Conference PaperPDF Available

Focus on the main modelling features of ASTEC V2.1 major version

Authors:

Abstract and Figures

A new major version of the European severe accident integral code ASTEC, developed by IRSN with some GRS support, was delivered in November 2015 to the ASTEC worldwide community.Main modelling features of this V2.1 version are summarised in this paper. In particular, the in-vessel coupling technique between the reactor coolant system thermal-hydraulics module and the core degradation module has been strongly re-engineered to remove some well-known weaknesses of the former V2.0 series. The V2.1 version also includes new core degradation models specifically addressing BWR and PHWR reactor types, as well as several other physical modelling improvements, notably on reflooding of severely damaged cores, Zircaloy oxidation under air atmosphere, corium coolability during corium concrete interaction and source term evaluation.Moreover, this V2.1 version constitutes the back-bone of the CESAM FP7 project, which final objective is to further improve ASTEC for use in Severe Accident Management analysis of the Gen.II-III nuclear power plants presently under operation or foreseen in near future in Europe. As part of this European project, IRSN efforts to continuously improve both code numerical robustness and computing performances at plant scale as well as users' tools are being intensified.Besides, ASTEC will continue capitalising the whole knowledge on severe accidents phenomenology by progressively keeping physical models at the state of the art through a regular feed-back from the interpretation of the current and future experimental programs performed in the international frame.
Content may be subject to copyright.
Main modelling features of the ASTEC V2.1 major version
P. Chatelard
a,
, S. Belon
a
, L. Bosland
a
, L. Carénini
a
, O. Coindreau
a
, F. Cousin
a
, C. Marchetto
a
,
H. Nowack
b
, L. Piar
a
, L. Chailan
a
a
Institut de Radioprotection et de Sûreté Nucléaire (IRSN), PSN-RES, Cadarache, BP3, 13115 Saint Paul lez Durance, France
b
Gesellschaft für Anlagen und Reaktorsicherheit (GRS) mbH, Schwertnergasse 1, 50677 Köln, Germany
article info
Article history:
Received 31 July 2015
Received in revised form 4 December 2015
Accepted 7 December 2015
Available online 1 February 2016
Keywords:
ASTEC
Severe accidents
Code simulation
abstract
A new major version of the European severe accident integral code ASTEC, developed by IRSN with some
GRS support, was delivered in November 2015 to the ASTEC worldwide community.
Main modelling features of this V2.1 version are summarised in this paper. In particular, the in-vessel
coupling technique between the reactor coolant system thermal-hydraulics module and the core degra-
dation module has been strongly re-engineered to remove some well-known weaknesses of the former
V2.0 series. The V2.1 version also includes new core degradation models specifically addressing BWR
and PHWR reactor types, as well as several other physical modelling improvements, notably on reflood-
ing of severely damaged cores, Zircaloy oxidation under air atmosphere, corium coolability during corium
concrete interaction and source term evaluation.
Moreover, this V2.1 version constitutes the back-bone of the CESAM FP7 project, which final objective
is to further improve ASTEC for use in Severe Accident Management analysis of the Gen.II–III nuclear
power plants presently under operation or foreseen in near future in Europe. As part of this European pro-
ject, IRSN efforts to continuously improve both code numerical robustness and computing performances
at plant scale as well as users’ tools are being intensified.
Besides, ASTEC will continue capitalising the whole knowledge on severe accidents phenomenology by
progressively keeping physical models at the state of the art through a regular feed-back from the inter-
pretation of the current and future experimental programs performed in the international frame.
Ó2016 Elsevier Ltd. 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. Main ASTEC applica-
tions are therefore source term determination studies, level 2
http://dx.doi.org/10.1016/j.anucene.2015.12.026
0306-4549/Ó2016 Elsevier Ltd. All rights reserved.
Abbreviations: ASTEC, Accident Source Term Evaluation Code; BARC, Bhabha Atomic Research Centre, India; BIP, Behaviour of Iodine Project; BWR, Boiling Water Reactor;
CATHARE, Code Avancé de THermohydraulique pour les Accidents sur les Réacteurs à Eau; CCI, corium concrete interaction; CEA, Commissariat à l’Énergie Atomique et aux
Énergies Alternatives, France; CESAM, Code for European Severe Accident Management; CHIP, CHemistry of Iodine in the Primary circuit; CORDEB, CORium/DEBris behaviour
in the lower head; CSNI, Committee on Safety of Nuclear Installations; CT, calandria tube; EC, European Commission; EPICUR, Experimental Programme for Iodine Chemistry
Under Radiation; FP, fission product; FP7, 7th Framework Programme; GRS, Gesellschaft für Anlagen- und Reaktorsicherheit gGmbH; IRSN, Institut de Radioprotection et de
Sûreté Nucléaire; ISTP, International Source Term Programme; IVMR, in-vessel melt retention; JASMIN, Joint Advanced Severe accident Modelling and INtegration for sodium-
cooled fast neutron reactors; KIT, Karlsruhe Institute of Technology; LCDA, Limited Core Damage Accident; LOCA, loss of coolant accident; LWR, Light Water Reactor; MCCI,
molten core concrete interaction; MOZART, Mesure de l’Oxydation du Zirconium par l’AiR en Température; NITI, Nauchno-Issledovatel’skii Tekhnologicheskii Institute,
Russia; NPP, nuclear power plant; OECD, Organisation for Economic Cooperation and Development; PAR, passive autocatalytic recombiner; PASSAM, Passive and Active
Systems on Severe Accident source term Mitigation; PEARL, Programme Expérimental Analytique sur le Renoyage des Lits de débris; PHWR, Pressurised Heavy Water Reactor;
PRELUDE, Préliminaire sur le Renoyage ExpérimentaL d’Un lit de DEbris; PSA2, level 2 Probabilistic Safety Assessment; PT, pressure tube; PWR, Pressurised Water Reactor;
RCS, reactor coolant system; SA, severe accident; SAFEST, Severe Accident Facilities for European Safety Targets; SAM, Severe Accident Management; SARNET, Severe Accident
Research NETwork; SBLOCA, small break LOCA; SCDA, Severe Core Damage Accident; SM, structural materials; SSWICS, Small Scale Water Ingression and Crust Strength;
STEM, Experimental programme for study of Source Term and Mitigation; ST, source term; SUNSET, Statistic UNcertainty and Sensitivity Evaluation Tool; THAI, Thermal
Hydraulics And Iodine.
Corresponding author.
E-mail address: patrick.chatelard@irsn.fr (P. Chatelard).
Annals of Nuclear Energy 93 (2016) 83–93
Contents lists available at ScienceDirect
Annals of Nuclear Energy
journal homepage: www.elsevier.com/locate/anucene
Probabilistic Safety Assessment (PSA2) studies, accident manage-
ment studies and physical analyses of experiments to improve
the understanding of the phenomenology.
ASTEC progressively became the reference European severe
accident integral code for water-cooled reactors through the capi-
talisation of new knowledge acquired in the frame of the SARNET
European Network of Excellence from 2004 to 2013 (SARNET,
2008; Van Dorsselaere et al., 2011, 2015a). Successive revisions
of the V2.0 version (Chatelard et al., 2014a) were continuously
assessed by SARNET partners through comparison either with
results of the most important international experiments covering
the main SA physical phenomena (Chatelard et al., 2014b) or with
results of other severe accident simulation codes (Chatelard et al.,
2014c), thus confirming or allowing to identify the topics on which
modelling efforts should be paid.
In accordance, a second ASTEC V2 major version has been pro-
gressively developed at IRSN since 2011 with some GRS support,
aiming at removing some shortcomings of the V2.0 series, while
continuing keeping ASTEC models close to the state of the art
through the integration of new or improved physical models gath-
ering a large part of the knowledge generated by SARNET (Van
Dorsselaere et al., 2015a) and the ISTP (International Source Term
Programme) (Clement and Zeyan, 2005). This V2.1 new version
notably includes new core degradation models specifically
addressing Boiling Water Reactors (BWR) and Pressurised Heavy
Water Reactors (PHWR). Besides, the V2.1 version constitutes the
back-bone of the CESAM FP7 project (CESAM, 2013; Chatelard
et al., 2014d; Van Dorsselaere et al., 2015b), which final objective
is to further improve ASTEC for use in Severe Accident Manage-
ment (SAM) analysis of the Gen.II–III nuclear power plants pre-
sently under operation or foreseen in near future in Europe.
This paper presents the main V2.1 advances in terms of capabil-
ities and/or physical modelling with respect to the former V2.0
series. After a brief recall of some general ASTEC features, the
main part is devoted to a description of few selected important
V2.0-to-V2.1 evolutions, along with a summary of their prelimi-
nary validation and very first applications at plant scale. In
addition, the main perspectives of further improvements in the
next years are briefly summed up.
2. ASTEC V2 code structure and main programming features
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:
stand-alone mode for running each module independently (useful
for module validation) and coupled mode where all (or a subset) of
the modules are run sequentially within a macro-time step.
The ASTEC V2.1 general organisation is very similar to the one
of former V2.0 versions (Chatelard et al., 2014a), the main differ-
ence lying in the activation now for complete plant applications
(coupled running mode) of the SOPHAEROS module to deal with
fission product (FP) and aerosol transport in the containment
(see Fig. 1), the CPA module being now only in charge of the con-
tainment thermal-hydraulics. Besides, the IODE module has been
fully integrated into SOPHAEROS.
3. Main evolutions from ASTEC V2.0 to V2.1
3.1. Coupling between reactor coolant system (RCS)
thermal-hydraulics and core degradation modules
In former ASTEC V2.0 series, the treatment of the thermal-
hydraulics in the core region was made using sequentially two dif-
ferent numerical approaches during a given calculation, depending
Fig. 1. Schema of the ASTEC V2.1 modules, code structure and running mode.
84 P. Chatelard et al./ Annals of Nuclear Energy 93 (2016) 83–93
on the phase of the accident (front-end phase or degradation
phase). During the front-end phase, CESAR alone calculated the
thermal-hydraulics in the whole RCS, i.e. including the vessel, up
to an automatic switch to ICARE. Such a switch was generally
triggered around time of start of first core dry-out. However, for
SA sequences involving a late hydro-accumulator (HA) operation,
the ICARE start was often postponed to the end of the HA discharge
phase to prevent possible numerical problems linked to the
CESAR/ICARE numerical coupling. After the switch, CESAR calcu-
lated only thermal-hydraulics in the loops and the vessel upper
plenum, while in addition to the degradation phenomena ICARE
calculated thermal-hydraulics in the remaining part of the vessel
(core, bypass, lower plenum and downcomer) all along the
degradation phase, but using a simplified thermal-hydraulics mod-
elling based on 0D liquid water components below (r-z) gas flows.
That means the former V2.0 versions were not designed to
properly handle transient situations when water is re-entering into
the core after a first core dry-out and heat-up, the switch to
ICARE being an irreversible event during an ASTEC V2.0 complete
calculation.
To allow best-estimate simulations of two-phase flows in the
core during the degradation phase, the in-vessel coupling tech-
nique between the RCS thermal-hydraulics module and the core
degradation module has been strongly re-engineered (see Fig. 2).
In V2.1, CESAR is now covering the thermal-hydraulics in the
whole RCS (vessel and loops) during the whole transient, while
ICARE deals with the thermal behaviour and degradation of all ves-
sel structures (heat-up, creep, oxidation, material relocation, ...)
since the beginning of the transient.
As to the physical relevance of the ASTEC V2 simulations, such a
new CESAR/ICARE coupling notably allows removing the above-
mentioned V2.0 weaknesses relating to the simulation of acciden-
tal sequences involving a core reflooding occurring after a first
heat-up and oxidation phase. The improved coupling makes
accordingly the V2.1 version fully applicable to SA scenarios
involving a delayed core quenching, thus providing ASTEC users
with an adequate frame to deal with the reflooding of degraded
cores, as illustrated in (Drai et al., 2015) through a recently revised
ASTEC simulation of the TMI-2 accidental transient.
Besides, a 2D extension of CESAR has been simultaneously
developed at IRSN to support a radial/axial discretization of the
core region as in ICARE, thus allowing accounting in the V2.1 ver-
sion for in-core 2D two-phase flow patterns. In addition, to prop-
erly handle the core degradation phase, CESAR was adapted to
account for geometry changes in the core control volumes.
As to the numerics, in order to gain computing time, each mod-
ule is running at its own time-step thus allowing the resolution of
the core degradation processes to be possibly achieved with a lar-
ger time step than the one used to solve RCS/vessel two-phase
thermal-hydraulics since CESAR can sometimes require rather
small time steps. Meeting points occur at the end of an intermedi-
ate macro time-step.
In order to briefly illustrate how this new coupling scheme can
influence the course of an ASTEC simulation, some first V2.1 results
obtained at IRSN on the basis of a PWR small break loss of coolant
accident (SBLOCA) sequence are briefly discussed in Section 4and
compared to V2.0 results. Furthermore, few information about
computing time are provided too.
3.2. BWR and PHWR core degradation
The main conclusions from the investigations that were done by
several partners in the frame of the SARNET FP7 project (Chatelard
et al., 2014c) were that most of the ASTEC V2.0 modules were suit-
able enough to support analyses addressing BWR and PHWR (in
particular containment models) while it was not the case for the
ICARE module that was unable to represent the real core geometry
for BWR and PHWR. Indeed, in ASTEC V2.0, the ICARE modelling
was based on a schematic view of the core inspired by PWR design.
As a consequence the geometry of objects that were used to repre-
sent the core was only cylindrical. On a thermal-hydraulics point of
view, the core was discretized in concentric rings as the meshing is
only axisymmetrical. To answer BWR and PHWR modelling
requirements, specific developments have been made at IRSN in
ASTEC V2.1, mainly consisting in setting up a new core meshing
architecture to allow multi-channels description and creating
new geometrical objects to model some BWR core structures.
Besides, these new V2.1 features are also useful to model spent fuel
pools.
3.2.1. BWR core modelling
The method chosen in ASTEC V2.1 to model a BWR core consists
in describing several core areas in which fuel assemblies could be
represented by an average representative assembly. An average
representative fuel assembly (numerically weighted to model a
group of fuel assemblies located in the same core area) is com-
posed of a representative fuel rod (composed of fuel pellets sur-
rounded by a cladding), a representative water rod and a
representative canister while each core area can also contain rep-
resentative control blades. Fuel rods and water rods are repre-
sented using basic cylindrical objects that were already existing
in V2.0 while new objects have been specifically created to repre-
sent the canisters (square box channel which can contain fluid and
other structures) and the cross-shape control blades.
Moreover, from a thermal-hydraulics point of view, the new
multi-channels meshing allows dealing with both intra-canister
and inter-canister flows, up to canister failure and then fluid mix-
ing due to canister degradation.
Beside the set up of new meshing and new objects, early phase
physical models (heat transfers, oxidation...) have been also
adapted to properly account for plane geometries and new mesh-
ing, but no BWR specific correlations have been implemented in
ASTEC V2.1. New or improved physical models for BWR analyses
will be implemented later in subsequent V2.1 revisions according
to specifications to be proposed in the CESAM frame as a feed-
back from the V2.1 validation by partners. Notably, as to the late
degradation phase, modelling needs have been already identified,
such as e.g. corium progression in BWR vessel (outside and inside
guided tubes located below the core) or mechanical behaviour of
the vessel lower head (the current model does not take into
account lower head penetrations).
Fig. 2. Principle of the CESAR/ICARE new coupling.
P. Chatelard et al./ Annals of Nuclear Energy 93 (2016) 83–93 85
3.2.2. PHWR core modelling
PHWR core comprises number of horizontal channels sub-
merged in relatively cold moderator. First the above-mentioned
new multi-channels meshing goes far beyond the strict scope of
BWR applications since it also fully benefits to the description of
PHWR cores. Besides, core thermal-hydraulics CESAR modelling
has been extended by IRSN to address fluid flows in horizontal
cores, what was not possible with former V2.0 versions.
PHWR core damage accident scenarios are classically divided in
two distinct phases based on the availability of the moderator heat
sink: Limited Core Damage Accident (LCDA) and Severe Core Dam-
age Accident (SCDA).
In the frame of the IRSN-BARC (India) collaboration, the BARC
team focused its modelling efforts on the simulation of the LCDA
phase where channels are intact but amenable to deformation.
Indeed, the peculiarity of horizontal type of reactors is that high
temperature thermal creep of the pressure tube (PT) in the direc-
tion of gravity would eventually lead to circumferential contact
of PT and calandria tube (CT) at the bottom. So, BARC developed
a specific thermo-mechanical model addressing firstly the PT creep
sagging in addition to ballooning up to PT-CT contact, and secondly
the heat transfer modelling due to the PT-CT contact. The way this
model has been integrated into ASTEC V2.1 is detailed in
(Majumdar et al., 2012); its successful combination with the new
multi-channels meshing has been then demonstrated at the reac-
tor scale by BARC (Majumdar et al., 2013).
Next step regarding the adaptation of ASTEC to PHWR applica-
tions will focus on the SCDA phase. It will consist in describing the
configuration of a molten pool located inside the cylindrical calan-
dria vessel cooled on its external face by water contained inside the
vault, with final goal to answer some key issues regarding safety
for such SCDA configuration.
3.3. Core reflooding
Core reflooding is an important SAM measure in order to stop or
at least delay the progression of a severe accident in a PWR.
Depending on the time during the accident at which water sources
could be available, two quite different geometrical core configura-
tions have to be considered for the reflooding process: (1) a
rod-like geometry representative of a quasi-intact core (with only
ballooned claddings) or of the early degradation phase (possibly
exhibiting some candling of molten materials but somewhat
limited, i.e. without any local flow blockages); (2) a debris bed
geometry representative of the late degradation phase.
In order to predict accurately such phenomena, adequate mod-
els were developed few years ago in the mechanistic IRSN ICARE/
CATHARE V2 code (Drai et al., 2007), and they have been imple-
mented in ASTEC V2.1, taking into account the specifics of the
CESAR thermal-hydraulics modelling with respect to the best-
estimate CATHARE one (Robert et al., 2003).
In this paper, focus is made on the model for severely damaged
cores. As concerns the improved reflooding model addressing rod-
like geometries, details about the adopted physical and numerical
modelling can be found in (Chikkhi and Fichot, 2010) and (Drai
et al., 2015), along with results from respectively past ICARE/CATH-
ARE V2 and more recent ASTEC V2.1 model assessment vs. CEA
PERICLES large scale experiments.
For debris bed reflooding, the particularity is the higher temper-
atures, the smaller hydraulic diameter and tortuosity so that the
coolability is more difficult to achieve. The ASTEC V2.1 model takes
into account the two-phase flow through the heated porous bed
(Bachrata, 2012). It is based on modified momentum balance equa-
tion and on specific heat transfer laws between the debris and the
fluid.
Friction forces between solid and fluid phases are based on the
classical extension of Darcy’s law to two-phase flows. Since the
CESAR thermal-hydraulics model is based on a mean fluid momen-
tum equation, the viscous and inertial drag forces are calculated
using intrinsic permeability and passability coefficients correlated
with the average particle diameter and the porosity by the
Carman-Kozeny (Carman, 1937) relation and Ergun law (Ergun,
1952). Corresponding drift velocity is obtained considering equilib-
rium between the gravity force and the wall frictions.
The existing CESAR constitutive heat transfer relations, that are
described in terms of a complete boiling curve, are adapted in
order to take into account small hydraulic diameter channel char-
acteristics. The nucleate boiling regime is modified when the
hydraulic diameter reaches its critical size below which the bubble
separation from the wall is influenced. In this case, the bubble sep-
aration from the wall will be delayed and will be limited, and
therefore the critical heat flux will reduce. This induces the modi-
fication of the critical heat flux correlation for values of hydraulic
diameter lower than a critical size. A transition regime is described
by a specific interpolation function. The particularity of the model
is a transition criterion in order to determine the nucleate boiling
regime, the transition regime and the film boiling regime. The heat
transfer layer length L1, function of Weber number, compared to
the difference between the mesh elevation and the quench front
position (dz), determines the film boiling (dz > L1) or the transi-
tion/nucleate boiling (dz < L1) regimes.
The model for debris bed reflooding is validated on small-scale
IRSN PRELUDE experiments with initial debris bed temperature at
400 °C or 700 °C(Repetto et al., 2013). The heated steel debris bed
is separated from a metallic grid by a quartz particles layer in order
to avoid possible grid heat-up due to induction. Validation is per-
formed at atmospheric pressure for different bottom water injection
velocities (2 m/h, 5 m/h, 10 m/h and 20 m/h) and for particle diam-
eter of 4 mm. Obtained results show a good agreement of ASTEC
V2.1 predictions with experimental data, as illustrated on Fig. 3
for the test 58 (T
initbed
= 400 °C, T
injec
=20°C and V
injec
= 2 m/h).
Next steps will be to improve this reflooding model of severely
damaged cores using in particular the data to be produced by the
larger scale IRSN PEARL experimental programme (Chikhi et al.,
2015) where 2D effects should be observed.
3.4. Zircaloy oxidation under air atmosphere
Air ingress scenarios are an important issue for SA codes since
such situations are known to likely provide an acceleration of
the cladding oxidation, fuel rod degradation and release of some
fission products. Such situations could be notably met inside the
vessel after lower head rupture, or in spent fuel pools.
However, such situations could not be properly dealt with using
the ASTEC V2.0 former series due firstly to limitations that were
inherent to the use of a simplified core thermal-hydraulics not
allowing to properly dealing with O
2
/N
2
mixtures.
The ASTEC V2.1 model for Zircaloy oxidation by air is derived
from the model that was originally developed for the ICARE/CATH-
ARE V2 code. In the low temperature range (up to 1100 °C), it
deals with the different stages of the physical process that have
been experimentally observed. The model considers that the inter-
action between Zr and O
2
occurs, in a first stage, following a para-
bolic kinetics since the role of nitrogen in that pre-transition
regime is considered to be relatively weak. In a second stage, due
to the presence of nitrogen, the initially dense and protective oxide
layer rapidly cracks, which leads to a kinetic transition (or break-
away); oxidation is no more controlled by the diffusion of oxygen
through the oxide layer and the acceleration of oxidation kinetics is
taken into account according to IRSN MOZART experimental data
(Coindreau et al., 2010). The subsequent transition to the
86 P. Chatelard et al. / Annals of Nuclear Energy 93 (2016) 83–93
post-break-away regime is based on a critical value of the oxygen
mass gain, function of temperature (the weight gain at which the
kinetic transition happens increases with increasing temperature).
Finally, the post-break-away regime is modelled by a linear law,
according to MOZART kinetics.
If the fuel cladding is oxidised in air at high temperature
(>1200 °C), there is no more breakaway, but nitriding can occur in
case of lack of oxygen. A preliminary model has been implemented
in ASTEC V2.1 to deal with the nitriding phenomena in case of total
oxygen starvation. As a first attempt, the nitriding rate follows a lin-
ear kinetics derived from KIT experimental investigations.
A very first application of this new ASTEC V2.1 model, that was
done in 2014 at IRSN in the frame of a code-to-code benchmark vs.
two KIT QUENCH integral experiments (resp. Q10 and Q16 tests),
can be found in (Beuzet et al., 2015). Applications at plant scale
of the Zr/air oxidation model are not discussed in this paper since
they remain up to now limited to few IRSN private calculations car-
ried out in the frame of spent fuel pool accident analyses.
Next modelling steps will be to include in the low temperature
model the presence of steam and the effect of the oxygen partial
pressure in the computation of the kinetic rate. Then, a model
addressing the zirconium nitrides oxidation (responsible for the
formation of porous oxide and fast cladding degradation) should
be also progressively developed at IRSN.
3.5. Fission product and structure materials transport and chemistry in
circuits and containment
An important modelling issue is addressing the transport of
iodine through the RCS by implementations in SOPHAEROS of
kinetics aspects (feed-back from lessons drawn from the ASTEC
V2.0 validation vs. Phébus FP tests). Indeed, FP chemistry inside
the RCS has a deep impact on deposition rate along the RCS, on
the gas/aerosol partition entering the containment (especially for
iodine) and thus on the source term release to environment.
A very first model for kinetics of gaseous phase chemistry was
set up in 2013 at IRSN and included in the V2.0 latest version,
focusing at that time on the Cs–I–O–H system. This modelling
has been then extended to few other elements (Mo for fuel and B
for absorber), mainly based on the interpretation of the IRSN CHIP
experimental programme (Grégoire and Mutelle, 2012). Therefore,
the V2.1 version allows now addressing the Cs–I–O–H–Mo–B sys-
tem. In order to keep the SOPHAEROS kinetics at the state of the
art, this model is now being progressively further extended with
aim at notably accounting at short term for other control rod mate-
rials (such as Cd, Ag and In) in the frame of the CHIP + programme.
Examples of the significant impact on iodine speciation of the new
(I,O,H) kinetic scheme implemented in SOPHAEROS can be found in
(Cantrel et al., 2013) and (Gouello et al., 2013) that are respectively
summarising the SOPHAEROS V2.1 validation vs. Phébus FPT1 and
FPT3 tests and vs. some CHIP data relating to the modelling of the
Mo–Cs–I transport.
Beside this physical improvement of the RCS chemistry mod-
elling, significant SOPHAEROS evolutions have been done at IRSN
to enlarge its scope of application to containment. While the appli-
cation was restricted in former ASTEC versions to the RCS domain,
SOPHAEROS is now in charge in the V2.1 version of FP and struc-
ture materials (SM) transport and chemistry in both RCS and con-
tainment. For that purpose, the IODE module has been integrated
into SOPHAEROS which is now also entirely managing FP/SM phys-
ical phenomena that were previously treated (up to V2.0) in CPA
AFP sub-module (Chatelard et al., 2014a). This development allows
achieving in V2.1 a complete and consistent treatment of FP and
SM transport in the whole reactor. Especially, it allows computing
iodine chemistry in containment, iodine oxide formation and pos-
sible interaction with aerosol behaviour. Moreover, with respect to
older code versions, iodine mass balance is also simplified in con-
tainment because it is the same module which now computes
iodine behaviour.
SOPHAEROS has been firstly extended to account for FP and SM
liquid transport, knowing that in this case, only elements are trans-
ported (speciation is lost when FP and SM settle or are deposited in
sump; only iodine speciation in liquid phase is computed linked
with iodine chemistry). SOPHAEROS is now also managing iodine
capture by containment spray system while aerosol washing is also
implemented in SOPHAEROS using the same model as formerly in
CPA AFP. Moreover, kinetic reactions that have been developed for
the RCS are taken into account in containment, especially for
iodine chemistry.
Besides, the SOPHAEROS extension also allows dealing in con-
tainment with FP and SM speciation for other element such as
Cs, Mo, Te as well as taking into account key-phenomena such as
FP and SM nucleation to form aerosol, vapour condensation/evap-
oration onto aerosol and FP and SM condensation/evaporation onto
wall. An example of a very first application at plant scale of the
SOPHAEROS V2.1 extended version can be found in (Chevalier-
Jabet et al., 2015).
3.6. Iodine chemistry
Recently, significant improvements were made in the under-
standing of gaseous iodine formation on the basis of the ISTP and
Fig. 3. Validation of the ASTEC V2.1 core reflooding model vs. PRELUDE 1D test 58.
P. Chatelard et al. / Annals of Nuclear Energy 93 (2016) 83–93 87
OECD/NEA/CSNI STEM, BIP2 and THAI-2 project results (Mun et al.,
2015; Glowa et al., 2015; Funke et al., 2015). Two new correspond-
ing ASTEC models are briefly explained below.
STEM/LD tests series performed in IRSN EPICUR facility from
2011 to 2013 have shown that gaseous I
2
and CH
3
I are significantly
released from the paint by radiolytic processes. For each species
two kinetics are exhibited: a fast one in the short term (<12 h)
and a slower one in the long term. Influence of the iodine concen-
tration on the paint, the temperature, the dose rate and the relative
humidity was also experimentally investigated. Current model in
ASTEC V2.0 (Funke, 1999) did not model well all these available
data. Accordingly, the new iodine-paint model takes into account
the influence of those parameters on the released kinetics for I
2
and CH
3
I. Its application to the Phébus FPT3 test (Di Giuli et al.,
2016) has shown that the gaseous I
2
concentration is quite well
reproduced at short term whereas the gaseous CH
3
I concentration
is underestimated by one order of magnitude. The latter
discrepancy was attributed to a missing reaction in the gaseous
phase modelling, especially for the long term, between organics
compounds (C
n
H
m
) and I
2
under irradiation.
A theoretical approach was therefore set up to identify the
mechanisms leading to gaseous iodine formation under irradiation.
The reaction between gaseous organics (like methane) and gaseous
I
2
can lead to the formation of gaseous CH
3
I by a radiolytic process.
A simple model modelling the reaction between organics (CnHm)
and gaseous I
2
under irradiation has been then developed on the
basis of some relevant and representative literature data
(Bartonicek and Habersbergerova, 1987) and implemented in
ASTEC V2.1. Its application to the simulation of the Phébus FPT3
and RTF1 tests exhibits a significant improvement of the CH
3
I glo-
bal behaviour, as illustrated hereafter on Fig. 4 by the observed
increase of the gaseous CH
3
I amount by about one order of magni-
tude. At this stage, this simplified model appears promising even if
it needs to be consolidated by more precise data. Next modelling
step will be therefore to refine it according notably to new data
to be produced by the OECD/STEM2 project (Mun et al., 2015)
starting in January 2016 and aiming at including the effect of the
parameters (such as e.g. temperature, reactants concentration,
humidity and dose rate) having a significant influence on iodine
volatility.
3.7. Molten core concrete interaction (MCCI) coolability
New models were implemented in the MEDICIS module on cor-
ium coolability by top quenching during MCCI (Cranga et al., 2012).
A dedicated debris bed layer is now modelled above and apart
from the upper crust, thus allowing to properly account for the
continuous cooling of already ejected debris by top water injection.
A simple energy balance for the debris bed is applied assuming that
the built-up debris remain at saturation temperature because of
the much larger dry-out heat flux than in case of the heat flux
extracted from the upper crust. Indeed, provided that the debris
bed porosity exceeds around 0.3 and the debris size a few mm,
which are likely assumptions, the dry-out heat flux deduced from
available correlations (Lipinski, 1982) reaches around 1 MW/m
2
which permits extracting the total decay power of a core inventory
even in the case of a large PWR reactor. In case of insufficient water
inventory, this dry-out is supposed to start at the bottom of the
debris bed. One can refer to (Cranga et al., 2012) to get details
about the mass and energy balance equations that are written for
the debris bed. A preliminary validation of this debris layer mod-
elling has been successfully achieved at IRSN in 2014 vs. the ANL
CCI-7 experimental data.
With regard to water ingression, the detailed model proposed
by Epstein, based on considerations on the material creep and
cracking behaviour (see (Epstein, 1999) and (Lomperski and
Farmer, 2007)), has been implemented and taken into account in
the heat flux continuity equation of the upper crust. This model,
that was validated against ANL SSWICS tests, is considered to be
sufficiently mechanistic.
As to melt eruption, it was accounted for in former V2.0 ver-
sions according to the Ricou-Spalding model. To set the ASTEC
MCCI modelling at the state of the art, the PERCOLA model
(Tourniaire et al., 2006) for the melt ejection hydrodynamics has
been implemented in addition in ASTEC V2.1. This ‘‘fountain”
model is combining a double phase upwards flow through the hole
(accounting for the impact of the melt viscosity) and a lateral liquid
pouring. A preliminary validation of this improved modelling has
been successfully carried out at IRSN in early 2015 vs. the CCI-7
and CCI-8 experimental data (both are experiments recently con-
ducted at ANL that focussed on corium coolability from the top
CH3R + I2 + γ => CH3I reaction not
included
CH3R + I2 + γ => CH3I reaction included
Fig. 4. Validation of new ASTEC V2.1 iodine chemistry models vs. Phébus RTF1 test.
88 P. Chatelard et al. / Annals of Nuclear Energy 93 (2016) 83–93
during early phases of MCCI). It has however to be stressed that, as
a first step, a fixed hole size and density is assumed in V2.1.
A combined use of PERCOLA with available models for determining,
as detailed as possible way, the precise ranges of hole diameter and
hole density corresponding to the addressed MCCI situation is only
planned in a second step.
3.8. Main other V2.1 new features
Beside the few modelling improvements that have been
described in this paper, the V2.1 version is also including several
other evolutions with respect to former V2.0 versions. One may
notably mention:
The account for NIS-type particle bed passive autocatalytic
recombiners (PAR) that is currently being developed in the
frame of the CESAM FP7 project in addition to the box-type
PAR produced by AREVA that was available in former V2.0
versions.
A new model devoted to characterise the core degraded geome-
tries and aiming at continuously evaluating in each core mesh
(according to its composition, i.e. rods, debris, magma, other
structures) the local effective properties that drive the heat
and momentum transfers.
A 0D neutronics model that was built-up at IRSN in the frame of
the European JASMIN FP7 project (Girault et al., 2013).
A dedicated mechanical model (explicitly coupled with the
CESAR module) aiming at providing users with a rough evalua-
tion of RCS induced failure risks that could be met in case of SA
high pressure scenario.
A preliminary simple model addressing the radiolytic decompo-
sition into gaseous I
2
of iodine oxides and multi-components
aerosols coming from the circuit (such as CsI, CdI
2
, AgI) that is
an outcome of the OECD/STEM project (Bosland and Cantrel,
2015).
...
In addition, efforts have been directed towards a general
improvement of both pre-processing and post-processing tools
for code applications. A new graphical user interface (GUI)
XASTEC was in particular set up at IRSN to provide ASTEC users
with an adequate user-friendliness tool to build-up or modify
complete plant input decks, launch calculations and then post-
process the results. Furthermore, the XASTEC tool gives also the
possibility to easily use the ASTEC/SUNSET (Chevalier-Jabet
et al., 2014) coupling to evaluate the influence of uncertainties
on data or models on the simulation results. This new GUI will
be continuously improved in the next years accounting notably
for the users’ feed-back.
4. Example of a first ASTEC V2.1 application to a complete
reactor severe accident sequence
4.1. Foreword
A SBLOCA sequence, applied to a French PWR 900MWe, is cho-
sen to briefly illustrate one possible impact of the CESAR/ICARE
coupling redesigning onto the ASTEC V2 predicted results. All
ASTEC V2 modules are activated (see Fig. 1) in the calculations,
thus allowing describing the whole SA sequence i.e. both in-
vessel and ex-vessel phases. The accident real time covered by
the ASTEC V2.0 and V2.1 simulations is about 2 days.
Focus is made hereafter on in-vessel processes through a V2.1
vs. V2.0 comparative analysis of the main results obtained for the
core degradation phase. As to ex-vessel phenomena, the ASTEC
results are simply illustrated in this paper through few outputs
taken only from the V2.1 simulation and relating respectively to
the iodine behaviour in containment and MCCI.
4.2. Results overview and discussion
According to the chosen hypothesis for the simulated scenario
(i.e. availability of the systems and retained operator actions),
more than one hour is necessary to depressurise the primary
circuit up to triggering at 42 bar the HAs discharge (Fig. 5).
In the V2.0 simulation, the switch to ICARE was delayed up to
10,000 s (i.e. after the HAs isolation) to let CESAR entirely manag-
ing alone the core refilling phase: proceeding that way implied
neglecting in-core oxidation phenomena during that 2 h long cool-
ing phase.
In the V2.1 simulation, the first partial core uncovery that starts
20 min before the HAs discharge operation (Fig. 6) involves a sig-
nificant heat-up of the fuel assemblies. This core heat-up, then
enhanced by the achievement of a first oxidation phase of the fuel
rod claddings, leads to produce 290 kg hydrogen (Fig. 7). Then,
the water supply into the core leads to recover large enough in-
core convective heat transfers to allow significantly reducing the
total energy stored in the core during 1 h, i.e. up to the end of
the HAs discharge phase. After the HAs isolation, the progressive
core dewatering restarts since other safety injection systems are
assumed unavailable (Fig. 6). There is no more any means to cool
the core and thus to prevent the in-core oxidation processes to
restart. A second oxidation phase of the fuel rod claddings is there-
fore observed, leading to produce 225 kg more hydrogen (Fig. 7).
So, the total amount of hydrogen produced during the in-vessel
degradation phase reaches more than 500 kg according to the
V2.1 simulation while it was found to remain a bit below 300 kg
according to the former V2.0 calculation (Fig. 7). As a consequence
of the higher total power produced by the oxidation reactions in
the V2.1 simulation, the second core dry-out kinetics is faster than
in the V2.0 simulation (Fig. 6), leading to a faster core degradation
kinetics and an earlier formation of a molten pool in the core (Fig. 8
– top views showing that about the same degradation level was
reached 50 min later with V2.0). The subsequent corium transfer
from the core into the lower head is also detected earlier, as illus-
trated on Fig. 5 by the pressure peak corresponding to the strong
vaporisation of the water pool located in the lower head at the time
of massive corium slump.
The vessel bottom head failure is detected a bit earlier in V2.1 in
comparison to V2.0, and at a higher elevation. Due to a different
corium composition in the lower head (higher oxide fraction in
the V2.1 simulation), a corium stratification develops with a light
metal layer at the top, inducing with V2.1 the rupture of the vessel
wall to be met in front of this upper layer (Fig. 8).
Due to the V2.1 higher failure location, the amount of corium
transferred to the cavity at the vessel rupture time is smaller and
less UO
2
-rich compared to V2.0. That leads driving a little weaker
MCCI than in older V2.0 calculation. In the end, as illustrated by
the V2.1 final cavity shape (Fig. 9), the basemat erosion (which is
here equal in both axial and radial directions according to the cho-
sen assumptions) reaches about 220 cm vs. 250 cm for the V2.0
calculation.
Looking now very shortly at the source term evaluation, the
obtained results with V2.1 are consistent with what was expected
(according to experiences got from past V2.0 simulations) since the
selected transient (NPP type and scenario) was not in favour of
highlighting the impact of V2.1 modelling improvements on gas-
eous iodine chemistry. Indeed, combining the high silver inventory
of a PWR 900MWe core with the assumption of a permanent CSS
operating (no loss of CSS at the recirculation) led respectively
P. Chatelard et al. / Annals of Nuclear Energy 93 (2016) 83–93 89
favouring the formation of iodine aerosols in the RCS (such as AgI)
and the trapping of iodine in the containment sump (which pH is
alkaline), thus limiting the amount of gaseous volatile iodine in
the containment atmosphere (Fig. 9).
4.3. Summary
In summary, focussing on the in-vessel degradation phase, the
main difference in the predicted results for the 2
00
cold leg LOCA,
when using ASTEC V2.1 versus the older V2.0, lies in the way to
account for oxidation processes. While the first oxidation phase
occurring before the onset of HAs discharge could not be reliably
addressed by the V2.0 version due to some penalising CESAR/ICARE
coupling restriction, it is now possible with ASTEC V2.1 to ade-
quately simulate the entire in-core oxidation processes regardless
of the accident time window they are taking place.
Anyway, this discussion aimed simply at providing in this paper
a typical example of some ASTEC plant simulation improvements
Fig. 5. PWR 900MWe SBLOCA: evolution of the primary pressure.
Fig. 6. PWR 900MWe SBLOCA: evolution of the water mass inventory in the vessel.
Fig. 7. PWR 900MWe SBLOCA: evolution of the in-vessel hydrogen production. Comparison between current V2.1 results (red) vs. older V2.0 results (orange).
90 P. Chatelard et al. / Annals of Nuclear Energy 93 (2016) 83–93
Fig. 8. PWR 900MWe SBLOCA: evolution of the core degradation. Comparison between current V2.1 results (left) vs. older V2.0 results (right).
Fig. 9. PWR 900MWe SBLOCA: simulation of ex-vessel processes with ASTEC V2.1: final cavity shape (left) and evolution of the iodine repartition (right) in the containment.
P. Chatelard et al. / Annals of Nuclear Energy 93 (2016) 83–93 91
brought by the V2.1 CESAR/ICARE re-engineered coupling, but in
no way to draw generalised conclusions for any kind of SA
sequence simulation. For instance, IRSN on-going detailed analyses
of several other 900MWe and 1300MWe PWR sequences have
shown that the 2D modelling of the in-vessel two-phase
thermal-hydraulics contributes also to get with V2.1 a more con-
sistent evaluation of the in-core convective heat transfers and thus
a more relevant evaluation of the (r-z) heat distribution in the
vessel during the core degradation phase.
4.4. Computing time
Lastly, to give some idea of the ASTEC V2.1 numerical perfor-
mances for complete reactor applications (i.e. with all in-vessel
and ex-vessel modules running together), the V2.1 computing time
is compared with V2.0 for the simulation of 2 days of accident for
two different SA scenarios applied to a 3-loop French PWR
900MWe: for a sequence initiated by a 2
00
cold leg LOCA, the com-
puting time is quite similar for V2.0 and V2.1 and about the real
time, while for a sequence initiated by a total loss of steam gener-
ator feed-water, the V2.1 simulation is faster than the V2.0 one
(respectively 1d4h and 1d20h) and faster than real time. Moreover,
this trend is confirmed by ASTEC applications to other NPPs since
e.g. the computing time to achieve the V2.1 simulation of 2 days
of accident for a 12
00
hot leg large break LOCA SA sequence applied
to a 4-loop French PWR 1300MWe is also less than the real time
(1d8h).
5. Conclusion and perspectives
The new V2.1 major version of the European severe accident
integral code ASTEC, developed by IRSN with some GRS support,
was delivered in November 2015 to the ASTEC worldwide
community.
As to the code general features, the in-vessel coupling technique
between the RCS thermal-hydraulics module and the core degrada-
tion module has been strongly re-engineered. The V2.1 version also
includes new core degradation models specifically addressing BWR
and PHWR, but also useful for spent fuel pool accidents analyses.
Besides, main other physical modelling efforts have been focused
on the reflooding of severely damaged cores, Zircaloy oxidation
under air atmosphere, corium coolability aspects during MCCI
and RCS gas phase chemistry kinetics in accordance to the main
lessons drawn from the V2.0 extended assessment. Dedicated
improvements have also addressed iodine chemistry in contain-
ment according to some recent better understanding of gaseous
iodine formation processes. Finally, the treatment of FP/SM trans-
port and chemistry has been harmonised between circuits and
containment domains in order to provide a more consistent mod-
elling with respect to the source term evaluation.
As concerns more particularly BWR, the V2.1 version exhibits a
significant progress with respect to former V2.0 versions. Thanks to
an improved representation of the real core geometry, it is now
possible to perform simulations of complete BWR SA sequences
as it was already the case with older versions for PWR. Modelling
efforts should now focus on the relevance of the ICARE physical
models according to BWR specifics. Furthermore, thanks to the
BARC partner involvement, a similar progress is also obtained for
PHWR. It is now possible to entirely simulate the LCDA phase with
the V2.1 version thanks to an improved representation of the real
core geometry combined with the implementation of dedicated
thermo-mechanical models. Next PHWR modelling steps will
specifically address the SCDA phase.
Moreover, ASTEC is providing the back bone of the on-going
CESAM FP7 European project that aims at the improvement of
the ASTEC code for use in SAM analysis of the nuclear power plants
of Generation II–III presently under operation or foreseen in near
future in Europe. In that frame, the assessment of ASTEC models
important for SAM, in particular considering the dominant phe-
nomena in Fukushima accidents (reflooding of degraded cores,
pool scrubbing, hydrogen combustion, spent fuel pools
behaviour....) is being done. Main CESAM other target is the exten-
sion of ASTEC for support to diagnosis, notably by interfacing with
atmospheric dispersion codes in order to enhance capabilities of
direct comparison with on-site measurement.
Besides, while current efforts towards a progressive industrial-
isation of the code will be intensified, ASTEC shall remain a repos-
itory of R&D knowledge for SA phenomenology. For that purpose,
the feedback from the interpretation of the current and future
experimental programs performed in the international frame will
be continuously taken into account: reflooding of severely dam-
aged cores according to PEARL new data; iodine and ruthenium
chemistry in RCS and in containment according to CHIP+, STEM/
STEM2, BIP-2/BIP-3 and THAI-2/THAI-3 projects; corium/debris
behaviour in the lower head according first to the on-going COR-
DEB project between French partners and NITI (Russia) and then
to the starting IVMR H2020 EURATOM project on power reactors
safety (IVMR, 2015; Fichot et al., 2015); in-vessel and ex-vessel
corium behaviour according to various new data to be produced
through the European SAFEST FP7 project (Miassoedov et al.,
2015); hydrogen distribution, combustion and recombination
according to the OECD THAI-2/THAI-3 and French ANR-
MITHYGENE (Bentaib et al., 2014) projects; pool scrubbing and
mitigation according to the French ANR-MIRE (Cantrel et al.,
2015) and European PASSAM projects (Albiol et al., 2015); corium
coolability during MCCI according to the last CCI experiments...
As concerns other types of applications, both modelling and
assessment activities addressing the ASTEC adaptations to Gen.IV
reactors, especially sodium-cooled fast neutron reactors, and the
treatment of accidents in the ITER Fusion installation will continue.
Acknowledgments
The authors would like to associate the whole IRSN ASTEC
development team to this paper, according to the significant con-
tribution of each developer in the build-up, validation and testing
of the V2.1 new version.
References
Albiol T., Herranz L.E., Riera E., Guieu S., Lind T., Morandi S., Karkela T., Losch N.,
Azambre B., 2015. The European PASSAM project on atmospheric source term
mitigation: half-way progress and main results. In: Proc. International OECD-
NEA/NUGENIA-SARNET Workshop on the Progress in Iodine Behaviour for NPP
Accident Analysis and Management. Marseille (France), March 30–April 1.
Bachrata A., 2012. Modélisation du renoyage d’un cœur de réacteur nucléaire
fortement dégradé (Ph.D. thesis). INP Toulouse (France).
Bartonicek B., Habersbergerova A., 1987. Formation of methyl iodide by ionizing
radiation. In: Proc. 4th working meeting on Radiation Interaction, Leipzig, pp.
491–495.
Bentaib A., Studer E., Chaumeix N., Magne S., Reinecke E., Pinet D., 2014.
MITHYGENE project Towards the improvement of hydrogen risk assessment
models and safety management procedures. In: Proc. International Workshop
on Nuclear Safety and Severe Accident (NUSSA-2014), Chiba (Japan), September
3–5.
Beuzet E., Haurais F., Bals C., Coindreau O., Fernandez-Moguel L., Vasiliev A., Park S.,
Stuckert J., Steinbrück M., Hozer Z., Duriez C., 2015. Cladding oxidation during
air ingress. Part II: synthesis of modeling results. In: Proc. 7th European Review
Meeting on Severe Accident Research (ERMSAR 2015), Marseille (France),
March 24–26.
Bosland L., Cantrel L., 2015. Iodine behaviour in the circuit and containment:
modelling improvements in the last decade and remaining uncertainties. In:
Proc. International OECD-NEA/NUGENIA-SARNET Workshop on the Progress in
Iodine Behaviour for NPP Accident Analysis and Management, Marseille
(France), March 30–April 1.
Cantrel L., Louis F., Cousin F., 2013. Advances in mechanistic understanding of
iodine behaviour in Phébus FP tests with the help of ab initio calculations,
92 P. Chatelard et al. / Annals of Nuclear Energy 93 (2016) 83–93
Annals of Nuclear Energy: Special issue on Phebus FP final seminar, vol. 61,
November 2013, p. 170–178.
Cantrel L., Herranz L.E., Guieu S., Albiol T., Collet R., Lind T., Karkela T., Mun C.,
Jacquemain D., Chebbi M., 2015. Overview of on-going and planned R&D works
on delayed releases and FCVs efficiencies. In: Proc. International Congress on
Advances in nuclear Power Plants (ICAPP-2015), Nice, France, May 3–6.
Carman, P., 1937. The determination of the specific surface area of powder. I. J. Soc.
Chem. Ind. 57, 225–234.
CESAM: Code for European Severe Accident Management, 2013. EC (European
Commission) 7th Framework Programme. Grant agreement 323264, February
2013.
Chatelard, P., Reinke, N., Arndt, S., Belon, S., Cantrel, L., Carénini, L., Chevalier-Jabet,
K., Cousin, F., Eckel, J., Jacq, F., Marchetto, C., Mun, C., Piar, L., 2014a. ASTEC V2
severe accident integral code main features, current V2.0 modeling status,
perspectives. Nucl. Eng. Des. 272, 119–135.
Chatelard, P., Arndt, S., Atanasova, B., Bandini, G., Bleyer, A., Brähler, T., Buck, M.,
Kljenak, I., Kujal, B., 2014b. Overview of the ASTEC V2.0-rev1 validation. Nucl.
Eng. Des. 272, 136–151.
Chatelard, P., Reinke, N., Ezzidi, A., Lombard, V., Barnak, M., Lajtha, G., Slaby, J.,
Constantin, M., Majumdar, P., 2014c. Synthesis of the SARNET FP7 WP4
activities on ASTEC topic: focus on ASTEC V2 plant applications. Ann. Nucl.
Energy 74, 224–242.
Chatelard P., Chevalier-Jabet K., Van Dorsselaere J.P., Nowack H., Herranz L.E., Pascal
G., Sanchez-Espinoza V.H., 2014. The CESAM European project on ASTEC code
improvements towards simulation of severe accident management. In: Proc.
International Workshop on Nuclear Safety and Severe Accident (NUSSA-2014),
Chiba (Japan), September 3–5.
Chevalier-Jabet, K., Cousin, F., Cantrel, L., Seropian, C., 2014. Source Term
assessment with ASTEC and associated uncertainty analysis using SUNSET.
Nucl. Eng. Des. 272, 207–218.
Chevalier-Jabet K., Cousin F., Bosland L., Cantrel L., Créach V., Fougerolle S., Denis J.,
Veilly E., 2015. Iodine Source Term computations with ASTEC, link with PSA
Level 2 tools and fast running Source Term tools for emergency organization. In:
Proc. International OECD-NEA/NUGENIA-SARNET Workshop on the Progress in
Iodine Behaviour for NPP Accident Analysis and Management, Marseille
(France), March 30–April 1.
Chikhi N., Garcin T., Foubert F., March P., Fichot F., 2015. First experimental results
of large scale debris bed reflood tests in the PEARL facility. In: Proc. 16th
International Topical Meeting on Nuclear Reactor Thermal-Hydraulics
(NURETH-16), Chicago, Michigan (USA), Aug. 30–Sept. 04.
Chikkhi, N., Fichot, F., 2010. Reflooding model for quasi-intact rod configuration:
quench front tracking and heat transfer closure laws. Nucl. Eng. Des. 240, 3387–
3396.
Clement B., Zeyen R., 2005. The Phébus FP and international source term
programmes, In: Proc. International Conference on Nuclear Energy for New
Europe (NENE-2005), Bled (Slovenia), September 5–8.
Coindreau, O., Ederli, S., Duriez, C., 2010. Air oxidation of Zircaloy-4 in the 600 °C–
1000 °C temperature range: modelling for ASTEC code application. J. Nucl.
Mater. 405, 207–215.
Cranga M., Mun C., Bottin M., 2012. Analysis of top quenching models during MCCI:
status of knowledge and proposals of developments in the ASTEC/MEDICIS
code. SARNET2 report MCCI-P12, June 2012.
Di Giuli M., Haste T., Biehler R., Bosland L., Herranz L.E., Fontanet J., Beuzet E.,
Torkhani M., Davidovich N., Klein-Heßling W., Weber S., Dickinson S., Horváth
G., Kruse P., Koch M., Paci S., Weber S.J., Salay M., Bujan A., Ivanov I., Kalychev P.,
Kim S.B., Morandi S., Del Corno A., Kotouc
ˇM., Dienstbier J., Kim H.-C., 2016.
SARNET benchmark on Phébus FPT3 integral experiment on core degradation
and fission product behaviour. Ann. Nucl. Energy 93, 65–82.
Drai P., Marchand O., Chatelard P., Fichot F., Fleurot J., 2007. Improvement of core
fusion modelling in ICARE/CATHARE. Application to the calculation of a six-
inches LOCA leading to severely degraded situation. In: Proc. 12th International
Topical Meeting on Nuclear Reactor Thermal-Hydraulics (NURETH-12),
Pittsburgh, Pennsylvania (USA), Sept. 30–Oct. 4.
Drai P., Chatelard P., Laborde L., Piar L., Fichot F., Brillant G., 2015. Capability of
ASTEC V2.1 to simulate a severe accident in a nuclear power plant – application
to the TMI-2 accident scenario. In: Proc. 16th International Topical Meeting on
Nuclear Reactor Thermal Hydraulics (NURETH-16), Chicago, IL, USA, August 30–
September 4.
Epstein M., 1999. Review of water ingression models, prepared by Fauske Associates
for EPRI. Technical report, ACEX (Draft), September 1999.
Ergun, S., 1952. Fluid flow through packed columns. Chem. Eng. Prog. 48, 89–94.
Fichot F., Bonnet J.M., Chaumont B., 2015. IRSN views and perspectives on in-vessel
melt retention strategy for severe accident mitigation, EUROSAFE Forum 2015,
Brussels (Belgium), Nov. 2–3.
Funke F., 1999. Data analysis and modelling of organic iodide production at painted
surfaces. OECD Workshop on Iodine Aspects of Severe Accident Management,
Vantaa, Finland, May 18–20, pp. 151–168.
Funke F., Gupta S., Weber G., Langrock G., Poss G., 2015. Interaction of gaseous I2
with painted surfaces and aerosols in large-scale THAI tests. In: Proc.
International OECD-NEA/NUGENIA-SARNET Workshop on the Progress in
Iodine Behaviour for NPP Accident Analysis and Management, Marseille
(France), March 30–April 1.
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: Proc. International Conference on
Advances in nuclear Power Plants (ICAPP 2013), Jeju Island (Korea), April 14–18.
Glowa G., Moore C.J., Boulianne D., 2015. The main outcomes of the Behaviour
Iodine Project. In: Proc. International OECD-NEA/NUGENIA-SARNET Workshop
on the Progress in Iodine Behaviour for NPP Accident Analysis and
Management, Marseille (France), March 30–April 1.
Gouello, M., Mutelle, H., Cousin, F., Sobanska, S., Blanquet, E., 2013. Analysis of the
iodine gas phase produced by interaction of CsI and MoO
3
vapours in flowing
steam. Nucl. Eng. Des. 236, 462–472.
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: Proc. International Conference on Nuclear Energy for New
Europe (NENE-2012), Ljubljana (Slovenia), September 5–7.
IVMR: In-Vessel Melt Retention Severe Accident Management Strategy for Existing
and Future NPPs, 2015. European Union’s ‘‘Horizon 2020” Framework
Programme for Research and Innovation, Grant agreement 662157, April 2015.
Lipinski R.J., 1982. A model for boiling and dry-out in particle beds. NUREG/CR-
2646, SAND82-0765, Prepared by SNL, June 1982.
Lomperski, S., Farmer, M.T., 2007. Experimental evaluation of the water ingression
mechanism for corium cooling. Nucl. Eng. Des. 237, 905–917.
Majumdar P., Chatterjee B., Mukhopadhyay D., Lele H.G., Guillard G., Fichot F., 2012.
ASTEC adaptation of PHWR module for limited core damage accident analysis.
In: Proc. 5th European Review Meeting on Severe Accident Research (ERMSAR
2012), Cologne (Germany), March 21–23.
Majumdar P., Lele H.G., Belon S., Fichot F., 2013. Analysis of large break LOCA with
simultaneous loss of power supply in 220 MWe Indian PHWR using ASTEC. In:
Proc. 6th European Review Meeting on Severe Accident Research (ERMSAR
2013), Avignon (France), October 2–4.
Miassoedov A., Journeau, C. Bechta S., Hozer Z., Manara D., Kiselova M., Langrock G.,
Schyns M., 2015. Severe accident facilities for European safety targets. The
SAFEST project. In: Proc. 7th European Review Meeting on Severe Accident
Research (ERMSAR 2015), Marseille (France), March 24–26.
Mun. C., Bosland L., Cantrel L., Colombani J., Leroy O., Ohnet M.N., Albiol T., 2015.
OECD-STEM project and its follow-up STEM2. In: Proc. International OECD-NEA/
NUGENIA-SARNET Workshop on the Progress in Iodine Behaviour for NPP
Accident Analysis and Management, Marseille (France), March 30–April 1.
Repetto, G., Garcin, T., Eymery, S., Fichot, F., 2013. Experimental program on debris
reflooding (PEARL) – results on PRELUDE facility. Nucl. Eng. Des. 264, 176–186.
Robert M., Farvacque M., Parent M., Faydide B., 2003. CATHARE2 V2.5: a fully
CATHARE version for various applications. In: Proc. 10th International Topical
Meeting on Nuclear Reactor Thermal-Hydraulics (NURETH-10), Seoul (Korea),
Oct 5–9.
SARNET: Network of Excellence for a Sustainable Integration of European research
on Severe Accident Phenomenology and Management – Phase 2, 2008. EC 7th
Framework Programme, Grant agreement 231747, April 2008.
Tourniaire, B., Seiler, J.M., Bonnet, J.M., Amblard, M., 2006. Liquid ejection through
orifices by sparging gas, PERCOLA models. Nucl. Eng. Des. 236, 2281–2295.
Van Dorsselaere J.-P., Auvinen A., Beraha D., Chatelard P., Journeau Ch., Kljenak I.,
Miassoedov A., Paci S., Tromm Th.W., Zeyen R., 2011. Some outcomes of the
SARNET network on severe accidents at mid-term of the FP7 project. In: Proc.
International Conference on Advances in nuclear Power Plants (ICAPP-2011),
Nice, France, May 2–5.
Van Dorsselaere, J.P., Auvinen, A., Beraha, D., Chatelard, P., Herranz, L.E., Journeau,
Ch., Klein-Hessling, W., Kljenak, I., Miassoedov, A., Paci, S., Zeyen, R., 2015a.
Recent severe accidents research: synthesis of the major outcomes from the
SARNET network. Nucl. Eng. Des. 291, 19–34.
Van Dorsselaere J.P., Chatelard P., Chevalier-Jabet K., Nowack H., Herranz L.E., Pascal
G., Sanchez-Espinoza V.H., 2015. ASTEC code development, validation and
applications for severe accident management within the CESAM European
project, In: Proc. International Congress on Advances in Nuclear Power Plants
(ICAPP-2015), Nice (France), May 3–6.
P. Chatelard et al. / Annals of Nuclear Energy 93 (2016) 83–93 93
... Besides an extension of the existing experimental database on existing and innovative filtration systems, the focus was put on trying to get a deeper understanding of the phenomena underlying their performance and to develop models/correlations that allow modelling of the systems in severe accident analysis codes, like ASTEC (Chatelard et al., 2015). ...
... Globally, in-depth analysis of the experimental results allowed a deeper understanding of the phenomena involved in the performance of the mitigation systems studied, and simple models or correlations which should be easy to implement in accident analysis codes, like ASTEC (Chatelard et al., 2015) could be proposed for several PASSAM experimental studies. ...
Article
The PASSAM (Passive and Active Systems on Severe Accident source term Mitigation) project was launched in the frame of the 7th framework programme of the European Commission. Coordinated by IRSN, this four year project (2013–2016) involved nine partners from six countries: IRSN, EDF and university of Lorraine (France); CIEMAT and CSIC (Spain); PSI (Switzerland); RSE (Italy); VTT (Finland) and AREVA GmbH (Germany). It was mainly an R&D project of experimental nature aimed at investigating phenomena that might enhance source term mitigation in case of a severe accident in a Nuclear Power Plant (NPP). Both existing systems (i.e., water scrubbing and sand bed filters plus metallic pre-filters) and innovative ones (i.e., high pressure sprays, electrostatic precipitators, acoustic agglomerators and, advanced zeolites and combined wet-dry filtration systems), were experimentally studied in conditions as close as possible from those anticipated for severe accidents. This paper presents the main experimental results of the project which represent a significant extension of the current database on these existing or innovative mitigation systems. Application of some of these data for improving existing models or developing new ones should eventually enhance the capability of modelling Severe Accident Management measures and developing improved guidelines
... under the 7 th Framework programme [2], which included source term studies [3]. Phébus FP and associated studies [4] provide outstanding insights into fission product release and transport and, particularly, containment iodine chemistry, which are encapsulated in recent versions of integral severe accident analysis codes like ASTEC 2.1 [5], MELCOR 2.1 [6] and MAAP-EDF [7], while data from international projects such as the International Source Term Project (ISTP) [8], EC/PASSAM [9] and Organisation for Economic Cooperation and Development (OECD)/BIP&BIP2, THAI&THAI2, and STEM [1] are being interpreted with a view to further code improvements. Nevertheless, some issues still remain and need addressing [10]. ...
... and EURATOM, e.g. the SARNET Network of Excellence under the 7 th Framework programme [2], which included source term studies [3]. Phébus FP and associated studies [4] provide outstanding insights into fission product release and transport and, particularly, containment iodine chemistry, which are encapsulated in recent versions of integral severe accident analysis codes like ASTEC 2.1 [5], MELCOR 2.1 [6] and MAAP-EDF [7], while data from international projects such as ISTP [8], EC/PASSAM [9] and OECD/BIP&BIP2, THAI&THAI2, and STEM [1] are being interpreted with a view to further code improvements. Nevertheless, some issues still remain and need addressing [10]. ...
Article
The integrated ICE (Ingress-of-Coolant Event) facility, scaled 1/1600 with respect to the ITER-FEAT design, was built at JAERI with the aim of reproducing the phenomenology occurring in an ICE accident. An ICE occurs when a rupture in the coolant pipes causes the pressurized coolant to enter into the Plasma Chamber, which is held under high vacuum condition. A suppression system is used to mitigate the overpressurization and to prevent mechanical damages to the structures. The CPA module of the ASTEC severe accident code (Study carried out with ASTEC V2, IRSN all rights reserved, [2020]), has been adopted for the modelling and the simulation of a test conducted in the ICE facility. The experimental results of the main thermal-hydraulic parameters have been compared to the code results to characterize the ASTEC capability to predict the phenomenology of a low-pressure two-phase flow transient occurring in a fusion reactor. By coupling the ASTEC code with the uncertainty tool RAVEN, developed by Idaho National Laboratory, an uncertainty analysis has been conducted on the transient. The aim of the present activity is to investigate the dispersion and the sensitivity of the code response to the variation of selected uncertain input parameters, which could influence the simulation of an ICE. The activity also provides a first application of uncertainty analysis through the RAVEN-ASTEC coupling.
Conference Paper
Full-text available
The 4-year JASMIN collaborative project, involving 9 organizations, was launched by IRSN end of 2011 within the 7th European R&D Framework Programme on the enhancement of Na-cooled Fast Neutron Reactors (SFR) safety for a higher resistance to severe accidents. The project aims at developing a new European simulation code, ASTEC-Na, with a modern architecture, sufficiently flexible to account for innovative reactor designs and eventually new types of fuel and claddings and accounting for results of recent research outcomes on water-cooled reactors. The code will be based on the ASTEC European code system, developed by IRSN and GRS for severe accidents in water-cooled reactors, and it will integrate and capitalize the state-of-the-art knowledge on SFR accidents through the improvement of existing physical models or the development of new ones. The code objectives will be to predict throughout the primary phase of the accidental sequence the cladding and fuel behaviour as well as the behaviour of the released fission products both in the primary circuit and in the containment vessel, including the extreme thermal-hydraulic conditions prevailing in case of Na fire. The main involved phenomena that will be investigated during the project include fuel element heat-up and failure, fuel-coolant-interaction, fuel dispersion or compaction, neutronics of the degraded core, fission product retention by sodium pool scrubbing, sodium aerosol depletion and physico-chemical transformations in the containment vessel. The first 18 month period of the project is mainly dedicated to build model specifications as well as code validation matrices related to fuel pin degradation in transient events and in-containment phenomena. The developed models will be validated as far as possible on existing in-pile (mainly the past CABRI experiments) and out-of-pile experimental data.
Conference Paper
Full-text available
Severe accident (SA) phenomenology related to iodine release to the environment has been closely investigated at IRSN over the last decades, by conducting semi-integral or analytical experimental programs like PHEBUS FP, ISTP (CHIP, EPICUR) and OECD/STEM and contributing to others such as OECD/BIP and THAI. These programs have led to the improvement of knowledge and modelling of iodine behavior in the reactor coolant system and the containment in a SA at a nuclear power plant (NPP). Computational tools developed at IRSN for source term (ST) assessment have benefited from these improvements. Three kinds of codes are developed and used at IRSN for different purposes: - The Accident Source Term Evaluation Code, ASTEC, is a dedicated tool used to simulate and predict accident evolution from the initiating event to the releases to the environment with a detailed modelling of the accident phenomenology tackling the coupling between various processes. The accuracy of the prediction depends on the status of knowledge and uncertainties of the involved processes; The fast running code MER is developed in order to get a rapid assessment of releases for Level 2 Probabilistic Safety Assessment. MER provides the associated iodine release for a large number and variety of accident situations, with a reasonable accuracy. - In the field of emergency response, the fast running tool PERSAN provides a relative conservative assessment of the releases to the atmosphere, also for a wide range of conceivable situations. The short calculation times needed for MER and PERSAN to achieve an evaluation of a release are obtained by imposing some features of the accident progression scenario, evaluated aside, and by using a simplified modeling. After a presentation of these three source term assessment tools, the article explains the implemented methodology at IRSN to build suitable models for each of the three applications. Simplifications are driven by the final need of realism or conservatism, using uncertainty modeling when necessary. Calculation examples are given for MER and PERSAN.
Conference Paper
Full-text available
The STEM (Source Term Evaluation and Mitigation) OECD project operated by IRSN, has been launched mid-2011 in order to improve, in the event of a severe accident (SA) on a nuclear power plant, the evaluation of Source Term (ST) and to reduce uncertainties on specific phenomena dealing with the chemistry of two major fission products: iodine and ruthenium. These data are also useful to design mitigation means like FCVS or others. More precisely, the STEM project consists in providing additional knowledge and improvements of calculation tools in order to allow a more robust diagnosis and prognosis of the progression of a SA, and a better evaluation of potential releases of radioactive materials. Initially, two phases were defined at the expert meeting hold in October 2010: the first one from mid-2011 to mid-2015 and the second one (STEM2) up to mid-2018. The first phase addressed three main issues: i/ middle-term iodine releases with specific attention to the stability of iodine aerosol particles under radiation (transformation into gaseous iodine species), ii/ short and middle-term iodine-paint interactions under irradiation, iii/ ruthenium transport chemistry in order to determine the speciation of Ru, in particular the partition between gaseous and condensed forms, during its transport through the Reactor Cooling System (RCS). This paper presents the main outcomes of the STEM first phase. Together with recent results of other R&D programs in the field, notably the OECD/BIP2, STEM results helped defining more precisely major remaining issues that should be investigated during the second phase of the project to improve significantly ST evaluations in SA. The paper provides a description of the updated proposal for the second phase of the project that will be conducted from mid-2015 to mid-2019.
Conference Paper
Full-text available
Within the past decade, international experimental programs were performed in the area of iodine behaviour in the Reactor Coolant system (RCS) and containment during a severe accident (SA) in a Nuclear Power Plant (NPP). The objectives were to better understand involved physico-chemical processes and develop relevant models in SA codes like ASTEC which are used to estimate Source Term (i.e., radioactive releases to the environment). Significant progresses have been made in the modeling of the iodine behavior in the RCS and the containment. This article provides a synthesis of the progresses made in the last decade, what has been understood from the experimental data and which models have been developed, validated and implemented in the ASTEC code. Significant progresses have been made on understanding and modelling of key processes for iodine behavior which gives more confidence on iodine Source Term (ST) assessments and which highlights remaining major source of uncertainties for such assessments. These uncertainties mostly concern processes that have been recently identified as potential important contributors to the releases at mid-term, notably for some NPPs in conjunction with the use of Emergency Containment Filtered Venting Systems (EFCVS). These processes are discussed and are the subject of on-going R&D programs (e.g., OECD-STEM, ANR-MIRE, EC-PASSAM) and future R&D initiatives (e.g., OECD-STEM2, OECDBIP3, OECD-THAI3) in the ST field.
Conference Paper
Deposition of gaseous molecular iodine (I2) onto dry or wet surfaces in nuclear power plant containments during severe accidents represents an important sink to reduce iodine volatility. Decontamination paint, covering large surface areas of walls and floors, is well-known to provide the necessary high reactivity towards I2. Aerosols, with their high specific surface of small particles, can significantly increase the I2 deposition surface area. Models to predict the I2 depletion from atmospheres by interaction with painted surfaces are available. These are commonly based on results obtained from small-scale tests including volumes of only up to about 0.3 m3. Up to now no models have been included in severe accident containment codes to quantify the I2 depletion by adsorption onto aerosols. The 60 m3 vessel of the THAI test facility is used to bridge the differing scales of laboratory tests and containment dimensions. At the same time coupled effects such as I2 / aerosol interaction and iodine deposition in aerosol form have been studied in the large-scale THAI facility. The reaction of gaseous I2 with painted surfaces was addressed in a number of THAI tests (Iod 15, Iod 17, Iod 20, Iod 21, Iod 24, Iod 27A, Iod 28, and Iod 30). An empirical deposition / desorption model was developed and included in COCOSYS/AIM 3, but it still needs more refinements: (1) paint is only a German epoxy paint, and there is no connection to compare its behaviour versus iodine with the effects of other paints, (2) aging of paint is not explicitly scaled but always referring to 15 years, (3) aging procedure is different from other experimental programs and a deeper analysis of this issue including connecting tests is necessary, (4) modeling of relative humidity and temperature effect on chemisorption are to be developed, (5) the partitioning between physisorbed and chemisorbed iodine on paint is to be checked and adapted to the high fraction of chemisorbed iodine measured. This issue could entrain the necessity of further tests in the laboratory-scale. The reaction of gaseous I2 with wet painted surfaces was studied with tests Iod 21 and Iod-24 and the expected phenomena of I2 deposition and wash-off were quantified. A new water-film-based model, which explicitly considers detailed condensation rate, and which considers the iodine chemistry in the water film, was designed and preliminary adjusted to THAI test Iod 24. More analytical and validation work is necessary for optimisation. The two THAI tests on the I2 / aerosol reactions (Iod 25, Iod 26) are providing the effect of I2 removal from containment atmospheres by interaction with containment aerosol, the reactive silver (Ag) aerosol providing a high reactivity, and the inert tin oxide (SnO2) a low reactivity under the given boundary conditions of the tests. The I2 chemisorption reaction with the Ag particle surface is as high as for the higher range of I2 / paint reactions, as evident from comparison of large-scale THAI tests and associated laboratory-scale tests. Real mixed aerosol in the containment can be expected to exhibit I2 removal rates between the bounding cases "reactive Ag" and "inert SnO2". If source term analyses reveal a significant effect of the I2 / aerosol interaction in the containment, the actual surface reactivity of an aerosol defined by the thermochemical phases available to gaseous I2 lays between the bounding cases "reactive Ag" and "inert SnO2". Additional tests using a more realistic aerosol (e.g. from simulated core melts) could turn out necessary, as well as the study of an effect of steam.