Sustainable integration of EU research in severe accident
phenomenology and management
Jean-Pierre Van Dorsselaere1, Thierry Albiol, Bernard Chaumont (IRSN, France), Tim Haste
(PSI, Switzerland), Christophe Journeau (CEA, France), Leonhard Meyer (KIT, Germany),
Bal Raj Sehgal (KTH, Sweden), Bernd Schwinges, David Beraha (GRS, Germany),
Alessandro Annunziato (JRC-Ispra), Roland Zeyen (JRC-IE)
1
Institut de Radioprotection et de Sûreté Nucléaire (IRSN), DPAM, B.250, Cadarache, BP3
13115, Saint-Paul-lez-Durance, Cedex, France
SUMMARY AND KEY MESSAGES
In order to optimise the use of the available means and to constitute sustainable
research groups in the European Union, the Severe Accident Research NETwork of
Excellence (SARNET) has gathered 51 organisations representing most of the actors involved
in Severe Accident (SA) research in Europe plus Canada. This project was co-funded by the
European Commission (EC) under the 6th Euratom Framework Programme. Its objective was
to resolve the most important pending issues for enhancing, in regard of SA, the safety of
existing and future Nuclear Power Plants (NPPs).
SARNET tackled the fragmentation that existed between the national R&D
programmes, in defining common research programmes and developing common computer
codes and methodologies for safety assessment. The Joint Programme of Activities consisted
in:
- Implementing an advanced communication tool for accessing all project information,
fostering exchange of information, and managing documents;
- Harmonizing and re-orienting the research programmes, and defining new ones;
- Analyzing the experimental results provided by research programmes in order to elaborate
a common understanding of relevant phenomena;
- Developing the ASTEC code (integral computer code used to predict the NPP behaviour
during a postulated SA) by capitalizing in terms of physical models the knowledge
produced within SARNET;
- Developing Scientific Databases, in which the results of research experimental
programmes are stored in a common format;
- Developing a common methodology for Probabilistic Safety Assessment of NPPs;
- Developing short courses and writing a text book on Severe Accidents for students and
researchers;
- Promoting personnel mobility amongst various European organizations.
This paper presents the major achievements after four and a half years of operation of
the network, in terms of knowledge gained, of improvements of the ASTEC reference code,
of dissemination of results and of integration of the research programmes conducted by the
various partners.
Most initial objectives were reached but the continuation of the SARNET network, cofunded by EC in the 7th Framework Programme (SARNET2 project), will consolidate the first
assets and focus mainly on the highest priority pending issues as determined during the first
period. The objective will be also to make the network evolve toward a complete selfsustainability.
FISA-2009, 22 June 2009, Prague
1
1
INTRODUCTION
Despite the accident prevention measures adopted in the nuclear power plants (NPPs),
some accident scenarios, in very low probability circumstances, may result in severe accidents
(SA) with core melting and plant damage and to dispersal of radioactive material into the
environment, thus constituting a hazard for the public health and for the environment. Large
progress has been reached since the 80’s thanks in particular to the numerous European
actions undertaken within the 4th and 5th Framework Programmes (FP4 and 5) of the
European Commission (EC), but several issues still needed research activities to reduce
uncertainties and consolidate SA management plans, as shown by the outcomes of the
EURSAFE EC FP5 project [1].
Facing the reduction of the national budgets on SA research, it is necessary to better
coordinate the national efforts to optimise the use of the available expertise and experimental
facilities in order to resolve the remaining issues for enhancing the safety of existing and
future NPPs.
2
THE SARNET PROJECT
In April 2004, 51 organizations involved in R&D on SA, including technical safety
organisations (TSO), industries, utilities and universities, decided to network in SARNET
(Severe Accident Research NETwork of Excellence), in the framework of FP6, their
capacities of research in the SA area in a sustainable way [2]. These organizations were
coming from 18 member states of the European Union (Austria, Belgium, Bulgaria, Czech
Republic, Finland, France, Germany, Greece, Hungary, Italy, Lithuania, the Netherlands,
Romania, Slovakia, Slovenia, Spain, Sweden and United Kingdom), Switzerland, Canada and
the Joint Research Centres of the EC.
The general objectives of SARNET consisted of:
 Tackling the fragmentation that exists between the different R&D organizations, notably in
defining common research programmes and developing/qualifying computer tools;
 Harmonizing the methodologies applied for assessing risk and improve Level 2
probabilistic safety assessment (PSA) tools;
 Disseminating the knowledge to newcomers to the European Union more efficiently and
associating them with the definition and the conduct of research programmes more closely;
 Bringing together top scientists to constitute a world leadership in advanced computer
simulation for SA risk assessment.
The network was organised on the basis of a two-level structure: the Governing Board
in charge of strategic decisions, advised by an Advisory Committee on strategic orientations
of the research activities; and the Management Team in charge of the day-to-day technical
administration.
The Joint Programme of Activities (JPA) was broken down in 20 work-packages (Fig.1)
pertaining to three types of activities:
- Integrating activities to strengthen links between organizations;
- Joint research activities to resolve remaining outstanding issues;
- Spreading of excellence activities to diffuse the knowledge.
The joint research activities that constituted the R&D basis of the network aimed at
resolving the priority pending issues. They were split into three areas: corium behaviour,
containment integrity and radiological source term. In all three areas, the same method has
been adopted:
- Review and selection of available relevant experiments,
- Contribution to the definition of test matrices,
FISA-2009, 22 June 2009, Prague
2
-
Synthesis of the interpretation of experimental data,
Integrating activities
Spreading of
excellence activities
WP 1 : ACT
WP 2 : USTIA
WP 9,10,11 : CORIUM
ASTEC Users Support
and Training,
Integration, and
Adaptation
WP 17 : ET
Advanced
Communication Tool
Early phase core degradation
Late phase core degradation
Ex-vessel corium recovery
Education and Training
WP 3 : PHYMA
WP 12,13 : CONTAINMENT
ASTEC PHYsical
Model Assessment
Hydrogen behaviour
Fast Interaction in Containment
WP 6 : IED
Implementation of
Experimental
Database
WP 7 : SARP
Severe Accident
Research Priorities
-
Jointly executed
research activities
WP 4 : RAB
WP 8 : IA
ASTEC Reactor
Application
Benchmarking
Integration
Assessment
WP 5 : PSA2
Level 2 PSA
methodology and
advanced tools
WP 18 : BOOK
Book on severe accident
phenomenology
WP 19 : MOB
Mobility programme
WP 14,15,16 : SOURCE TERM
FP Release and Transport
Aerosol Behaviour impact on Source
Term
Containment Chemistry impact on
Source Term
WP 20 : Management
Benchmark exercises between codes,
Review of models, synthesis and proposals of new or improved models for ASTEC.
Fig.1: SARNET Joint Programme of Activities
3
SCIENTIFIC AND TECHNICAL ACHIEVEMENTS
3.1
Corium phenomena
The corium area ranges from early phase of core degradation to late phase core
degradation and ex-vessel corium stabilization.
Joint activities have been deployed in 22 organizations [3] on the following
experimental programmes: QUENCH-10 on air ingress in bundle geometry [4]; QUENCH-11
on bundle boil-down and quenching; QUENCH-12 on a VVER bundle; LIVE on corium
behaviour in vessel lower head; OECD-CCI on Molten-Core-Concrete-Interaction (MCCI);
COMET-L1 and L2 on MCCI in 2D geometry; VULCANO on MCCI in real materials.
Similar activities have been carried out for ongoing and new Russian projects from the
International Science and Technology Centre (ISTC): PARAMETER project on core top
flooding models, METCOR on the impact of thermo-mechanical interaction on the vessel
behaviour, CORPHAD on the corium thermodynamics.
As a first example, the QUENCH-13 experiment was successfully conducted at the
Karlsruher Institute für Technologie (KIT) on 8 November 2007. It concerned both corium
and source term fields. It was supported by PSI and AEKI regarding aerosol measurements
and by PSI, GRS and EdF regarding calculation support. It investigated the effects of the
presence of a PWR control rod on early-phase bundle degradation and on reflood behaviour
(here by cold water). Production of silver-indium-cadmium aerosols was measured
continuously for the first time, along with speciation at intervals. Two aerosol peaks of short
duration were observed at 11500 s, followed by a sustained period of aerosol transport.
Impactor samples were taken at significant times, for elemental composition analysis. Posttest analyses confirmed the release of cadmium before release of indium and silver (Fig.2).
Another example of joint activities concerned a code benchmark on MCCI at the reactor
scale in a simplified geometry for a siliceous concrete or a limestone concrete. A very large
scattering in axial ablation results was observed in the case of stratified corium pool
configuration (Fig.3). The convective heat transfer coefficient between the two stratified
FISA-2009, 22 June 2009, Prague
3
layers is the key parameter, and only experiments performed at a small scale with simulant
materials are available. This work has underlined the need to continue the analytical work,
based on all available data, in order to get a consistent interpretation of the different
experiments, both with simulant and with real material. There is the need to investigate the
interface structure (crust or not), the interfacial temperature, and the interfacial heat transfer
coefficient. Clearly new experiments in real material are needed.
Fig.2: Analysis of aerosol release in QUENCH-13 experiment
LCS concrete with pool stratification
Evolution of maximum axial ablation depth
LCS concrete pool stratification : cavity shape at 4 days
3
6
UPM
GRS
2
FZK
5
FZK
1
UPM
4
FZK
IRSN
GRS 3,91days
IRSN
0
UPM
IRSN
YCAV, m
Depth, m
GRS
IRSN
3
UPM
-1
-2
GRS
-3
2
-4
1
-5
0
FZK
RCAV, m
-6
0
2
4
6
8
10
12
14
0
1
2
3
4
5
6
7
Time, Days
Fig.3: Reactor benchmark between codes on a stratified pool on a limestone concrete – Left:
axial ablation versus time – Right: cavity shape at 4 days interaction
Other examples of significant achievements [3] are:
 Progress on understanding of the oxidation phenomena in steam and in air, in particular
showing the importance of material composition. In particular data on B4C oxidation
allowed a common interpretation of the integral test Phebus FPT3.
 New series of modelling and analytical work on in-vessel pool behaviour in relation with
the LIVE experiments (KIT).
 Common understanding of the OLHF-1 test (performed at Sandia National Laboratories,
USA) using different models of vessel failure by creep rupture.
FISA-2009, 22 June 2009, Prague
4
 Increased coolability of 2D inhomogeneous debris beds compared to earlier 1D particle
beds. This launched a new interest in this issue both experimentally and numerically, in
particular the debris bed formation and its characteristics of coolability importance.
 Marked ablation anisotropy for silica-rich materials in the 2D MCCI tests, which was an
unexpected result. Interpretation and modelling of this behaviour must be pursued, as well
as knowledge on the processes in case of layered corium pools must be improved.
 Efforts on the development and improvement of the corium thermodynamic and material
physical property databases, mainly NUCLEA thanks to the analysis of EC co-funded
experiments.
3.2
Containment phenomena
The research efforts on energetic phenomena that could potentially threaten
containment integrity concern hydrogen behaviour and fast interactions in the containment.
The main achievements were detailed in [5].
For the former, the hydrogen combustion and the associated risk mitigation were
studied, concentrating on the formation of combustible gas mixtures, local gas composition
and potential combustion modes, including reaction kinetics inside catalytic recombiners.
Hydrogen distribution within the containment was studied to assess the risk of high
concentrations. Experimental programmes on combustion with gradients (ENACCEF at
IRSN) and recombiner kinetics (REKO-3 at FZJ) have been performed and/or are ongoing.
ENACCEF experimental results have been used in a benchmark using different 3D
computational fluid dynamic (CFD) codes (FLUENT, TONUS-3D and REACFLOW). The
results revealed some weaknesses of the combustion models for stratified mixtures from rich
to lean hydrogen concentrations.
As for containment atmosphere mixing, spray experiments were performed in small
scale (TOSQAN at IRSN) and large scale (MISTRA at CEA) facilities (Fig.4). Both
atmosphere depressurization and stratification break-up phenomena were addressed. The
feasibility of simulations using CFD and lumped-parameter codes was demonstrated, which
could also be applied to actual containments, provided that sufficient computing capacities are
available. Besides, CFD simulations of the operation of passive autocatalytic recombiners
(PAR) in simplified 2D containment models represent a first significant step towards
comprehensive simulations of PAR-atmosphere interaction in real plants.
Fig.4: Views of experimental facilities: left, TOSQAN (7 m³) – right: MISTRA (100 m³)
Containment behaviour is strongly affected by condensation, influencing the levels of
pressurization and atmosphere mixing. Simulations of wall condensation CONAN
FISA-2009, 22 June 2009, Prague
5
experiments (University of Pisa) allowed the assessment of different steam condensation
models that are used in CFD codes (Fig.5). The results obtained by the exercise underline the
need to refine CFD tools in accordance with the available experimental information.
Benchmark-1 10
kW
3.5
Calculated Condensation Rate [g/s]
CEA
FzJ-conj_heat_transfer
3.0
FzJ-eq_heat_transfer
FzK
2.5
JRCP
JSI
NRG
2.0
UJV
UNIPI-eq_heat_transfer
1.5
UNIPI-conj_heat_transfer
VEIKI
Experimental Uncertainty
on Condensation Rate < 1%
1.0
0.5
0.0
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
Experimental Condensation Rate [g/s]
Fig.5: Benchmark on CONAN wall condensation experiments
Concerning fast interactions, FCI (Fuel Coolant Interaction) was studied to increase the
knowledge of parameters affecting steam explosion energetics during corium relocation into
water, and determine the risk of vessel or containment failure by investigation of specific
processes like premixing, melt fragmentation and particle heat transfer mode. The work was
closely linked to the phase 1 of the OECD/SERENA programme. Mainly the MC3D (IRSN)
and IKEMIX/IDEMO (IKE) codes have been used for this purpose. In addition a number of
experiments were performed in MISTEE and DEFOR (KTH), and KROTOS (CEA) facilities.
A consensus was reached that in-vessel steam explosion would not induce failure of the
vessel, thus closing the in-vessel steam explosion issue from the risk perspective, and that exvessel steam explosion could induce some damage to the cavity. However, the level of loads
in the latter could not be predicted due to a large scatter of the results. Major reasons of this
scattering were found to be the uncertainties on void distribution in the pre-mixing region,
inducing large discrepancies on the initial conditions of the explosion, and on explosion
behaviour of corium melts, inducing more or less arbitrary tuning of the explosion
parameters. These uncertainties will be addressed experimentally and analytically in Phase 2
of OECD/SERENA programme and of SARNET.
The second issue concerning fast interactions is Direct Containment Heating (DCH).
This includes melt dispersion into various reactor compartments, heat transfer and chemical
processes such as production and combustion of hydrogen. The consequences of DCH are
essentially related to cavity geometry; therefore a database has been established for the plant
types EPR, French PWR-1300, VVER-1000 and the German Konvoi by an experimental
programme performed in the DISCO facilities (KIT). For EPR and VVER1000 plants the
DCH issue can be considered as closed, due to their cavity design. Benchmark exercises have
revealed severe deficiencies in the current modelling in all system codes, mainly their lack of
predictive capabilities. Based on DISCO data, the scaling of combustion of hydrogen jets in
an air-steam-hydrogen atmosphere must be established by applying dedicated combustion
codes (COM3D, REACFLOW). The efforts to improve the predictive capabilities of the CFD
code MC3D and the ASTEC system code, using as test bed the COCOSYS code, must
continue.
FISA-2009, 22 June 2009, Prague
6
3.3
Source term phenomena
In the source term area, fission product (FP) release, transport and deposition were
studied, including air ingress (i.e. influence of an oxidising environment on release and
transport phenomena). The main achievements were detailed in [6].
FP release and transport
Extensive effort was devoted to the experimental International Source Term Programme
(Phebus.FP, ISTP) launched by IRSN, CEA and EDF with the EC support [7]. The
interpretation of available AECL and RUSET (AEKI) data showed that Ruthenium release
occurs in oxide form after an incubation period during which full oxidation of fuel and
cladding occurs. RUSET and VTT tests showed that oxide forms can stay volatile enough in
lower temperature regions to be transported to the reactor containment in a stable volatile
form, which is a very significant result.
The review of two ISTC projects was performed and SARNET proposals on the test
matrices were adopted: VERONIKA experiments on FP release from highly irradiated VVER
fuel, and EVAN experiments on iodine chemistry. FIPRED experiments (INR) provided data
on UO2 pellet self-disintegration.
Concerning FP transport, the interpretation of iodine chemistry in the circuit was based
on data from Phebus FP (IRSN) and VERCORS HT (CEA). Under reducing conditions, and
without absorber material, it seems relatively straightforward, the iodine being transported
mainly as caesium (and rubidium) iodide (CsI). In oxidizing conditions it is more complicated
since iodine can either be principally CsI or tends to form other metal iodides such as with
control rod materials or, if these are not present, conditions become conducive to HI
formation. These statements still need to be confirmed. The experimental CHIP programme
(IRSN) [7] will provide kinetic and thermodynamic data on iodine transport through the
primary circuit, particularly concerning key systems such as {I-Cs-O-H}, which will help in
the scaling to reactor conditions.
The QUENCH-13 new data (cf. section 3.1 on Corium) are being correlated with earlier
analytical tests EMAIC (CEA), through post-test calculations on these and on Phebus FP data.
Aerosols transport and behaviour
The scenarios of special significance for risk were investigated: by-pass sequences
(particularly steam generator tube rupture or SGTR), through-containment cracks and thermal
and mechanical aerosols remobilization.
Several facilities investigated aerosol retention within the steam generator under SGTR
conditions: PSAERO/HORIZON (VTT), PECA/SGTR (CIEMAT) and ARTIST (PSI). These
tests showed that wet scenarios (i.e. with the breach under the secondary side water level)
would provide effective scrubbing of particles, and even dry scenarios could capture a small
fraction of the particles entering the secondary side. The VTT tests showed that resuspension
is important in aerosol retention within horizontal tubes and is enhanced by sudden velocity
changes. All the available resuspension models are being assessed by comparison to data
(new VTT results, JRC Ispra STORM) and with each other. Revolatilisation tests in the smallscale REVAP facility (JRC/ITU) on Phebus FP samples showed that Cs revaporisation can be
very high (~95%) on flat metallic substrates.
Retention of aerosols in containment cracks can be effective, particularly in the
presence of steam as shown in SIMIBE experiments (IRSN) as well as in the EC-funded
PLINIUS/COLIMA-Cracks test (CEA, Demokritos, CESI Ricerca, CIEMAT).
Containment chemistry
The facilities involved were: at bench-scale PARIS (AREVA NP GmbH), EPICUR
(IRSN) [7] (Fig.6 for the latter) and facility at Chalmers University; at intermediate scale
FISA-2009, 22 June 2009, Prague
7
CAIMAN (CEA); at large-scale SISYPHE (IRSN), ThAI (Becker Technology) and Phebus
FP; along with data recently released from RTF P9T3 (AECL). The iodine data book, written
by Waste Management Technology (WMTL) provided an overall critical review of chemistry
data and models.
Fig.6: Schematic diagram of the EPICUR facility at IRSN
Particular attention was paid to two specific issues: effects of radiation on aqueous and
gaseous iodine chemistry, and mass transfer of iodine between aqueous and gaseous phases.
In the first of these, CAIMAN results showed that, in the presence of paints, irradiation and
high temperature, organic iodide can be the dominant form of volatile iodine; in alkaline
conditions, gaseous iodine concentrations decrease by several orders of magnitude. Data from
EPICUR were interpreted cooperatively using codes such as ASTEC, INSPECT and LIRIC,
with results being fed back into definition of the experimental programme, and ASTEC being
improved. Comparison between calculations and EPICUR and CAIMAN data suggests that
the aqueous phase chemistry is reasonably well understood, though some uncertainties
remain. Interpretation of integral experiments such as Phebus FPT2 suggested that radiationinduced conversion of molecular iodine into particulate species (such as iodine oxides) could
be responsible for the gaseous iodine depletion seen in the long-term, but further
improvements in understanding and modelling are still needed.
Mass transfer between sump and gas phase was addressed in SISYPHE: evaporating
conditions increase the transfer rate from the liquid to the gas phase and change the steadystate iodine concentrations, the sump iodine concentration being reduced. A mechanistic
model is now under validation. On the integral scale, cooperative evaluation with several
codes of the ThAI-Iod9 integral test on containment iodine chemistry was completed under
the coordination of GRS. While the iodine behaviour was quite well simulated, the need was
identified for further improvements in modelling of iodine wash-down from the walls,
sump/atmosphere mass transfer, and iodine/steel reactions.
The effect of FP heating on passive autocatalytic recombiners was studied in the benchscale RECI tests (IRSN), which showed that CsI and CdI2 are not stable and yield gaseous
iodine when heated in a humid atmosphere under temperatures representative of recombiner
operation; scaling effects are now being considered by modelling studies with ASTEC and
CFD methods. Finally, ruthenium in-containment behaviour, including irradiation effects, has
been studied experimentally and theoretically by IRSN (further EPICUR tests) and by
Chalmers University, with new models again prepared for ASTEC.
FISA-2009, 22 June 2009, Prague
8
4
INTEGRATION OF ACTIVITIES
The integrating elements constitute key elements of the JPA. The previous section 3
shows that a real integration of the research activities has been achieved thanks to:
- Collaborations on pre and post calculations of experiments: for instance PSI performed
pre and post-test calculations of the KIT QUENCH tests,
- Joint realization of experiments: for instance VTT experimentalists installed and operated
specific instrumentation on the IRSN CHIP experiments,
- Joint definition and interpretation of experiments through many “interpretation circles”
that were created and really active,
- Benchmarking of different codes,
- Exchanges on the application of R&D results to the reactor scale,
- Round-robin exercise on analyses of a prototypic corium sample,
- Yearly technical meetings in each area (corium behaviour, containment integrity and
source term), complemented by a large number of specialists’ meetings.
Indeed a key integration aspect was the creation of the technical circles, each covering a
specific detailed topic; these helped to bring experimenters and modellers closer together,
concerning test definition, interpretation, model development etc... The Fig.7 shows the
example of the circle working on Ruthenium aspects, with 12 different partners.
AECL
VTT
PSI
IRSN
INR
GRS
AECL
AEKI
CEA
INR
IRSN
VTT
Chalmers
AEKI
RU/
RUTH
CEA
EDF
ENEA
FZK
Chalmers
Fig.7: Example of Ruthenium technical circle (black: experiments, red: organisations)
4.1
Advanced Communication Tool (ACT)
The ACT has been implemented for enabling communication between project partners
and for document management. Today's portal technology provides unified support for
efficient collaboration within the network, in particular:
- Access, search, publication of documents and codes (concept of knowledge storage),
- Contact and communicating with partners (interactive and collaborative services),
- Joint co-ordination of actions/programmes (co-operative management of the network),
- List of links to satellite community projects (R&D projects, related sites).
Access is given by web browsers, enabling logging in from any Internet connection.
Around 250 users of SARNET have been granted access to this tool, and the ACT was used
intensively and efficiently: e.g. 500 to 1000 accesses per month, 8000 items stored.
Besides, a public web site (www.sar-net.org) providing major information on SARNET
has been implemented.
4.2
Experimental database (DATANET)
This database has been developed and maintained to ensure preservation, exchange and
processing of SA experimental data, including all related documentation. The data are both
existing experimental data that SARNET partners are willing to share within the network and
all new data produced within SARNET. DATANET is based on the STRESA tool developed
by JRC Ispra [8] and consists of a network with several local databases (or nodes). From the
FISA-2009, 22 June 2009, Prague
9
central database, it is possible to connect with other local databases; direct connections to the
local databases are also possible, which increases the potential and the power of this type of
system. Currently, 7 nodes exist: the central one at JRC Ispra, and local ones at KIT, IRSN,
CIEMAT, Fortum-VTT, AEKI and KTH. Training weeks have been periodically organized at
JRC Ispra. The results of about 190 experiments from 33 facilities have been implemented so
far.
4.3
ASTEC code
The SA integral code ASTEC [9], jointly developed by IRSN and GRS, simulates the
behaviour of a whole NPP under SA conditions, including SA management by engineering
systems and procedures. It is a key integrating component since one of the ultimate goals of
the joint research activities is to provide physical models to be integrated in ASTEC. The
exchange of information on the detailed models developed by the various experts through
interpretation of experiments leads to generic common models in the different detailed codes
(examples of ICARE/CATHARE at IRSN and ATHLET-CD at GRS for core degradation).
Model improvements for the ASTEC code are then derived from these detailed models.
Twenty-seven organizations collaborated with IRSN and GRS on the code development
and assessment. A close and efficient collaboration between ASTEC users and developers has
been set up using ACT and the MARCUS web tool for code maintenance. One workshop on
code use was organized. Three main code versions were released to the partners, the latest
V1.3 rev2 in Dec.07. Three ASTEC users’ club meetings allowed fruitful direct discussions
between users (representing more than 60 qualified scientists) and developers.
Joint analyses of models in the frame of the specialists’ circles already lead to
recommendations for improved or new ASTEC models, such as for example: heat transfer in
the corium in lower head, clad air oxidation and B4C effects on core degradation, lower head
rupture, MCCI, corium spreading, release of silver-indium-cadmium alloys and Ru, iodine
mass transfer in containment, radiolytic oxidation in containment, etc….
Some partners have investigated the needs of model adaptations to other NPP types than
PWR. First, they concluded on the full applicability to Generation II VVER, either 440 or
1000, from the good results of validation (e.g. PACTEL and CORA-W2 experiments) and
benchmarking with other codes on plant sequences. For BWR and CANDU, as a conclusion,
all ASTEC V1 models can already be applied, sometimes with minor adaptations or further
need of validation, except for core degradation. For RBMK, exploratory plant calculations
showed the partial code applicability to the core degradation, as well as the full applicability
to the behaviour of the confinement building (including its Accident Localization System).
A first demonstration of the applicability of models to simulate the In-Vessel Melt
Retention (IVMR) with vessel external cooling was given. Validation on the SULTAN (at
CEA) and ULPU (at University of California, USA) experiments gave good results despite
some remaining oscillations in the calculations, part of them being physical ones
experimentally observed. A coupling between core degradation and thermal-hydraulic models
has been tested with success on a VVER-440 plant calculation for a more detailed IVMR
simulation. Validation will have to continue on the LIVE experiments.
An extensive validation of the code successive versions was carried out by the partners
on 65 experiments (analytical and integral ones), often OECD/CSNI International Standard
Problems. Generally, the results can be ranked as good, even very good for circuit thermalhydraulics and FP/aerosols behaviour. Many plant applications were performed on various
NPPs (PWR, Konvoi 1300, Westinghouse 1000, VVER-440, VVER-1000 and RBMK),
including benchmarks with other codes. The agreement was good with the integral codes
MELCOR and MAAP4 on the trends and orders of magnitude of the main sequence results,
FISA-2009, 22 June 2009, Prague
10
and very good with detailed results of mechanistic codes such as ATHLET-CD and
ICARE/CATHARE for core degradation.
The whole above work allowed a code assessment by users independently from the code
developers. IRSN and GRS have taken into account the feedback of this work on the V1
versions, for instance through a large improvement of the code documentation and of the code
numerical robustness. Several model improvements, from the different topics described
above, have been already implemented by IRSN and GRS. The latest V1.3 version can
currently simulate all types of scenarios for PWR and VVER reactors in operation states,
taking almost all phenomena into consideration, except late reflooding of a degraded core and
air ingress (steam explosion has always been out of the scope of the code). All models are at
the current State of the Art, except the model of reflooding of a degraded core that is still
inadequate like for all codes. All safety systems and SAM for the existing PWR and VVER
can be represented: e.g. voluntary primary circuit depressurization, steam generator
management, containment spray and venting.
IRSN and GRS are now taking into account all users' requirements on ASTEC evolution
for the new series of ASTEC V2 versions that will extend the scope of application to BWR
and CANDU. The 1st of these, V2.0, released in spring 2009, is applicable to the EPR, in
particular its core-catcher, and includes most of the ICARE2 (IRSN code for core
degradation) mechanistic models.
4.4
Level 2 PSA
Level 2 PSA is a powerful tool to assess plant-specific vulnerability regarding NPP SA.
It evaluates possible SA scenarios in terms of frequency, loss of containment integrity and
radioactive release into the environment and quantifies the contribution of prevention and
mitigation measures in terms of risk reduction. Different approaches are used in Europe,
derived more or less from what has been implemented in the USA. A description and
comparison of the main elements of methods used by the different partners to develop their
PSA has been written. For many issues regarding Level 2 PSA, questionnaires have been set
up and the answers have been analyzed in order to define next steps of harmonization. Case
studies for harmonization have then been performed on specific issues like hydrogen
combustion, iodine chemistry, MCCI, large early releases, reactor end states definitions and
methods for level 1 and 2 PSA interface. A certain level of harmonization has been reached
and recommendations have been written. A state on the art report on dynamic reliability
methods has been produced and the limitations of classical methods, which could be exceeded
using these reliability methods, were identified. Examination of the benefit of one of the
possible methods (Monte Carlo Dynamic Event Tree) has been achieved on a station blackout scenario. A benchmark exercise allowed the comparison of dynamic reliability methods
with classical ones by quantifying the risk of containment failure induced by the activation of
safety system during the vessel core degradation phase.
General requirements for the use of integral codes in level 2 PSA have been issued by
some PSA users. The main conclusions are that the ASTEC code characteristics tend to be
similar to the other integral codes (MAAP, MELCOR) used for level 2 PSA. The ASTEC
models are able to simulate most of the SA phenomena in a LWR but two needs were
underlined: modelling the physical processes after vessel failure for EPR and calculating SA
in shutdown states (both improvements are under way in ASTEC V2). The code high
modularity is well suited for level 2 PSA uncertainty or sensitivity studies. The descriptions
of the code input data and of the physical models need to be improved.
FISA-2009, 22 June 2009, Prague
11
5
REMAINING SCIENTIFIC CHALLENGES
The Severe Accident Research Priority (SARP) SARNET Work-Package allowed
identifying research priorities with the aim of re-orientating progressively the existing
national programmes and of contributing to launch new ones in a coordinated way,
eliminating duplications and developing complementarities. Taking as a basis the Phenomena
Identification and Ranking Table (PIRT) carried out in EURSAFE [1], the whole spectrum of
SA situations, extending from core uncovery to long term corium stabilization, long term
containment integrity, and FP release to the environment was considered. Special attention
was brought to a risk-oriented approach in order to really focus on the most relevant pending
issues. As a result, a consensus was reached and 18 main issues were ranked into four
categories.
Six issues have been considered to be investigated further with high priority, the first
three being closely related to SAM essential actions:
- Core coolability during reflood and debris cooling,
- Ex-vessel melt pool configuration during MCCI and ex-vessel corium coolability by top
flooding,
- Hydrogen mixing and combustion in containment, including the effect of mitigation
measures such as recombiner effect on global convection, generation of local stratification
by recombiner, ignition by recombiner,
- Melt relocation into water, ex-vessel FCI,
- Oxidising impact (Ru oxidising conditions/air ingress for high burn-up and MOX fuel
elements) on source term,
- Iodine chemistry in reactor cooling system (RCS) and in containment.
Four issues were assessed with medium priority: these items should be investigated
further as already planned in the different research programmes. The risk significance is
reduced due to considerable progress of knowledge, but some questions are still open:
- Hydrogen generation during reflood and melt relocation in vessel,
- Corium coolability in lower head, i.e. the corium pool behaviour and its thermal loading
on the vessel,
- Integrity of the Reactor Pressure Vessel due to external vessel cooling,
- Direct Containment Heating.
Five issues were assessed with low priority: the current knowledge was considered as
sufficient with regards to the state of knowledge and the risk and safety relevance and taking
into account ongoing activities; they could be closed after the related ongoing activities are
finished:
- Corium coolability in core catcher with external cooling, i.e. the predictability of the
thermal loading on the core catcher device,
- Corium release following vessel rupture,
- Crack formation and leakages in concrete containment,
- Aerosol behaviour impact on source term, in SG tubes and containment cracks,
- Core reflooding impact on source term.
Three issues were marked as “issue could be closed”. Due to the risk significance and
the current state of knowledge, which is regarded as sufficient in comparison to other issues
with greater risk relevance and larger uncertainties, no further experimental programme is
needed:
- Integrity of RCS and heat distribution, i.e. predictability of heat distribution in the RCS,
especially in the SGs, to quantify the risk of RCS failure and possible containment bypass,
- Ex-vessel core catcher behaviour, mainly on corium-ceramics interaction and on cooling
by water bottom injection,
FISA-2009, 22 June 2009, Prague
12
-
FCI including steam explosion in weakened vessel.
This ranking of issues will allow redistribution of competence and manpower on high
priority issues, both in EC FP7 and in other international projects (e.g. OECD projects).
6
EDUCATION AND TRAINING
The first type of activity concerned educational courses on SA, with teachers from the
most experienced organizations, gathering from 40 to 100 persons:
- 1st course on SA phenomenology addressing PhD students and young researchers,
organised and hosted by CEA in January 2006,
- 2nd course on “Accident progression (data, analysis and uncertainties)”, addressing more
experienced nuclear safety specialists, organised and hosted by IRSN in March 2007,
- 3rd course covering both phenomenology and severe accident scenario codes, organised by
CEA and AEKI, hosted by AEKI in April 2008.
A second type of activity was the writing of a textbook on SA phenomenology (the
edition will take place in 2009). It covers historical aspects of light water reactor safety and
principles, phenomena concerning in-vessel accident progression, both early and late
containment failure, FP release and transport. It contains also a description of analysis tools or
codes, of management and termination of SA, as well as environmental management. It gives
elements on Generation III reactors.
A mobility programme offered the possibility to students and researchers to visit
different laboratories of SARNET for training: 33 delegations, with an average duration of 3
months, have been funded by SARNET.
Three conferences (European Review Meetings on Severe Accident Research –
ERMSAR) have been organized in France, Germany and Bulgaria as a forum to the Severe
Accident community. They are becoming one of the major events in the world on this topic.
Finally, at the end of SARNET, about 300 papers have been presented in conferences or
published in scientific journals.
7
PERSPECTIVES IN SARNET NETWORK
The SARNET contract with the EC covered a four and a half year period from April
2004 to September 2008. From 2006 a working group composed of nine representatives of the
Governing Board prepared the continuation of the network. Such new project, named
SARNET2, has been accepted by EC and started to work on April 1st, 2009, for four years.
During this period, the networking activities will focus mainly on the 6 high priority issues
underlined in section 5.
Forty-one partners from Europe, plus Canada, Korea and the United States will network
their research capacities in SARNET2 (Fig.8). In the continuity of SARNET, the project has
been defined in order to optimize the use of the available means and to constitute a
sustainable consortium in which common research programmes and a common computer
code on SA (ASTEC integral code) are developed.
To increase efficiency at the decision-making level, the SARNET Governing Board is
replaced by a Steering Committee of 10 members in charge of strategy, advised by an
Advisory Committee of managers of end-user organisations. A General Assembly, composed
of one representative of each Consortium Contractor, plus the EC representative, will be
called yearly for information and consultation on the progress of the network activities, the
work orientations and the decisions taken by the Steering Committee. On the second level, a
Management Team (Coordinator and seven work-package leaders) will be entrusted with the
task of the day-to-day management of the Network.
FISA-2009, 22 June 2009, Prague
13
The key integrating and spreading of excellence activities will be pursued: gathering of
available experimental data in a common format in the DATANET database; integration of
acquired knowledge in ASTEC and extension of its applicability to BWR and CANDU;
spreading of the knowledge through the public SARNET web site and the ACT; organization
of conferences and seminars; organization of education and training courses; encouragement
of exchange of students and researchers. Besides, research priorities will be periodically
updated. Level 2 PSA activities will not be pursued within SARNET, as the ASAMPSA2 EC
project [10] on harmonization of level 2 PSA studies has been set up in 2008.
Fig.8: SARNET2 FP7 project
Differently to the SARNET first phase, the joint research activities will include
experimental programmes. Following EC recommendations, experimental efforts will be
mainly devoted to two topics considered of highest importance, and for which real progress
toward the closure of the issue is expected: corium/debris coolability and MCCI. Common
analyses of experimental results and benchmarking activities in order to elaborate a common
understanding of relevant phenomena will be pursued through various technical circles.
Finally, the network will evolve toward real self-sustainability, through the creation of a
legal entity. The definition of the most adequate solution will be an essential task of the 1st
SARNET2 period.
Links with other international programmes and organizations
Links with OECD/NEA, ISTC and other programmes co-funded by the EC will be
maintained and reinforced, for instance:
- ASAMPSA2: discussion on level 2 PSA requirements for integral codes such as ASTEC,
and on the sufficiency of knowledge to assess the most relevant phenomena in level 2
PSA,
- PLINIUS transnational infrastructure (CEA): joint performance of experiments and
interpretation by different European teams,
- ENEN [11]: collaboration on education and training courses for students,
- ISTC: discussion in the CEG-SAM experts’ group about the Russian projects,
- SNE-TP: progressive opening of the activities to new generations of reactors.
Discussions will take place on possible cooperation with Third Countries organisations
(including other newcomers in Europe).
FISA-2009, 22 June 2009, Prague
14
8
CONCLUSION
Most objectives of the SARNET network of excellence have been reached:
- Preservation of knowledge produced by thousands of person-years of R&D and
dissemination to end-users through the ASTEC code and the DATANET database,
- First step towards harmonization of level 2 PSA methods, which continues in the
ASAMPSA2 project in FP7,
- Making Europe a leader in SA computer code and risk assessment methodology, in
particular through the assessment and improvements of the ASTEC integral code that is
becoming the European reference tool,
- Efficient transfer of knowledge to younger generations through conferences, courses and
an efficient mobility programme,
- Large progress to solve remaining outstanding issues through joint work on Corium,
Containment and Source Term topics and to provide model recommendations for ASTEC.
- Direct impact of the research priorities on national programmes.
A large step towards the ambitious but highly important objective to achieve the
sustainable integration of the European SA research capacities has been now reached.
Reaching this integration will be the objective of the continuation of the SARNET network
for four years from 2009. Beyond the objective to solve the latest pending issues for current
NPPs safety, the self-sustainability of the network should be obtained at the end of the
SARNET2 project. The network will progressively coordinate all the research activities in
this field, which will contribute to an optimised use of European resources. This unique
kernel of competences should be kept alive and play an increasing role in future on the safety
“cross-cutting” issues in the SNE-TP frame, for instance progressively working on the new
reactor Generation IV designs.
ACKNOWLEDGMENTS
Thanks are given to all the SARNET members: WMTL (formerly AEAT), AEKI,
ARCS, AVN, Budapest University of Technology and Economics (BUTE), CEA, ERSE
(formerly CESI Ricerca), Chalmers University, CIEMAT, CSN, Demokritos, University of
Pisa, EA, EDF, ENEA, Fortum, AREVA NP SAS (formerly Framatome ANP SAS), AREVA
NP GmbH (formerly Framatome ANP-GmbH), FZJ, KIT (formerly FZK), FZR, GRS,
University of Stuttgart (IUSTT-IKE), INR, INRNE, IVS, JRC Ispra, JRC ITU, JRC IE Petten,
JSI, University of Stockholm (KTH), LEI, NNC, NRG, PSI, University of Ruhr Bochum
(RUB), Swedpower AB, Technicatome, Thermodata, Suez-Tractebel SA, Technical
University of Sofia, Université Libre de Bruxelles, Université Catholique de Louvain, UJD,
UJV, Polytechnic University of Madrid (UPM), VEIKI, VTT, VUJE, Becker Technologies,
AECL, BNRA, Newcastle University. The authors are also very grateful to the technical
contributors in these organizations, too numerous to name.
Finally the authors thank the European Commission for funding the SARNET network
in FP6 in the area “Nuclear Fission: Safety of Existing Nuclear Installations” under contract
number FI6O-CT-2004-509065.
REFERENCES
[1] Magallon D. et al., “European Expert Network for the Reduction of Uncertainties in
Severe Accident Safety Issues (EURSAFE)”, Nuclear Engineering and Design, vol. 235,
pp. 309-346 (2005).
FISA-2009, 22 June 2009, Prague
15
[2] Micaelli J.-C. et al., “SARNET: A European Cooperative Effort on LWR Severe Accident
Research”, Proc. European Nuclear Conference, Versailles, France (2005).
[3] Journeau C. et al., “European Research on the Corium Issues within the SARNET
Network of Excellence”, Proc. Int. Congress on Advances in Nuclear Power Plants
(ICAPP’08), Anaheim, California, USA (2008).
[4] QUENCH programme, http://hikww2.fzk.de
[5] Wilkening H. et al.,“European Research on Issues Concerning Hydrogen Behaviour in
Containment within the SARNET Network of Excellence”, ICAPP’08.
[6] Haste T. et al., “SARNET: Integrating Severe Accident Research in Europe: Key Issues in
the Source Term Area”, Proc. Int. Congress on Advances in Nuclear Power Plants (ICAPP
’06), Reno, Nevada, USA (2006).
[7] Clément B. and Zeyen R., “The Phébus FP and International Source Term Programmes”,
Proc. Int. Conf. on Nuclear Energy for New Europe, Bled, Slovenia (2005).
[8] STRESA tool, http://www.asa2.jrc.it/stresa
[9] Albiol T., Van Dorsselaere J.-P., Reinke N., “SARNET: a success story. Survey of major
achievements on severe accidents and of knowledge capitalization within the ASTEC
code”, Proc. EUROSAFE Forum, Paris, France, Nov.2-3 (2008).
[10] ASAMPSA2 project – Advanced Safety Assessment Methodologies: Level 2 PSA,
http://www.asampsa2.eu
[11] European Nuclear Education Network (ENEN), http://www.enen-assoc.org
For general information on safety of nuclear power plants, see the web site of the
EUROSAFE European global approach on safety: http://www.eurosafe-forum.org
FISA-2009, 22 June 2009, Prague
16