
γ
3
I
H
2
O
HOI
γ
γ
3
I
H
2
O
HOI
γ
γ
3
I
H
2
O
HOI
γ
γ
3
I
H
2
O
HOI
γ
γ
3
I
H
2
O
HOI
γ
γ
3
I
H
2
O
HOI
γ
γ
3
I
H
2
O
HOI
γ
γ
222
I-
Iodine aerosol sediment and settle
on walls. If
solu
bles (C
sI…),
they
form iodide
s ions (I
-) in
the
aque
ous phase. The inso
luble
aerosols (AgI…) stay in
the bottom
of the sump
I-
Iodine aerosol sediment and settle
on walls. If
solu
bles (C
sI…),
they
form iodide
s ions (I
-) in
the
aque
ous phase. The inso
luble
aerosols (AgI…) stay in
the bottom
of the sump
Iodine aerosol sediment and settle
on walls. If
solu
bles (C
sI…),
they
form iodide
s ions (I
-) in
the
aque
ous phase. The inso
luble
aerosols (AgI…) stay in
the bottom
of the sump
gazeux
aérosolsI
I
circuit
gazeux
aérosolsI
I
circuit
aérosolsI
I
circuit
circuit
Iodine oxides sediment and settle on
the surfaces (walls, surface developed
by aerosols in suspension)
Iodine oxides sediment and settle on
the surfaces (walls, surface developed
by aerosols in suspension)
Volatile iodin
e reac
ts wi
th air radio
lytic produc
ts , and oxidizes
a fraction of I
2
and RI => formation of iodine oxides (considered
as fines particles)
RIRI
O
IO
3
3
H
2
O
(v)
IO
3
3
H
2
O
(v)
O
IO
3
3
H
2
O
(v)
IO
3
3
H
2
O
(v)
γ
γ
γ
γ
I
2
I
2
Fig.
7.
ASTEC
modelling
of
iodine
chemistry
in
containment.
to
now
(“suspended
vapours”,
“suspended
aerosols”,
“vapour
con-
densed
on
walls”,
“deposited
aerosols”,
“sorbed
vapours”).
These
new
states
(“liquid
state”
and
the
“wet
and
dry
walls”
as
in
the
IODE
module)
are
required
to
deal
with
FPs
and
aerosols
behaviour
in
case
of
a
two-phase
carrier
fluid
and
iodine
chemistry
in
the
containment.
3.8.
Iodine
and
ruthenium
behaviour
in
the
containment
The
IODE
module
(Bosland
et
al.,
2010)
simulates
iodine
and
ruthenium
behaviour
inside
the
containment,
except
the
transport
through
compartments
of
the
associated
species,
which
is
com-
puted
by
the
CPA
module
(i.e.
as
for
transport,
the
IODE
module
is
directly
using
the
junction
flow
rates
given
by
the
CPA
containment
module).
The
module
is
validated
using,
for
example,
experiments
conducted
under
the
auspices
of
SARNET
(Haste
et
al.,
2012).
While
ASTEC
V1
dealt
only
with
iodine
behaviour
in
the
con-
tainment,
the
IODE
module
has
been
extended
in
ASTEC
V2
to
the
detailed
simulation
of
the
ruthenium
radiochemistry
in
the
con-
tainment,
according
in
particular
to
the
high
radio-toxicity
of
this
species
through
its 103Ru
and 106Ru
isotopes.
In
severe
accident
conditions,
gaseous
ruthenium
can
be
formed
and
released
out-
side
the
containment,
thus
inducing,
at
short
and
middle
term,
significant
radiological
consequences
that
have
to
be
considered
for
emergency
safety
plan.
As
radiochemical
reactions
involve
a
large
amount
of
elemen-
tary
reactions,
which
compose
large
and
complex
mechanisms,
an
easier
way
is
to
consider
a
phenomenological
approach,
as
applied
in
ASTEC.
Therefore,
as
concerns
iodine,
the
IODE
module
is
com-
posed
of
around
40
phenomenological
models
that
focus
on
the
predominant
chemical
reactions
in
sump,
gas
phase
and
at
the
interface
with
surfaces
(Fig.
7).
More
precisely,
it
describes
in
a
kinetic
way
(i.e.
non-equilibrium)
the
chemical
transformations
of
iodine
in
the
reactor
containment
building,
taking
into
account
three
kinds
of
reaction:
thermal
reactions,
radiolytic
reactions
and
mass
transfer
processes.
The
iodine
species
involved
in
the
reac-
tions
are
I2,
CH3I,
I−,
IO3−,
HOI
and
AgI.
The
methyl
iodide
is
assumed
to
represent
all
the
organic
forms,
knowing
that
CH3I
is
the
most
volatile
form.
The
inorganic
forms
present
in
gas
phase
are
molecular
iodine
I2and
iodine
oxides
represented
by
IO3species.
The
other
species
are
Ag
(from
in-core
control
rods),
Ag2O
(Ag
released
and
supposed
to
be
oxidized
in
the
containment),
Rp
and
R
(respectively
initial
amount
of
organics
in
the
wall
paints
and
gaseous
concentration
in
the
containment
resulting
from
the
paint
releases)
and
O3formed
in
gas
phase.
As
concerns
ruthenium,
the
IODE
module
is
focusing
on
the
three
predominant
chemical
reactions
in
gas
phase.
The
main
ruthenium
species
involved
in
these
reactions
are
ruthenium
diox-
ide
RuO2under
aerosol
form
and
gaseous
ruthenium
tetroxide
RuO4.
(a)
Mass
transfer
reactions
-
The
main
modelled
phenomena
are:
-
Adsorption/desorption
of
molecular
iodine
on
painted,
metal
and
concrete
walls;
-
Mass
transfer
between
sump
and
gas
phase
for
diffu-
sion/convection
processes,
based
on
the
interpretation
of
IRSN
SISYPHE
experiments
(Spitz
et
al.,
2005).
This
model
is
also
avail-
able
in
evaporating
conditions;
-
Condensation
of
steam
on
the
walls
and
on
the
sump
surface;
-
Transfer
of
non
volatile
iodine
oxides
towards
the
sump;
-
Effect
of
spray
on
molecular
iodine:
mass
transfer
between
gas
phase
and
droplet,
interfacial
equilibrium
at
the
droplet
surface,
liquid
mass
transfers
inside
the
droplet,
chemical
reactions
in
the
bulk
liquid.
The
module
computes
kinetics
of
the
overall
process
during
the
droplet
fall
down
and
the
output
information
is
the
rate
of
capture
of
iodine
I2for
each
compartment.
(b)
Liquid
phase
chemistry
-
As
concerns
the
iodine
aqueous
chemistry
in
the
sump,
the
main
modelled
reactions
are:
-
Hydrolysis
of
molecular
iodine
I2;
-
HOI
dissociation/disproportionation;
-
Oxidation
of
I-
by
the
oxygen
dissolved
in
the
sump
water;
-
Radiolytic
oxidation
of
I−into
I2in
the
sump,
where
oxidation
depends
on
the
production
rate
of
OH
radicals,
and
where
I2
reduction
is
temperature-,
pH-
and
[I-]-dependent;
-
Reduction
of
iodates
by
radiolysis
into
molecular
iodine;
-
Silver
iodide
(AgI)
formation
by
heterogeneous
reactions
(both
Ag2O/I−and
Ag/I2reactions
are
considered,
the
reaction
of
Ag2O
with
I−depending
on
the
solubility
of
Ag2O
and
AgI);
-
Formation
of
organic
iodide
RI
by
homogeneous
reaction
in
the
liquid
phase
with
the
Taylor’s
homogeneous
model:
solvents
are
released
from
paint
in
liquid
phase,
then
oxidized
under
radiation

P.
Chatelard
et
al.
/
Nuclear
Engineering
and
Design
272
(2014)
119–135
129
to
form
organic
acids,
and
finally
RI
are
formed
by
interaction
between
I2and
solvents
or
sub-products;
-
Decomposition
of
organic
iodides
in
the
liquid
phase,
according
to
two
possible
destruction
processes
(either
radiolysis
or
hydrol-
ysis).
(c)
Gas
phase
chemistry
-
As
concerns
the
iodine
chemistry
in
the
containment
gaseous
phase,
the
main
modelled
reactions
are:
-
Kinetics
of
ozone
(O3)
formation;
-
Oxidation
of
molecular
iodine
in
iodate
by
O3air
radiolysis
prod-
ucts;
-
Decomposition
of
iodine
oxides
into
organic
iodides;
-
Organic
iodide
formation,
managed
either
by
an
homogeneous
model
or
by
the
Funke
heterogeneous
model;
-
Gas
phase
radiolytic
destruction
of
organic
iodine,
based
on
work
performed
in
the
ICHEMM
FP5
project
(Dickinson
et
al.,
2003),
especially
on
AEA-T
experiments.
As
concerns
the
ruthenium
chemistry
in
the
gaseous
phase,
a
dedicated
modelling,
based
on
the
on-going
IRSN
EPICUR
(Mun
et
al.,
2008)
(Giordano
et
al.,
2010)
and
STEM
experiments
(Clément
and
Simondi-Teisseire,
2010),
has
been
implemented
in
ASTEC
V2
in
order
to
complete
FP
chemistry
models
inside
the
containment.
The
main
modelled
reactions
are:
-
Decomposition
in
bulk
phase
(dry
and
moist
air)
of
the
ruthenium
tetroxide
RuO4(g)
coming
from
the
RCS
into
ruthenium
deposit;
-
Ruthenium
ozonation
(reaction
between
ruthenium
deposit
and
ozone);
-
Oxidation
of
ruthenium
deposit
due
to
the
action
of
air
radiolytic
products
(revolatilisation
from
RuO2surface
deposit
producing
RuO4(g)
at
low
temperature).
3.9.
Other
ASTEC
modules
3.9.1.
Dose
rate
in
the
containment
The
DOSE
module,
which
was
designed
for
use
in
most
of
the
IODE
gas
phase
chemical
reactions,
was
specifically
implemented
in
ASTEC
V2
to
answer
IRSN
PSA2
requirements.
This
module
allows
evaluating
the
dose
rate
in
bulk
gas
phase
for
each
zone
of
the
con-
tainment,
as
well
as
the
inner
wall
dose
rate,
knowing
that
the
dose
rates
include
␤
and
␥
radiation
contributions
relative
to
each
isotope.
Anyway,
it
has
however
to
be
underlined
that,
up
to
now,
this
module
was
only
validated
by
comparison
with
dedicated
IRSN
codes.
3.9.2.
Decay
heat
and
isotopes
The
ISODOP
module
simulates
decay
of
FP
and
actinide
isotopes
in
different
zones
of
the
reactor
and
of
the
containment.
It
starts
the
calculation
using
an
initial
isotope
inventory
gen-
erated
by
a
dedicated
code
(for
instance
the
CEA
code
PEPIN)
and
allows
estimating
decay
heat
and
activity
in
the
core,
in
the
RCS,
in
the
containment
and
in
the
environment.
In
V2.0,
the
ISODOP
module
was
based
on
the
DOP
database
from
CEA
containing
the
description
of
720
isotopes
while
the
JEFF
(Joint
Evaluated
Fis-
sion
and
Fusion)
database
dealing
with
∼3800
isotopes
is
now
also
available
in
the
subsequent
V2.0
revisions
as
an
alternative
to
the
original
CEA
database.
3.9.3.
Management
of
engineered
safety
systems
The
SYSINT
module
allows
the
user
to
easily
simulate
the
man-
agement
of
engineered
safety
features,
such
as,
for
instance,
safety
injection
systems,
pressurizer
spray
and
heaters,
management
of
Fig.
8.
Schematic
of
ASTEC
account
for
some
engineered
safety
systems
in
a
PWR.
steam
generators,
containment
spray
system
in
direct
or
recircula-
tion
mode,
hydrogen
recombiners,
etc.
.
.
as
illustrated
hereafter
in
Fig.
8.
4.
Specific
coupling
modes
between
ASTEC
modules
As
explained
in
Section
2,
the
ASTEC
V2.0
modules
are
gener-
ally
running
sequentially
within
a
macro-time
step,
the
coupling
between
the
modules
being
therefore
explicit.
There
are
how-
ever
two
main
exceptions
to
this
general
rule
which
concern
the
CESAR/ICARE
and
CPA/MEDICIS
couplings.
4.1.
CESAR/ICARE
coupling
The
new
coupling
technique,
which
shall
allow
CESAR
to
cover
the
thermal-hydraulics
in
the
whole
RCS
(vessel
and
loops)
during
the
whole
transient,
is
planned
to
be
implemented
in
the
future
V2.1
version
(see
Section
8).
Waiting
for
such
a
new
CESAR/ICARE
running
mode
to
become
available,
the
treatment
of
the
thermal-hydraulics
in
the
core
region
is
still
made
in
ASTEC
V2.0
(like
in
V1
versions)
using
sequentially
two
different
numerical
approaches
during
a
given
cal-
culation,
depending
on
the
phase
of
the
accident
(front-end
phase
or
degradation
phase).
During
the
front-end
phase,
CESAR
alone
calculates
the
thermal-
hydraulics
in
the
whole
RCS,
i.e.
including
the
vessel
(lower
plenum,
core,
bypass,
downcomer
and
upper
plenum).
An
automatic
switch
to
ICARE
for
simulation
of
in-vessel
core
degradation
phenomena
is
then
applied,
depending
on
specific
criteria
e.g.
void
fraction
in
primary
circuit
loop,
void
in
the
upper
plenum,
void
fraction
at
the
top
of
the
core,
steam
temperature
at
the
top
of
the
core,
and
non
isolated
accumulators
mass
fraction.
This
switch
becomes
effective
around
time
of
start
of
core
uncovery,
and
in
any
case
before
the
start
of
Zry
cladding
oxidation
by
steam.
After
the
switch,
CESAR
calculates
only
thermal-hydraulics
in
the
loops
and
the
ves-
sel
upper
plenum,
while
in
addition
to
the
degradation
phenomena
ICARE
calculates
thermal-hydraulics
in
the
remaining
part
of
the
vessel
(core,
bypass,
lower
plenum
and
downcomer)
all
along
the
degradation
phase.
During
the
degradation
phase
(i.e.
after
the
switch),
a
spe-
cific
prediction-correction
numerical
coupling
approach
is
used

130
P.
Chatelard
et
al.
/
Nuclear
Engineering
and
Design
272
(2014)
119–135
between
CESAR
and
ICARE.
At
each
macro
time-step,
first
the
RCS
thermal-hydraulics
is
predicted
by
ICARE
using
as
inputs
the
CESAR
outputs
from
the
previous
macro
time-step.
Then
CESAR
computes
the
two-phase
thermal-hydraulics
in
the
circuits
using
source
and
sink
terms
supplied
by
ICARE
at
the
vessel
junctions
and
it
finally
evaluates
the
possible
water
fallback
into
the
vessel
at
the
end
of
the
macro
time-step.
Both
numerical
schemes
are
therefore
discon-
nected
and
each
module
is
running
at
its
own
time-step.
Meeting
points
occur
at
the
end
of
an
intermediate
macro
time-step.
This
macro
time-step
management
had
been
differentiated
from
the
other
modules
in
order
to
gain
CPU
time.
In
addition
to
this
usual
coupling
at
the
top
of
core
and
top
of
downcomer,
the
CESAR
and
ICARE
modules
can
be
also
chained
one
after
the
other
(in
a
fully
explicit
manner)
in
the
lower
head
region
to
support
preliminary
analyses
about
the
IVR
concept.
4.2.
CPA/MEDICIS
coupling
A
specific
prediction-correction
coupling
approach
was
devel-
oped
between
MEDICIS
and
CPA
when
representing
the
cavity
as
a
CPA
volume.
Two
levels
of
coupling
have
been
developed:
firstly
the
simplified
coupling
assuming
that
the
cavity
gases
are
fully
absorbent
and
then
the
detailed
coupling
(more
realistic
but
more
CPU
time
consuming)
assuming
that
the
cavity
gases
can
be
trans-
parent,
partially
absorbent
or
fully
absorbent.
First,
in
the
prediction
step,
MEDICIS
calculates
the
whole
behaviour
of
the
cavity
and
notably
evaluates
the
gas
tempera-
ture
in
the
cavity.
With
the
detailed
coupling,
MEDICIS
calculates
also
all
the
radiative
exchanges
between
the
upper
surface
of
the
corium
pool,
the
internal
wall
surfaces
of
the
reactor
cavity
(i.e.
lat-
eral
upper
cavity
walls
or
lower
head
reactor
vessel)
and
the
gas.
Besides,
the
gas
flow
rates
coming
from
the
MCCI
are
also
taken
into
account.
All
these
heat
fluxes
are
then
transferred
to
CPA
which,
in
the
correction
step,
calculates
again
the
cavity
thermal-hydraulics
in
the
same
time
as
the
other
containment
zones,
taking
into
account
the
gas
mass
flow
rates
entering
into
this
zone
(in
particular
the
gaseous
sources
issued
from
the
MCCI
process)
or
going
out
of
this
zone.
For
the
detailed
coupling,
CPA
calculates
the
convective
heat
fluxes
on
the
wall
surfaces,
which
are
then
taken
into
account
by
MEDICIS
in
the
following
time
step.
5.
Tools,
LIBRARIES
and
documentation
for
ASTEC
V2
users
5.1.
MDB
library
The
library
material
data
bank
(MDB)
includes
all
the
recent
research
on
the
nuclear
material
properties
done
in
international
projects:
CIT
FP4
and
ENTHALPY
FP5
projects
for
FP
(Adroguer
et
al.,
1999)
(De
Braemecker
et
al.,
2003)
and
RASPLAV
and
MASCA
OECD
projects
for
corium
(Asmolov,
1998)
(Asmolov
and
Tsurikov,
2004).
The
evaluation
of
corium
properties
is
based
on
the
Euro-
pean
NUCLEA
database
for
corium
thermo-chemistry
(Bakardjieva
et
al.,
2008).
On
a
practical
point
of
view,
the
MDB
library
groups
together
all
material
properties
under
a
unique
simple
readable
format.
This
includes:
-
All
simple
materials
of
a
water-cooled
reactor
(solid,
liquid
and
gas)
and
associated
usual
properties
(enthalpy,
conductivity,
density.
.
.),
-
FP
ideal
chemistry
(equilibrium
reactions),
-
Iodine
chemistry
in
containment
(kinetics),
-
FP
isotopes
(decay
heat
and
transmutation
rates),
-
Complex
materials
such
as
molten
corium.
A
graphic
user
interface
is
provided
allowing
users
to
draw
any
material
property,
updating
them
if
necessary.
In
the
ASTEC
V2.0
version,
MDB
is
only
used
in
the
following
modules:
ELSA
and
SOPHAEROS
for
FP
gas
equilibrium
databanks,
ISODOP
for
isotopes
evolution,
IODE
for
iodine
chemical
reactions,
and
MEDICIS
for
corium
properties.
The
extension
of
its
use
to
other
modules
(CESAR,
ICARE,
RUPUICUV
and
CPA)
is
planned
in
the
next
ASTEC
V2.1
version.
5.2.
User’s
tools
The
computer
interface
is
mostly
based
on
the
SIGAL-ODESSA
package,
which
is
combining
the
SIGAL
tool
(originally
developed
at
IRSN
for
the
ICARE2
mechanistic
code
and
used
then
in
ASTEC
V0
and
ASTEC
V1)
with
the
more
recent
one
ODESSA.
This
advanced
specific
SIGAL-ODESSA
package,
which
consists
of
a
set
of
general
and
portable
libraries,
modules
and
tools,
allows
in
particular
the
automatic
checking
at
the
pre-processing
step
(checking
of
syntax
and
consistency
of
input
data),
the
dynamic
management
of
the
memory
size,
the
on-line
management
and
handling
of
databases
stored
on
files
or
already
located
in
the
memory
(thanks
to
the
so-
called
“analyzer”
data
interpreter),
as
well
as
subsequent
graphic
post-processing.
Moreover,
since
ASTEC
V2.0,
the
coupling
to
the
IRSN
JADE
pre-processing
tool
provides
users
with
a
new
interac-
tive
way
to
build
and
check
ASTEC
input
decks
through
a
dedicated
graphical
user
interface
(GUI).
One
can
note
that
ODESSA
has
mainly
been
designed
for
multi-
threading
when
SIGAL
cannot
support
that.
In
addition,
ODESSA
has
the
following
main
advantages:
-
Much
better
performances
especially
when
storing
new
data
in
a
database,
-
Complete
C
and
C++
interfaces,
-
A
rewritten
analyzer
version
which
removes
many
limitations
and
is
much
faster
than
the
SIGAL
analyzer,
-
New
tools
such
as
for
instance
the
builder
to
build
executable
programmes
written
in
C,
C++
and
Fortran
in
the
same
way
on
Linux
or
Windows,
-
The
possibility
to
use
the
whole
computer
memory.
With
SIGAL,
the
memory
amount
was
necessarily
defined
in
an
include
file.
As
concerns
post-processing,
SIGAL-ODESSA
includes
on-line
visualization
tools
and
the
front-end
GUI
processor
GTIC
that
helps
to
build
graphical
files
with
the
IRSN
TIC
tool,
which
is
an
interactive
curve
plotter.
In
addition,
the
coupling
with
the
GRS
ATLAS
graphical
tool
(Beraha
et
al.,
1999)
allows
off-line
visu-
alization
of
results,
either
in
the
form
of
2D
colour
graphics
of
containment
or
RCS,
or
of
curves
of
evolution
of
variables
in
the
SIGAL
database
along
time.
ATLAS
(with
its
graphical
editor
APG)
allows
the
user
to
set
up
its
own
plant
graphics
with
its
own
nodalisation
scheme
for
dynamic
visualization
of
the
selected
parameters.
The
coupling
of
ASTEC
to
the
IRSN
SUNSET
statistical
tool
(Chevalier-Jabet
et
al.,
2014)
allows
making
easier
the
realization
of
sensitivity
analyses
thanks
to
an
evaluation
of
the
influence
of
uncertainties
on
data
or
models
on
the
code
results.
5.3.
QA
procedures
and
documentation
The
ASTEC
code
is
developed
and
managed
within
a
framework
of
Quality
Assurance
procedures,
resorting
in
particular
to
special
programming
rules
as
well
as
to
a
software
tool
for
version
man-
agement
and
change
control.
In
addition,
the
V2.0
code
version
is
associated
to
a
complete
and
up-to-date
documentation
(gen-
eral
description
of
the
V2.0
version,
theoretical
and
users’
manuals

P.
Chatelard
et
al.
/
Nuclear
Engineering
and
Design
272
(2014)
119–135
131
Fig.
9.
ASTEC
web
portal:
access
to
V2.0rev2p2
documentation
and
download.
for
each
module,
users’
guide
to
use
post-processing
tools,
users’
guidelines
to
perform
plant
applications.
.
.).
A
dedicated
ASTEC
web
site
(to
be
accessed
at
https://www-astec.irsn.fr)
was
created
in
2009
to
allow
registered
users
(i.e.
users
having
rights
to
proceed
though
a
password
access,
in
accordance
to
a
software
agreement
signed
with
either
IRSN
or
GRS)
to
easily
download
any
code
evolution
(new
major
version
or
new
revision
of
an
existing
version)
or
any
documentation
update
(Fig.
9).
For
exchange
of
information
between
users
and
the
main-
tenance
team,
a
specific
tool
MARCUS
was
set
up
in
2004
and
a
new
MARCUS
version
is
now
operating
(to
be
accessed
at
https://www-marcus.irsn.fr).
This
tool
assures
the
survey
and
stor-
age
of
all
request
cards
sent
by
users
to
indicate
anomalies
when
running
the
code,
as
well
as
of
all
the
corresponding
answers
(either
simple
advice
to
improve
input
decks
or
corrective
action
in
the
code
source)
supplied
by
the
maintenance
team
to
overcome
the
problem.
6.
Assessment
of
the
V2.0
version
vs.
experiments
As
it
was
already
the
case
for
ASTEC
V1,
the
ASTEC
V2
validation
is
supported
by
a
large
set
of
French,
German
and
international
experiments
that
cover
most
aspects
of
severe
accident
phenomen-
ology
(Chatelard
et
al.,
2012).
The
validation
matrix,
which
mainly
includes
separate-effect
tests
(SET)
or
coupled-effect
tests
(CET),
also
includes
integral
applications
such
as
the
TMI-2
accident
and
the
integral
experiments
of
the
Phébus.
FP
programme
(Clément
and
Zeyen,
2005),
particularly
the
application
to
the
OECD
ISP
No.
46
on
the
Phébus
FPT1
experiment
(this
application
coupled
all
the
modules
involved
for
the
primary
circuit
and
for
the
containment,
except
DCH
and
MCCI).
The
ASTEC
V2
validation
task
is
a
continuous
process.
Before
its
freezing
and
delivery,
the
ASTEC
V2.0
version
was
assessed
at
IRSN
against
the
25
experiments
belonging
to
the
basic
validation
matrix
which
is
partly
based
on
ISPs
(International
Standard
Prob-
lems
of
OECD)
and
this
regular
testing
has
been
then
repeated
for
both
V2.0-rev1
and
-rev2
versions.
Moreover,
independently
from
this
systematic
checking,
any
new
V2
version
or
revision
is
being
separately
and
deeply
further
assessed
at
IRSN
and
GRS
against
sev-
eral
other
experiments.
Illustrations
of
such
in-depth
validation
of
new
models
generally
performed
by
code
developers
using
each
ASTEC
module
in
a
stand-alone
mode
can
be
found
for
instance
in
(Carénini
and
Fleurot,
2014)
for
corium
behaviour
in
the
lower
head,
in
(Brillant
et
al.,
2013b)
for
FP
release
phenomena,
in
(Bentaïb
and
Bleyer,
2009),
(Bentaïb
and
Anatasova,
2010)
for
CPA
thermal-
hydraulics,
in
(Spitz
et
al.,
2001)
for
CPA
aerosol
behaviour,
in
(Cranga
et
al.,
2008)
and
(Cranga
et
al.,
2010)
for
MCCI
processes.
Moreover,
an
independent
V2.0
validation
work
has
been
initi-
ated
end
of
2009
in
the
particular
frame
of
SARNET
network
and
it

132
P.
Chatelard
et
al.
/
Nuclear
Engineering
and
Design
272
(2014)
119–135
has
been
then
intensively
continued
over
the
3
last
years
using
successively
the
V2.0-rev1
and
-rev2
versions,
thus
allowing
in
particular
to
reassess
ASTEC
V2
versus
most
of
the
experiments
which
had
been
yet
calculated
in
2008
with
the
latest
ASTEC
V1
version.
Several
illustrations
of
such
a
validation
cooperative
work
are
provided
in
(Chatelard
et
al.,
2012),
along
with
a
synthesis
of
the
overall
ASTEC
V2.0
assessment
including
numerous
references
to
the
abovementioned
in-depth
ASTEC
V2
validation
by
IRSN
and
GRS.
7.
Plant
applications
domain
covered
by
the
V2.0
version
Focusing
on
its
general
capabilities
with
respect
to
full
scale
analyses,
the
main
progress
of
ASTEC
V2
compared
to
ASTEC
V1
is
its
ability
to
extend
the
safety
analyses
to
EPR
(thanks
in
particu-
lar
to
an
adequate
ex-vessel
modelling),
as
well
as
the
possibility
to
cover
now
the
simulation
of
accidents
in
reactor
shutdown
states.
So,
the
ASTEC
V2.0
allows
particularly
complete
calculations
up
to
iodine
behaviour
in
containment
of
various
plant
severe
accident
sequences
on
several
Gen.
II
and
Gen.
III
reactor
types
(such
as
for
instance
PWR-900,
PWR-1300,
KONVOI-1300,
VVER-440,
VVER-
1000,
and
EPR
including
a
detailed
modelling
of
the
core
catcher)
operating
at
full
power,
such
as
station
black-out,
loss
of
steam
generator
feed-water,
steam
generator
tube
rupture
(possibly
com-
bined
with
a
main
steam
line
break),
as
well
as
small,
medium
and
large
break
loss
of
coolant
accidents,
or
in
shutdown
states.
In
particular,
as
to
benchmarks
on
plant
applications,
several
com-
parisons
of
the
SA
front-end
phase
for
several
French
PWR
plant
designs
and
notably
EPR
are
currently
done
at
IRSN
between
ASTEC
V2.0
and
the
French
reference
thermal-hydraulic
code
CATHARE2
(Camous
et
al.,
2005).
Such
work
confirms
the
good
results
that
were
obtained
in
the
ASTEC
V1.3-to-CATHARE
benchmarking
activ-
ity
in
the
frame
of
the
IRSN
PSA2
applied
to
a
French
PWR
1300
MWe
by
(Trégourès
et
al.,
2009).
One
can
underline
that
such
a
In
the
same
way,
the
V2.0-rev1
version
was
successfully
used
at
GRS
to
achieve
in
2011
the
consolidation
with
ASTEC
of
a
former
PSA2
on
a
German
Konvoi
PWR
1300
MWe
performed
few
years
ago
with
MELCOR
through
benchmarks
between
both
codes
(Application
of
ASTEC
V2.0
to
severe
accident
analyses
for
German
KONVOI
type
reactors,
Nowack
et
al.,
Proceedings
NURETH
14,
Toronto,
24–30,
September,
2011).
Moreover,
as
for
the
ASTEC
validation
vs.
experimental
data
(see
Section
6),
an
independent
V2.0
benchmarking
activity
has
been
also
continuously
carried
out
by
several
partners
in
the
particular
frame
of
the
SARNET
FP7
work-package
No.
4
using
successively
the
V2.0-rev1
and
-rev2
versions,
thus
allowing
to
cover
a
wide
matrix
of
plant
applications
(several
different
SA
sequences
applied
on
PWRs
and
VVERs):
in
general,
the
main
trends
of
results
were
similar
to
MELCOR
or
MAAP4
results
(despite
some
quantitative
differences
due
to
modelling
differences,
e.g.
on
core
degradation),
as
illustrated
for
instance
in
(Lombard
et
al.,
2014).
In
addition,
to
support
preliminary
analyses
of
the
mitigation
strategy
relying
on
IVR
concept
(which
has
been
adopted
for
some
advanced
Gen.
III
PWRs,
such
as
AP600,
AP1000,
VVER-1000
AES-
92,
but
also
for
older
VVER-440/V213),
it
is
possible
to
simulate
any
vessel
external
cooling
circuit
either
thanks
to
a
chaining
of
the
CESAR
and
ICARE
modules
in
the
lower
head
region
or
to
more
sim-
ply
perform
ICARE
stand-alone
applications
with
adequate
external
boundary
conditions,
as
illustrated
for
instance
in
(Matejovic
et
al.,
2014)
and
(Zvonarev
et
al.,
2014).
For
CANDU
reactors
and
more
generally
Pressurized
Heavy
Water
Reactors
(PHWR),
most
models
are
already
applicable,
mainly
for
fission
products
and
containment
phenomena,
while
necessary
work
is
on-going
for
model
adaptation
to
core
degra-
dation
due
to
the
horizontal
core
geometry
(see
Section
8).
For
BWR
reactors,
the
situation
is
similar:
while
the
ASTEC
appli-
cability
to
BWR
containments
had
already
been
shown
by
GRS
few
years
ago
(Van
Dorsselaere
et
al.,
2008),
current
efforts
are
put
on
the
modelling
of
core
degradation,
along
with
some
adaptation
of
core
thermal-hydraulics.
With
ongoing
development
of
a
more
suitable
ICARE-CESAR
coupling
(see
Section
8),
considerable
efforts
are
focused
in
priority
on
model
adaptation
to
the
core
degradation
(Van
Dorsselaere
et
al.,
2008).
Waiting
for
such
in-vessel
modelling
improvements
(which
will
only
become
available
in
the
next
ASTEC
V2.1
version),
preliminary
ASTEC
simulations
of
the
Fukushima-
Daiichi
accidents
could
be
however
successfully
carried
out
at
IRSN
with
the
V2.0-rev2
version,
assuming
some
simplifications
of
the
core
modelling
description
(Bonneville
and
Luciani,
2012).
Finally,
as
for
severe
accident
management
(SAM)
measures,
most
safety
systems
for
the
existing
PWR,
VVER
and
BWR
can
be
yet
represented:
volunteer
primary
circuit
depressurisation,
steam
generator
management,
spray
system
and
venting
in
the
contain-
ment.
One
can
however
point
out
that
the
peculiar
conditions
of
late
quenching
will
be
only
addressed
by
the
future
V2.1
major
version
(see
Section
8).
8.
Development
of
the
future
V2.1
major
version
The
ASTEC
models
are
today
close
to
the
state
of
the
art
(notably
with
respect
to
source
term
evaluation),
but
several
key
safety
issues
still
remain
open,
such
as
in
particular
in-vessel
and
ex-
vessel
corium
coolability
as
well
as
kinetics
of
iodine
and
ruthenium
chemistry
in
the
circuits.
So,
according
to
the
progressive
under-
standing
of
new
experimental
data,
the
integration
into
ASTEC
of
physical
models
will
go
on
in
priority
in
those
fields.
In
particu-
lar,
in
accordance
to
one
of
the
main
outcomes
from
the
extended
ASTEC
V2
assessment
(Chatelard
et
al.,
2012),
modelling
issues
on
the
reflooding
of
severely
damaged
cores
constitute
for
ASTEC,
as
for
most
other
SA
codes,
one
of
the
main
challenges
for
the
future
major
code
version,
to
be
addressed
using
in
particular
the
future
data
to
be
produced
by
the
IRSN
PEARL
experimental
programme
(Repetto
et
al.,
2011).
So,
while
the
current
V2.0
version
is
still
being
continuously
improved
(through
the
elaboration
and
periodical
release
of
suc-
cessive
revisions)
in
order
to
keep
in
particular
the
FP
and
MCCI
models
at
the
state
of
the
art,
the
preparation
of
the
second
major
ASTEC
V2
version
(to
be
identified
as
V2.1)
has
started
yet
to
provide
an
adequate
frame
to
welcome
future
quenching
models
under
development
at
IRSN
(Bachrata
and
Fichot,
2012).
These
efforts
will
progressively
intensify
at
IRSN
and
GRS
in
2013
up
to
its
freezing
and
delivery,
which
is
expected
mid-2014.
Beyond
the
development
of
an
adequate
model
for
the
reflood-
ing
of
a
degraded
core,
the
main
other
modelling
improvements
to
be
implemented
in
the
V2.1
version
will
concern
complete
kinetics
of
chemical
reactions
in
RCS
gaseous
phase,
an
exten-
sion
of
the
MDB
library
to
all
ASTEC
modules
(CESAR,
ICARE,
CPA,
RUPUICUV),
the
extension
of
the
ICARE-MEDICIS
coupling
to
the
thermal
behaviour
of
the
lower
head
vessel
wall
after
its
first
mechanical
failure,
a
revised
modelling
of
DCH
phenomena,
a
more
relevant
modelling
of
pool-scrubbing
phenomena
in
the
contain-
ment
and
a
series
of
ASTEC
structuring
adaptations
to
BWRs
and
CANDUs
(mainly
in
the
ICARE
module
through
the
development
of
multi-channel
components
in
each
core
ring
and
dedicated
mod-
elling
of
specific
core
components
such
as
respectively
canisters
or
pressure
and
calandria
tubes),
the
latter
benefiting
from
the
new
CESAR/ICARE
coupling
technique
described
hereafter.
Indeed,
on
a
code
structure
point
of
view,
one
significant
V2.1
evolution
will
concern
the
modification
of
the
CESAR/ICARE
cou-
pling
technique,
assigning
more
distinct
roles
for
the
two
modules:
CESAR
will
deal
with
the
thermal-hydraulics
in
the
whole
RCS

P.
Chatelard
et
al.
/
Nuclear
Engineering
and
Design
272
(2014)
119–135
133
(vessel
and
loops)
during
the
whole
transient
while
ICARE
will
deal
with
core
degradation,
i.e.
material
relocation
since
the
beginning
of
the
transient,
thus
allowing
to
definitely
suppress
the
CESAR-to-
ICARE
switch
at
the
end
of
the
front-end
phase
which
was
proved
to
have
some
drawbacks
(feed-back
from
ASTEC
V1
applications).
The
use
of
a
unique
thermal-hydraulics
in
the
whole
RCS
will
also
make
possible
the
simulation
of
SA
scenarios
with
in-core
late
water
injection
(which
are
possibly
involving
a
partial
or
total
refilling
of
the
RCS
loops
with
water)
up
to
a
stable
core
configuration,
which
was
not
possible
with
the
V2.0
version.
Besides,
a
2D
extension
of
CESAR
is
simultaneously
being
developed
at
IRSN
to
support
a
radial
discretisation
of
the
core
region
as
in
ICARE,
thus
allowing
to
take
into
account
in-core
2D
two-phase
flow
patterns.
Moreover,
thanks
to
the
combination
of
this
new
CESAR/ICARE
coupling
and
the
transfer
in
ASTEC
of
the
existing
ICARE
Zry/air
oxidation
model
which
was
developed
few
years
ago
in
the
IRSN
ICARE/CATHARE
code
(Coindreau
et
al.,
2010),
the
V2.1
version
will
allow
to
entirely
simulate
the
situations
of
air
ingress
(along
with
a
complete
ruthenium
behaviour
in
the
RCS),
either
after
vessel
lower
head
rupture
or
for
mid-loop
states.
In
addition,
the
V2.1
version
will
also
allow
making
more
relevant
analyses
of
spent
fuel
pool
accidents
since
this
account
for
air/steam
flows
within
fuel
rod
assemblies
could
be
now
combined
with
an
adequate
modelling
in
the
ICARE
module
of
special
canister
components.
Moreover,
efforts
will
be
paid
to
further
consolidate
the
ASTEC
capability
in
supporting
relevant
analyses
about
the
IVR
mitigation
strategy.
In
addition,
the
feedback
from
the
interpretation
of
the
current
experimental
programmes
performed
in
the
international
frame
will
be
of
course
continuously
taken
into
account:
iodine
and
ruthe-
nium
chemistry
in
RCS
and
in
containment
(respectively
CHIP,
EPICUR
and
STEM
at
IRSN
and
BIP-2
OECD
project
at
AECL
(for
the
latter,
see
Glowa
and
Moore,
2011));
effect
of
high
fuel
burn-up,
of
MOX
fuel
and
air
ingress
on
core
degradation
and
FP
release
(VER-
DON
at
CEA
(Ducros
et
al.,
2009));
corium
behaviour
in
the
lower
head
(CORDEB
project
that
is
starting
end
of
2012
at
NITI
in
Russia),
hydrogen
distribution,
combustion
and
recombination
(THAI-2
at
Becker
Technology,
(Sonnenkalb
and
Poss,
2012));
heat
flux
spatial
distribution
in
the
corium
pool
during
MCCI
(VULCANO
at
CEA,
CCI
at
ANL);.
.
.
As
concerns
more
particularly
new
ICARE
models
for
BWR
core
degradation,
they
will
be
assessed
vs.
both
experimental
data
(such
as
for
instance
selected
CORA
tests)
and
real
plant
data.
Indeed,
for
the
latter,
the
IRSN
on-going
V2.0
preliminary
simulations
of
the
Fukushima-Daiichi
accidents
(see
Section
7)
will
be
of
course
updated
using
the
new
V2.1
version.
As
concerns
the
ASTEC
adaptation
for
other
types
of
applica-
tions,
both
the
modelling
and
assessment
activities
will
go
on,
covering
in
particular:
-
Necessary
extension
to
Gen.
IV
reactors,
especially
SFR
in
the
frame
of
the
JASMIN
FP7
project
(Girault
et
al.,
2013),
coordinated
by
IRSN,
that
started
in
Dec.2011
for
4
years.
For
instance
the
pos-
sibility
to
deal
with
sodium
coolant
transport
in
the
RCS
has
been
yet
implemented
in
the
version
V2.1
under
development
at
IRSN
and
the
integration
of
a
0D
neutronics
model
is
underway.
-
Adaptation
to
accidents
of
air
or
water
ingress
in
the
vacuum
ves-
sel
of
the
ITER
Fusion
installation
(Van
Dorsselaere
et
al.,
2009b).
For
instance
the
SOPHAEROS
module
has
been
yet
extended
to
large
volumes
to
allow
accounting
for
chemical
speciation
in
the
containment.
-
Use
of
some
ASTEC
modules
for
the
IRSN
emergency
response
tools
(Foucher
et
al.,
2014).
Finally,
as
concerns
user
tools
and
user-friendliness,
efforts
will
continue
to
make
easier
both
the
pre-processing
and
post-
processing
of
the
code
applications,
as
well
as
the
use
of
the
ASTEC/SUNSET
coupling
to
evaluate
the
influence
of
uncertainties
on
data
or
models
on
the
simulation
results.
In
addition,
another
important
issue
is
to
extend
the
capabilities
of
ASTEC
from
progno-
sis
to
diagnosis
of
accidental
situations,
in
strong
link
in
particular
with
the
on-going
Fukushima-Daiichi
accidents
analyses
(Cranga
et
al.,
2012).
ASTEC
will
be
indeed
extended
towards
a
“diagno-
sis”
version,
first
by
interfacing
with
atmospheric
dispersion
codes
in
order
to
enhance
capabilities
of
direct
comparison
with
on-site
measurements,
while
a
methodology
based
on
Bayesian
networks
will
be
investigated
in
parallel
for
evaluating
the
probability
of
the
different
possible
accident
scenarios
from
the
uncertain
informa-
tion
provided
by
the
plant
instrumentation.
9.
Conclusion
Considerable
efforts
are
being
continuously
made
by
IRSN
and
GRS
on
the
development
of
the
source
term
evaluation
code
ASTEC,
ranging
from
the
initial
stage
of
core
uncovery
to
the
final
sit-
uation
of
long-term
corium
stabilization,
long-term
containment
integrity,
and
fission
product
retention
or
release
to
environment.
In
the
frame
of
the
SARNET
European
network,
jointly-executed
research
activities
have
been
performed
with
the
ultimate
objec-
tive
of
providing
the
best
physical
models
for
integration
into
ASTEC
and
make
the
code
the
European
reference.
This
effort
will
then
be
pursued
beyond
SARNET
in
the
frame
of
the
CESAM
FP7
new
project
(code
for
European
severe
accident
management:
see
(CESAM,
2013)),
coordinated
by
GRS
and
starting
most
probably
in
April
2013
with
4
years
duration.
It
will
aim
at
improving
the
code
capabilities
to
simulate
the
main
SAM
measures
important
for
the
European
NPPs,
accounting
also
for
the
feedback
of
the
Fukushima-
Daiichi
accidents.
A
strong
link
will
be
maintained
with
several
other
international
R&D
cooperative
actions
such
as
for
instance
the
PASSAM
FP7
project
(passive
and
active
systems
on
severe
accident
source
term
mitigation:
see
(Albiol
et
al.,
2012)),
coordinated
by
IRSN
and
starting
in
January
2013
for
4
years,
where
experimental
and
theoretical
works
will
be
performed
on
iodine
and
ruthenium
filtration,
studying
various
filtration
concepts
(pool
scrubbing,
solid
filtration
on
existing
and
innovative
materials
and
systems).
Most
ASTEC
models
are
today
close
to
the
state
of
the
art.
The
main
advantages
of
the
code
are
the
high
quality
and
validation
level
of
its
physical
models,
in
particular
fission
product
models
which
are
essential
for
source
term
evaluation.
Nevertheless,
since
R&D
still
progresses,
key
model
improvements
have
already
been
identified
for
the
next
V2
versions,
consistently
with
SARNET
R&D
priorities,
in
particular
in-vessel
and
ex-vessel
corium
coolability.
In
accordance,
the
main
on-going
modelling
efforts
are
spent
in
pri-
ority
at
IRSN
and
GRS
on
the
reflooding
of
degraded
cores,
on
MCCI
(in
particular
on
the
coolability
aspects)
as
well
as
on
kinetics
of
iodine
and
ruthenium
chemistry
in
the
circuits,
on
pool-scrubbing
phenomena
in
the
containment
and
in
lower
priority
on
DCH.
In
any
case,
ASTEC
will
remain
a
repository
of
knowledge
gained
from
international
R&D
since
the
feedback
from
the
interpretation
of
cur-
rent
and
future
experimental
programmes
(CHIP,
VERDON,
PEARL,
CORDEB,
THAI,
STEM,
CCI,
.
.
.)
performed
in
the
international
frame
will
be
continuously
taken
into
account.
In
addition,
a
new
CESAR/ICARE
coupling
methodology
has
been
developed
at
IRSN
to
remove
in
the
next
V2.1
major
version
some
shortcomings
of
the
current
V2.0
version.
The
V2.1
version
will
allow
after
mid-2014
both
to
simulate
the
situations
of
late
core
reflooding
and
to
deal
with
the
situations
of
air
ingress
either
after
vessel
lower
head
rupture
or
during
shutdown
or
mid-loop
states.
Moreover,
this
future
V2.1
version,
which
will
constitute
an
inter-
national
reference
version
integrating
a
large
part
of
the
knowledge
generated
by
SARNET
and
ISTP,
will
also
allow
relevant
simulations
of
BWR
and
CANDU
severe
accident
sequences
as
well
as
spent
fuel

134
P.
Chatelard
et
al.
/
Nuclear
Engineering
and
Design
272
(2014)
119–135
pool
accidents
thanks
in
particular
to
a
dedicated
description
of
specific
core
components
such
as
multi-channels,
canisters.
In
addition,
ASTEC
will
be
soon
extended
towards
a
“diagno-
sis”
version
by
interfacing
with
atmospheric
dispersion
codes
in
order
to
enhance
capabilities
of
direct
comparison
with
on-site
measurement.
Other
mean-term
objectives
are
on
the
one
hand
to
continue
the
on-going
extension
of
the
scope
of
application
of
the
ASTEC
code
to
Gen.
IV
SFR
reactors
and
to
accidents
in
the
ITER
Fusion
installa-
tion,
and
on
the
other
hand
the
use
for
emergency
response
tools,
while
making
ASTEC
evolving
towards
a
severe
accident
simulator
constitutes
the
main
long-term
objective.
References
Adroguer,
B.,
Barrachin,
M.,
Bottomley,
P.,
Hofmann,
P.,
Miassoeodov,
A.,
Stuckert,
J.,
Chevalier,
P.Y.,
Cheynet,
B.,
Fischer,
M.,
Hellmann,
S.,
et
al.,
1999.
Corium
Interac-
tion
and
Thermochemistry
(CIT).
In:
FISA-99
Symposium:
EU
Research
in
Reactor
Safety,
November
29–December
1,
Luxembourg.
Albiol,
T.,
Herranz,
L.E.,
Riera,
E.,
Guieu,
S.,
Lind,
T.,
Manzini,
G.,
Auvinen,
A.,
Losch,
N.,
2012.
New
studies
on
passive
and
active
systems
towards
enhanced
severe
accident
source
term
mitigation
–
The
PASSAM
project.
In:
Poster
EUROSAFE,
November
5–6,
Brussels,
Belgium.
Allelein,
H.-J.,
Arndt,
S.,
Klein-Heßling,
W.,
Schwarz,
S.,
Spengler,
C.,
Weber,
G.,
2008.
COCOSYS:
status
of
development
and
validation
of
the
German
containment
code
system.
Nuclear
Engineering
and
Design
283
(4),
872–889.
Asmolov,
V.,
1998.
Last
findings
of
RASPLAV
project.
In:
Workshop
on
in-vessel
core
debris
retention
and
coolability,
March
3–6,
Garching,
Germany.
Asmolov,
V.G.,
Tsurikov,
D.F.,
2004.
MASCA
project:
major
activities
and
results.
In:
MASCA
seminar,
June
10–11,
Aix-en-Provence,
France.
Bachrata,
A.,
Fichot,
F.,
2012.
Code
simulation
of
quenching
of
a
high
temperature
debris
bed:
model
improvement
and
validation
with
experimental
results.
In:
ICONE-20
–
POWER-2012,
July
30–August
3,
Anaheim,
CA,
USA.
Bandini,
G.,
Buck,
M.,
Hering,
W.,
Godin-Jacqmin,
L.,
Ratel,
G.,
Matejovic,
P.,
Barnak,
M.,
Paitz,
G.,
Stefanova,
A.,
Trégourès,
N.,
Guillard,
G.,
Koundy,
V.,
2010.
Recent
advances
in
ASTEC
validation
on
circuit
thermal-hydraulic
and
core
degradation.
Progress
in
Nuclear
Energy
52,
148–157.
Bakardjieva,
S.,
Barrachin,
M.,
Bechta,
S.,
Bottomley,
D.,
Brissoneau,
L.,
Cheynet,
B.,
Fischer,
E.,
Journeau,
C.,
Kiselova,
M.,
Mesentseva,
L.,
Piluso,
P.,
Wiss,
T.,
2008.
Improvement
of
the
European
thermodynamic
database
NUCLEA
in
the
frame
of
UE-funded
experiment.
In:
3rd
European
Review
Meeting
on
Severe
Accident
Research
(ERMSAR
2008),
September
23–25,
Nesseber,
Bulgaria.
Becquin,
G.,
Maas,
L.,
2010.
CESAR
calculations
on
two
LOFA
transients
conducted
in
HE-FUS3
helium
facility.
In:
5th
HTR
Conference
(High
Temperature
Reactor
Technology),
October
18–20,
Prague,
Czech
Republic.
Bentaïb,
A.,
Bleyer,
A.,
2009.
ASTEC
validation
on
experiments
from
the
PANDA
SETH
benchmark.
In:
NURETH-13
Conference,
September
27–October
2,
Kanazawa
city,
Japan.
Bentaïb,
A.,
Anatasova,
B.,
2010.
ASTEC
Validation
on
PANDA
tests.
In:
Nuclear
Power
for
the
People,
September
26–29,
Nesseber,
Bulgaria.
Beraha,
D.,
et
al.,
1999.
ATLAS:
applications
experiences
and
further
develop-
ments.
Second
OECD
specialist
meeting
on
simulators
and
plant
analyzers.
In:
Proceedings
of
Symposium
1997,
Espoo,
VTT-SYMP-194,
in
NEA/CSNI/R(97)37.
Blanchat
T.K.,
Pilch
M.M.,
Lee
R.Y.,
Meyer
L.,
Petit
M.,
1999.
Report
NUREG/CR-5746,
SAND99-1634.
Direct
containment
heating
experiments
at
low
reactor
coolant
system
pressure
in
the
SURTSEY
test
facility.
Bonneville,
H.,
Luciani,
A.,
2012.
Simulation
of
the
core-degradation
phase
of
the
Fukushima
accidents
using
the
ASTEC
code.
In:
International
ANS
Meet-
ing
on
Severe
Accident
Assessment
and
Management:
Lessons
Learned
from
Fukushima
Daiichi,
November
12–15,
San
Diego,
CA,
USA.
Bosland,
L.,
Cantrel,
L.,
Girault,
N.,
Clément,
B.,
2010.
Modelling
of
iodine
radiochem-
istry
in
the
ASTEC
severe
accident
code:
description
and
application
to
FPT-2
PHEBUS
test.
Nuclear
Technology
171,
88–107.
Brähler,
T.,
Koch,
M.K.,
2011.
Simulation
of
the
THAI
HD-12
and
HD-22
tests
with
the
flame
front
model
of
ASTEC
using
different
correlations
for
the
turbulent
burn-
ing
velocity.
In:
14th
International
Topical
Meeting
on
Nuclear
Reactor
Thermal
Hydraulics
(NURETH-14),
September
25–30,
Toronto,
Canada.
Brillant,
G.,
Marchetto,
C.,
Plumecocq,
W.,
2013a.
Fission
product
release
from
nuclear
fuel
I.
Physical
modelling
in
the
ASTEC
code.
Annals
of
Nuclear
Energy,
88–95.
Brillant,
G.,
Marchetto,
C.,
Plumecocq,
W.,
2013b.
Fission
product
release
from
nuclear
fuel
II.
Validation
of
ASTEC/ELSA
on
small
and
large
scale
experiments.
Annals
of
Nuclear
Energy
61,
96–101.
Camous,
F.,
Chatelard,
P.,
Schwarz,
A.-V.,
Freitas,
R.,
Sabotinov,
L.,
Messer,
N.,
2005.
Proceedings
of
NURETH-11
Conference,
October
2–6,
Avignon,
France.
Safety
assessment,
code
validation
and
R&D
studies
at
IRSN
using
the
best-estimate
CATHARE2
code.
Carénini,
L.,
Fleurot,
J.,
Fichot,
F.,
2014.
Validation
of
ASTEC
V2
models
for
the
behaviour
of
corium
in
the
vessel
lower
head.
Nuclear
Engineering
and
Design
272,
152–162.
CESAM
(Code
for
European
Severe
Accident
Management),
2013.
FP7
Fission
2012-
2.1.2
Collaborative
Project
No.
323264.
Theme:
Nuclear
Fission,
Safety
and
Radiation
Protection.
Chatelard,
P.,
Fleurot,
F.,
Marchand,
O.,
Drai,
P.,
2006.
ICONE-14
conference,
July
17–20,
Miami,
FL,
USA.
Assessment
of
ICARE/CATHARE
V1
severe
accident
code.
Chatelard,
P.,
Arndt,
S.,
Atanasova,
B.,
Bandini,
G.,
Bleyer,
A.,
Brähler,
T.,
Buck,
M.,
Kljenak,
I.,
Kujal,
B.,
2012.
5th
European
Review
Meeting
on
Severe
Accident
Research
(ERMSAR
2012),
March
21–23,
Cologne,
Germany.
Overview
of
the
ASTEC
V2.
0
and
V2.
0-rev1
validation.
Cheynet,
B.,
Chevalier,
P.Y.,
Fischer,
E.,
2002.
Thermosuite.
CALPHAD
Computer
Cou-
pling
of
Phase
Diagrams
and
Thermochemistry
26
(2),
167–174.
Chevalier-Jabet,
K.,
Cousin,
F.,
Cantrel,
L.,
Seropian,
C.,
2014.
Source
Term
assess-
ment
with
ASTEC
and
associated
uncertainty
analysis
using
SUNSET.
Nuclear
Engineering
and
Design
272,
207–218.
Clément,
B.,
Zeyen,
R.,
2005.
The
Phébus
FP
and
International
Source
Term
Pro-
grammes.
In:
Proceedings
of
the
International
Conference
on
Nuclear
Energy
for
New
Europe,
September
5–8,
Bled,
Slovenia.
Clément,
B.,
Simondi-Teisseire,
B.,
2010.
An
IRSN
project
on
source
term
evaluation
and
mitigation.
Transactions
of
the
American
Nuclear
Society
103,
475–476,
ISSN
0003-018X.
Coindreau,
O.,
Ederli,
S.,
Duriez,
C.,
2010.
Air
oxidation
of
zircaloy-4
in
the
600–1000 ◦C
temperature
range:
modelling
for
ASTEC
code
application.
Journal
of
Nuclear
Materials
405,
207–215.
Cousin,
F.,
Dieschbourg,
K.,
Jacq,
F.,
2008.
New
capabilities
of
simulating
fission
prod-
uct
transport
in
circuits
with
ASTEC/SOPHAEROS
v1.3.
Nuclear
Engineering
and
Design
238,
2430–2438.
Cranga,
M.,
Fabianelli,
R.,
Jacq,
F.,
Barrachin,
M.,
Duval,
F.,
2005.
The
MEDICIS
code,
a
versatile
tool
for
MCCI
modelling.
In:
ICAPP’05,
May
15–19,
Seoul,
Korea.
Cranga,
M.,
Mun,
C.,
Michel,
B.,
Duval,
F.,
Barrachin,
M.,
2008.
Interpretation
of
real
material
2D
MCCI
experiments
in
homogeneous
oxidic
pool
with
the
ASTEC/MEDICIS
code.
In:
ICAPP’08,
June
8–12,
Anaheim,
CA,
USA.
Cranga,
M.,
Michel,
B.,
Mun,
C.,
Marchetto,
C.,
2010.
MCCI
in
an
homogeneous
pool:
lesson
learnt
from
MCCI-OECD
and
VULCANO
real
material
experiments
in
dry
conditions.
In:
MCCI-OECD
seminar,
November
15–17,
Cadarache,
St-Paul-lez-
Durance,
France.
Cranga,
M.,
Chevalier-Jabet,
K.,
Marchetto,
C.,
Mun,
C.,
2012.
Evaluations
of
MCCI
risks
for
the
Fukushima
events:
related
IRSN
R&D
strategy
on
corium
retention
and
coolability.
In:
International
ANS
Meeting
on
Severe
Accident
Assessment
and
Management:
Lessons
Learned
from
Fukushima,
November
12–15,
San
Diego,
California,
USA.
De
Braemecker,
A.,
Barrachin,
M.,
Jacq,
F.,
Defoort,
F.,
Mignanelli,
M.,
Chevalier,
P.Y.,
Cheynet,
B.,
Hellmann,
S.,
Funke,
F.,
Journeau,
C.,
Piluso,
P.,
Marguet,
S.,
Hözer,
Z.,
Vrtilkova,
V.,
Belovsky,
L.,
Sannen,
L.,
Verwerft,
M.,
Duvigneaud,
P.H.,
Mwamba,
K.,
Bouchama,
H.,
Ronneau,
C.,
2003.
European
Nuclear
Thermody-
namic
Database
Validated
and
Applicable
in
Severe
Accident
Codes
(ENTHALPY).
In:
FISA-2003
Symposium:
EU
Research
in
Reactor
Safety,
November
10–13,
Luxembourg.
Dickinson,
S.,
Sims,
H.E.,
Guentay,
S.,
Bruchertseifer,
H.,
Liljenzin,
J.O.,
Glänneskog,
H.,
Kissane,
M.P.,
Cantrel,
L.,
Krausmann,
E.,
Rydl,
A.,
2003.
Iodine
Chemistry
and
Mitigation
Mechanism
(ICHEMM).
In:
FISA-2003
Symposium:
EU
Research
in
Reactor
Safety,
November
10–13,
Luxembourg.
Ducret,
D.,
Billarand,
Y.,
Roblot,
D.,
Vendel,
J.,
1996.
Study
on
collection
efficiency
of
fission
products
by
spray:
experimental
device
and
modelling.
In:
24th
DOE/NRC
nuclear
air
cleaning
and
treatment
conference,
July
15–18,
Portland,
OR,
USA.
Ducros,
G.,
et
al.,
2009.
Use
of
gamma
spectrometry
for
measuring
fission
product
releases
during
a
simulated
PWR
severe
accident:
Application
to
the
VERDON
experimental
program.
In:
1st
International
Conference
on
Advancements
in
Nuclear
Instrumentation
Measurement
Methods
and
their
Applications
(ANI-
MMA),
June
7–10,
Marseille,
France.
Farmer,
M.T.,
2001.
Modelling
of
ex-vessel
corium
coolability
with
the
CORQUENCH
code.
In:
ICONE-9
conference,
April
8–12,
Nice,
France.
Farmer
M.T.
et
al.,
2006.
OECD-MCCI
project:
2-D
Concrete
Interaction
(CCI)
Tests,
Final
report.
OECD-MCCI
-2005-TR05.
Fink
J.K.,
Thompson
D.H.,
Spencer
B.W.,
Sehgal
B.R.,
1992.
Aerosols
released
during
large-scale
integral
MCCI
tests
in
the
ACE
program.
Argonne
National
Laboratory
Report
ANL/CP-74552.
Fontanet,
J.,
Herranz,
L.E.,
Ramlakan,
A.,
Naicker,
L.,
2008.
Proceedings
of
the
4th
International
Topical
Meeting
on
High
Temperature
Reactor
Technology,
HTR2008,
September
28–October
1,
Washington,
DC,
USA.
PBMR
confinement
analysis
during
helium
pressure
boundary
breaks.
Foucher,
L.,
Cousin,
F.,
Fleurot,
J.,
Brethes,
S.,
2014.
Assessment
on
900-1300
MWe
PWRs
of
the
ASTEC-based
simulation
tool
of
SGTR
thermalhydraulics
for
the
IRSN
emergency
technical
centre.
Nuclear
Engineering
and
Design
272,
287–298.
Gelbard,
F.,
1982.
MAEROS
User
Manual.
Report
NUREG/CR-1391.
Giordano,
P.,
Auvinen,
A.,
Brillant,
G.,
Colombani,
J.,
Davidovich,
N.,
Dickson,
R.,
Haste,
T.,
Kärkelä,
T.,
Lamy,
J.S.,
Mun,
C.,
Ohai,
D.,
Pontillon,
Y.,
Steinbrück,
M.,
Ver,
N.,
2010.
Recent
advances
in
understanding
ruthenium
behaviour
under
air-
ingress
conditions
during
a
PWR
severe
accident.
Progress
in
Nuclear
Energy
52,
109–119.
Girault,
N.,
Van
Dorsselaere,
J.P.,
Bandini,
G.,
Buck,
M.,
Champigny,
J.,
Hering,
W.,
Herranz,
L.,
Raison,
P.,
Reinke,
N.,
Tucek,
K.,
Verwaerde,
D.,
2013.
The
European
JASMIN
project
for
the
development
of
a
new
safety
simulation
code,
ASTEC-Na,
for
Na-cooled
fast
neutron
reactors.
In:
Proceedings
of
ICAPP
2013,
April
14–18,
Jeju
Island,
Korea.
Glowa
G.,
Moore
C.,
2011.
Behaviour
of
Iodine
Project,
Final
Summary
Report.
NEA/CSNI/R(2011)11.

P.
Chatelard
et
al.
/
Nuclear
Engineering
and
Design
272
(2014)
119–135
135
Grégoire,
A.C.,
Haste,
T.,
2012.
FP
release
during
the
PHEBUS
FP
tests.
In:
Final
seminar
of
the
Phebus
FP
programme,
June
13–15,
Aix-en-Provence,
France.
Grégoire,
A.C.,
Mutelle,
H.,
2012.
Experimental
study
of
the
[B,
Cs,
I,
O,
H]
and
[Mo,
Cs,
I,
O,
H]
systems
in
the
primary
circuit
of
a
PWR
in
conditions
representative
of
a
severe
accident.
In:
Nuclear
Energy
for
New
Europe
2012
(NENE),
September
5–7,
Ljubljana,
Slovenia.
Haste,
T.,
Auvinen,
A.,
Colombani,
J.,
Funke,
F.,
Glowa,
G.,
Güntay,
S.,
Holm,
J.,
Kärkela,
T.,
Langrock,
G.,
Poss,
G.,
Simondi-Teisseire,
B.,
Tietze,
S.,
Weber,
G.,
2012.
Con-
tainment
iodine
experiments
in
the
SARNET2
project.
In:
5th
European
Review
Meeting
on
Severe
Accident
Research
(ERMSAR
2012),
March
21–23,
Cologne,
Germany.
Journeau,
C.,
Piluso,
P.,
Haquet,
J.F.,
Boccaccio,
E.,
Saldo,
V.,
Bonnet,
J.M.,
Malaval,
S.,
Carénini,
L.,
Brissonneau,
L.,
2009.
Two-dimensional
interaction
of
oxidic
corium
with
concretes:
the
VULCANO
VB
test
series.
Annals
of
Nuclear
Energy
36,
1597–1613.
Kim,
S.B.,
Park,
R.J.,
Kim,
H.D.,
Chevallier,
C.,
Petit,
M.,
1999.
Reactor
cavity
dis-
persal
experiments
with
simulant
at
intermediate
system
pressure.
In:
15th
International
Conference
on
SMIRT,
August
15–20,
Seoul,
Korea.
Kljenak,
I.,
Dapper,
M.,
Dienstbier,
J.,
Herranz,
L.E.,
Koch,
M.K.,
Fontanet,
J.,
2010.
Thermal-hydraulic
and
aerosol
containment
phenomena
modelling
in
ASTEC
severe
accident
computer
code.
Nuclear
Engineering
and
Design
240,
656–667.
Lombard,
V.,
Azarian,
G.,
Ducousso,
E.,
Gandrille,
P.,
2014.
ASTEC
V2.0-rev1
reactor
applications
French
PWR
900
MWe
accident
sequences
and
comparison
with
MAAP4.
Nuclear
Engineering
and
Design
272,
219–223.
Matejovic,
P.,
Barnak,
M.,
Bachraty,
M.,
Vranka,
L.,
2014.
ASTEC
V2
appli-
cations
to
VVER-440/V213
reactors.
Nuclear
Engineering
and
Design
272,
245–260.
Micaelli,
J.C.,
et
al.,
2005.
SARNET:
a
European
cooperative
effort
on
LWR
severe
accident
research.
In:
Proc.
European
Nuclear
Conference,
December
14–18,
Versailles,
France.
Meyer,
L.,
Albrecht,
G.,
Wilhelm,
D.,
2004.
Direct
containment
heating
investigations
for
European
pressurized
reactors.
In:
NUTHOS-6
Conference,
October
4–8,
Nara,
Japan.
Mun,
C.,
Cantrel,
L.,
Madic,
C.,
2008.
Radiolytic
oxidation
of
ruthenium
oxide
deposits.
Nuclear
Technology
164
(2),
245–254.
Plumecocq,
W.,
Layly,
V.D.,
Bentaïb,
A.,
2005.
Modelling
of
the
containment
mitiga-
tion
measures
in
the
ASTEC
code,
focusing
on
spray
and
hydrogen
recombiners.
In:
NURETH-11,
October
2–6,
Avignon,
France.
Ramlakan,
A.,
Naidu,
N.,
Sanyasi,
M.,
Naidoo,
D.,
2010.
ASTEC
application
to
HTRs.
In:
4th
European
Review
Meeting
on
Severe
Accident
Research
(ERMSAR
2010),
May
11–12,
Bologna,
Italy.
Repetto,
G.,
Garcin,
T.,
Eymery,
S.,
March,
P.,
Fichot,
F.,
2011.
Experimental
pro-
gram
on
debris
reflooding
(PEARL)
–
Results
on
PRELUDE
facility.
In:
NURETH-14
Conference,
September
25–30,
Toronto,
Canada.
Sonnenkalb,
M.,
Poss,
G.,
2012.
Overview
on
Current
OECD-THAI-2
Experiments
and
German
National
THAI
Test
Series
for
CFD
Code
Validation.
In:
CSARP
meeting
2012,
September
11–13,
Bethesda,
ML,
USA.
Spitz,
P.,
Van
Dorsselaere,
J.-P.,
Schwinges,
B.,
Schwarz,
S.,
2001.
ASTEC
participation
in
the
international
standard
problem
KAEVER.
In:
EUROSAFE
Forum,
November
5–6,
Paris,
France.
Spitz,
P.,
et
al.,
2003.
Mass
transfer
programme
in
the
SISYPHE
facility.
In:
Proceedings
of
the
5th
PHEBUS
FP
seminar,
June
24–26,
Aix-en-Provence,
France.
Trégourès,
N.,
Philippot,
M.,
Foucher,
L.,
Guillard,
G.,
Fleurot,
J.,
2009.
ASTEC-
CATHARE
benchmark
on
the
French
PWR
1300
MWe
reactors.
In:
NURETH-13
Conference,
September
27–October
2,
Kanazawa
city,
Japan.
Van
Dorsselaere,
J.P.,
Schwinges,
B.,
Buck,
M.,
Ma,
W.,
Constantin,
M.,
Jancovic,
J.,
Ratel,
G.,
2008.
ASTEC
extension
to
other
reactor
types
than
Gen.
II
PWR.
In:
ERMSAR-2008
Conference,
September
23–25,
Nesseber,
Bulgaria.
Van
Dorsselaere,
J.P.,
Seropian,
C.,
Chatelard,
P.,
Jacq,
F.,
Fleurot,
J.,
Giordano,
P.,
Reinke,
N.,
Schwinges,
B.,
Allelein,
H.J.,
Luther,
W.,
2009a.
The
ASTEC
integral
code
for
severe
accident
simulation.
Nuclear
Technology
165,
293–307.
Van
Dorsselaere,
J.P.,
Perrault,
D.,
Barrachin,
M.,
Bentaïb,
A.,
Cortès,
P.,
Seropian,
C.,
Trégourès,
N.,
Vendel,
J.,
2009b.
R&D
on
support
to
ITER
safety
assessment.
Fusion
Engineering
and
Design
84,
1905–1911.
Van
Dorsselaere,
J.P.,
Chatelard,
P.,
Cranga,
M.,
Guillard,
G.,
Trégourès,
N.,
Bosland,
L.,
Brillant,
G.,
Girault,
N.,
Bentaïb,
A.,
Reinke,
N.,
Luther,
W.,
2010a.
Validation
status
of
the
ASTEC
integral
code
for
severe
accident
simulation.
Nuclear
Technology
170
(3),
397–415.
Van
Dorsselaere,
J.P.,
Auvinen,
A.,
Beraha,
D.,
Chatelard,
P.,
Journeau,
C.,
Kljenak,
I.,
Sehgal,
B.R.,
Tromm,
W.,
Zeyen,
R.,
2010b.
Status
of
the
SARNET
network
on
severe
accidents.
In:
ICAPP’10,
June
13–17,
San
Diego
(USA).
Veteau,
J.M.,
et
al.,
2006.
Proceedings
of
ICAPP
2006,
June
4–8,
Reno,
NV,
USA.
ARTEMIS
program:
investigations
of
MCCI
by
means
of
simulating
materials
experiments..
Zvonarev,
Yu,
Volchek,
A.M.\,
Kobzar,
V.L.,
Budaev,
M.A.,
2014.
ASTEC
application
for
in-vessel
retention
modelling
in
VVER
plants.
Nuclear
Engineering
and
Design
272,
224–236.