The fourth Framework Programme summarised1 May 2000
In its fourth Framework Programme, 1994-1998 the European Union has invested $62.8 million in 67 projects aimed at improving safety.
Research performed in the EU’s fourth Framework Programme (FP-4) reinforces the nuclear safety concept of ‘defence in depth’. The 67 projects represented expenditure of E34.2 million by the EU and E28.8 million by more than 80 contractors from the EU and associated countries (including Switzerland). They are grouped into seven clusters which are each devoted to a key safety issue.
Eleven of the projects, those in the AGE cluster, consider ageing of the multiple barriers (especially the structural components). Their emphasis is on preventing deviations from normal operation and predicting margins of failure.
Some countries now in effect prohibit large-scale evacuation and long-term land contamination. As a result the defence in depth approach is being extended to a fourth level: ‘mitigation of severe accident consequences’. FP-4 includes 45 projects in this sector.
Finally, a cluster of 11 projects (referred to as INNO) cover development of innovative safety concepts for the next generation of plants.
An important objective of the AGE cluster is to understand age-related damage mechanisms and to develop models to predict the behaviour of components and structures, based on methods for the detection and on-line monitoring of degradation. Another important objective is to propose qualified methods for repair, replacement and/or back-fitting of safety-related components.
The research activities of the AGE cluster can be subdivided into projects that deal with three main areas: primary pipework, the pressure vessel and general issues. Progress is as follows:
•BIMET has achieved a better understanding of the safety margins given by fracture mechanics for dissimilar metal welds. Further work is required to confirm these findings on real components.
•DISWEC produced guidelines on the best use of methods to test dissimilar metal welds for their susceptibility to environmentally-assisted cracking.
•In the MODAGE and VORSAC programmes validation work was done on corrosion models and on analytical tools to calculate residual stress distributions, respectively.
Four projects focused on different aspects of irradiation embrittlement. A European code of practice for the reconstitution of irradiated Charpy specimens (used to monitor the embrittlement behaviour of the RPV) was developed within RESQUE. Validation work of different fracture toughness indexes was performed within REFEREE. More work is required in order to study the impact of the neutron spectrum on embrittlement behaviour and to develop a validated conversion table of the different damage indices used. Guidelines for a reference standard dosimeter were developed within AMES-Dosimetry. As a direct result of this project, this type of dosimeter is already used by industry. A pilot database, containing important dosimeter parameters used in different countries, was developed within MADAM.
Risk-informed inspection is becoming a very important technique in maximising the value of inspection and maintenance. EURIS is a utility-driven action within which guidelines for a European framework for risk-informed inspection are being developed. These guidelines will be used as the basis of a further discussion, first between utilities and then between utilities and regulators, in order to verify how they can be implemented in practice.
The use of non-destructive testing (NDT) techniques in order to assess material damage was studied in the AMES-NDT project. Some of the NDT techniques applied, such as thermal power measurements, showed promising results. Further validation work will verify the extent to which they can be applied in service on real components. For other NDT techniques, more development and validation work is required.
The INTACT project works towards making a general integrity assessment of ageing metallic and non-metallic components.
An aspect of nuclear safety which concerns the public is the process of managing severe accidents, and it is in this field that 45 of the FP-4 projects have been carried out. Their joint aim is to add a layer to the philosophy of ‘defence in depth’ that anticipates the likelihood – even if slight – of a severe accident and seeks to mitigate its effect.
In-vessel core degradation
The first cluster of projects, together known as the INV cluster, focuses on the integrity of the first and second barriers under severe accident conditions, with emphasis on the behaviour of the reactor core and on corium formation and behaviour. Those projects have a strong experimental basis with numerical modelling support for severe accident simulation in full-scale reactor conditions.
Late-phase phenomena in degraded core scenarios were investigated in the COBE project, emphasing the swelling of irradiated fuel, low-temperature quenching and hydrogen absorption. Real progress was achieved in applying severe accident codes to the modelling of debris bed and molten pool behaviour.
The air ingress scenario relevance was demonstrated in the OPSA project. An experimental step in preparing the Phebus FP-T5 test is planned for 2005, and devoted to fuel degradation and FP release in air conditions.
Thermodynamic corium properties and chemical interactions between core materials and corium structures were investigated in the CIT project, which also developed and validated thermodynamic databases for in- and ex-vessel corium scenarios.
The SARA project concluded that, for specific degraded BWR core conditions, recriticality is possible during reflooding with unborated water, and recommended various severe accident management schemes (increasing boron injection system capacities, limiting the maximum reflooding rate and delaying primary depressurisation).
For reactor applications, the MFCI (molten fuel coolant interaction) project, using up to 175kg of UO2/ZrO2, showed that a spontaneous steam explosion is very unlikely with prototypical corium. Energetic interactions involving molten reactor material could take place but would not jeopardise reactor integrity. The associated MVI project indicated that the risk of vessel failure due to melt impingement is extremely low. It also concluded that a potential accident management strategy at low power levels is external cooling of the vessel. This helps retain the melt inside the vessel, thanks partly to a focusing effect. This effect was confirmed by the RPVSA project, which demonstrated that the vessel head withstands the effects of a molten slug impact of mechanical energy 0.8GJ caused by a postulated in-vessel energetic MFCI. It also confirmed both the penetration of corium into nozzles over long tube distances and the formation of oxidic melt crusts with the development of a gap, acting as additional insulation, between that crust and the lower head.
Another in-vessel core retention approach, based on an internal core-catcher, was analysed in the IVCRS project (which managed to register several patents). This is in line with requirements for future PWRs established by European utilities and the IAEA.
The project REVISA developed further knowledge on damage, creep behaviour and size effects, concluding for various simulated loss of coolant accident scenarios that a delayed plastic collapse is the dominant vessel failure mode over creep damage.
Ex-vessel corium behaviour
The EXV cluster focuses on the integrity of the basemat, and investigates the behaviour of the core debris in the reactor cavity sump (or for evolutionary reactors on core-catchers).
The emphasis is on spreading, stabilisation and cooling of large quantities of corium as accident management measures for ex-vessel corium mitigation.
In the THMO project thermochemical data for various aerosols generated in molten corium concrete interactions (MCCI) were determined to assess the radioactive source term release and transport.
To support the INV cluster STRATIEX reviewed experimental programmes and numerical models, for both in- and ex-vessel scenarios, to find the conditions required for melt stratification. Corium stratification in the reactor pit plays a crucial role in basemat integrity, because the layers are inverted after the melt interacts with the sacrificial concrete layer. This reduces the liquidus temperature of the corium, improving the way it spreads on the core catcher.
The efficiency of various core-catcher designs, in terms of corium spreading and heat transfer efficiency, were investigated in the CSC project. Experimental and analytical studies focused on corium spreading under dry and wet conditions. Long-term corium coolability was investigated using either flooding on the corium top or water injection from the bottom. The results were as follows:
•For low melt flow rates, spreading was stopped by crust formation for small solidus-liquidus temperature ranges or by viscosity for large ones.
•Material melts presented cracks and a high porosity that aided cooling.
•Flooding the top after spreading never led to energetic melt water interactions.
The COMET-H programme at FZK demonstrated the feasibility of a core-catcher with a bottom cooling approach, based on an array of melt plugs crossing a ceramic layer simulating the sacrificial layer eroded by the melt.
In the COMAS project, several spreading tests using more than two tons of prototypical molten corium were performed for high melt flow rates.
The spreading behaviour was dominated first by inertia and later by viscous forces; mixed melts separated into horizontal metallic and oxide layers even at small density differences, which is crucial for designing sacrificial layers; the front immobilisation was caused by melt bulk freezing and crust formation at the top and front surfaces; and the spreading length was independent from the type of substratum but dependent on the temperature and metallic/oxide ratio for mixed melts. Consequently, it was also concluded for designers that spreading areas can be completely covered by a sufficiently high release melt rate.
Further tests as part of the linked project HTCM showed that an oxide crust is very rapidly formed on the surface of steel structures. This is due to interaction with prototypical melts, and in this way it avoids the problem of erosion. Protecting structures with ZrO2 ceramic covers did not appear to offer any real advantage during that early phase.
Radiological source term
There are 12 projects focusing on the radiological source term in the source term (ST) cluster. They complement six Phebus-FP experiments carried out by an international collaboration at Cadarache.
In-vessel fission product behaviour is being studied under the two projects MP and RSP. The first concentrates on the speciation and kinetics of FP and iodine release from degrading irradiated fuel that was measured during the first two Phebus-FP tests. The second project addresses the release kinetics of FP and core materials from molten pools. One major achievement is the determination of possible UO2 releases from such molten pools, which was essential for the design of Phebus FP-T4, a fuel debris bed test run in July 1999.
In the RVP project, separate-effect simulant studies are being carried out on revaporisation phenomena in the primary circuit. REVAP tests the adequacy of current FP models to cover revaporisation, based on comparisons by using data available from different separate-effect tests. The benchmarking exercise PHEBEN uses Phebus-FP data to validate codes and models concerning aerosol and vapour transport and deposition in the primary circuit and containment. First re-calculations of the Phebus FP-T0 test, run in December 1993 and devoted to melt progression and FP release in a steam rich environment, showed agreement within the limits of experimental uncertainty
Four projects address the ex-vessel FP behaviour. CHEM is investigating vapour-aerosol reactions, aerosol formation and volatile compound generation and has shown the important role of boric acid. IC and OIC are studying the behaviour of aqueous iodine in the containment sump and the formation of organic iodine in the containment atmosphere, respectively.
In the STU plant assessment severe accident scenarios were examined over eight reference designs. It suggested that the main uncertainties in PSA level 2 studies of severe accidents were aerosol behaviour in the containment and late revaporisation of deposited radionuclides.
CONT-LEAK assessed source term results and their applicability to real plants, while ASTERISM is designing an appropriate (pilot) archive for source term data and models.
The CONT cluster is examining containment integrity. Various risk studies have shown that early containment failure due to hydrogen combustion in severe accidents would be the major cause for large off-site contamination, so hydrogen explosions, which are generated by critical concentrations of hydrogen resulting from steam oxidation processes of hot metallic materials during core melting, are being carefully examined.
The CONT projects examine:
• Loads on the containment building, both thermal (including irradiation-induced heat effects) and mechanical (slow pressurisation and fast impacts). This includes containment thermohydraulics, taking into account the heat contents of both steam and radioactive source term to predict slow pressurisation processes and hydrogen deflagration, detonation (supersonic shock wave) and the transition mechanism (VOASM) as well as an impact generated by MFCIs. Initial conditions are provided by the clusters INV, EXV and ST, which are investigating hydrogen generation processes and molten fuel coolant interactions during late-phase core disassembly. Simple engineering correlations for deflagration-to-detonation criteria are particularly useful for optimising the containment design (H2DDT).
• The containment response to these loads, to predict damage, cracking and ultimately breaching of the containment walls. This includes micro- and macro-cracking of the concrete walls subject to pressurisation loads and the subsequent off-containment release rates of steam/aerosol mixtures (CESA), and the thermo-mechanical response of the primary containment penetrations, especially the personnel airlock, characterised by elastomeric sealings (ATHERMIP).
There is no doubt that research efforts related to the ultimate-barrier function of the containment are becoming increasingly important, as the safety requirements become more and more stringent (see the EUCOFA, VASA and ISARRP projects).
The CONT cluster is also examining phenomenological processes (DABASCO) and contributing to the development of engineered passive mitigation measures. This includes the developing passive autocatalytic recombiners – an accident management measure for PWRs that has been implemented in Belgian reactors and was recommended for the German reactors by the German Reactor Safety Commission (RSK) in December 1997. It is currently under discussion by the French nuclear safety authorities.
Accident management measures
The AMM cluster is composed of nine projects .
Three projects address severe accident measures. The ASIA project developed a generic methodology to assess and improve the survival potential of plant process instrumentation (in particular selected neutron flux instruments) under accident conditions, and provided recommendations on the use of simple algorithms for signal validation and accident identification. In the SAMEM project, the existing integrated methods for the assessment of the feasibility and effectiveness of accident measures were improved and applied to a number of case studies. It was concluded that no major further development of the PSA techniques and risk oriented accident analysis methodologies studied is believed necessary. Finally, SAMIME has reviewed the status of severe accident management in the EU and has provided valuable insights for better focusing future R&D activities in the field.
Two projects addressed different aspects of PSA level 2 studies. In BE-EJT “expert judgement” approaches have been documented and benchmarked against both a reference experiment from the FARO facility at JRC-ISIS Ispra and a scenario taken from a generic evolutionary PWR. The PSAL2 project assessed the feasibility of producing a database of information to assist in performing PSA level 2 studies. A pilot database for the hydrogen topic (including some 270 references) has been generated and might be the basis for the construction of a full database that helps to retrieve information in an easy, flexible and fast manner.
Two projects have addressed human factors from different angles. The ISANEW concerted action has concluded that there are available methodologies and elements for “integrated sequence analysis” which can analyse human behaviour, the interaction between human and technical systems, and the dynamic evolution of events. The ORFA project provided a framework to clarify the relationship between organisational factors and nuclear safety. It has concluded that the development and validation of methods and tools for the assessment of organisational factors is necessary and has identified the R&D required.
The EUBORA project concluded that the current status of assessment of the inhomogeneous boron dilution issue in PWRs is not yet complete (in particular mixing and transport phenomena) and proposed new experiments to be performed both in the existing 1/5-scale facilities and in the large scale PANDA facility based at PSI in Switzerland.
The main objective of the JSRI project was to compile information about current research related to nuclear safety aspects of LWRs undertaken in all the EU member states and to present the results as an ‘index’ using an electronic format already well defined. This index contains references for some 500 projects and has been widely disseminated in a CD form within the nuclear scientific community.
Exploring Innovative Approaches
The INNO cluster comprises 11 projects. Eight of them include experimental and analytical activities to study innovative self-acting passive systems for decay heat removal and for safety measures, and demonstrate their feasibility for different types of advanced LWR designs.
The experiments have been performed successfully and have contributed to demonstrating the feasibility and the robustness of most of the passive safety systems under investigation. They have provided a valuable bank of publicly available experimental data about different phenomena and aspects related to:
• Efficiency of pool immersed heat exchangers operating at low pressure (IPSS, POOLTHY, TEPSS).
• Efficiency of condensers with various geometrical tube arrangements for different steam/gas mixtures (IPSS, INCON, CONGA, POOLTHY).
•Behaviour of containment plate condensers (IPSS, INCON).
• Performance of gravity-driven safety injection systems (APSI, TEPSS).
• Performance of single-stage steam injector systems (SYNTHESIS)
• Performance of different types of passive initiators (IPSS).
• Heat transfer degradation due to the deposit of aerosols on the inside (TEPSS) and outside (CONGA) surfaces of heat exchanger tubes.
• Behaviour of aerosols in large water pools (CONGA).
• Natural circulation within a flooded spreading compartment (POOLTHY).
• 3D power distribution under static and dynamic conditions (BWRCA).
The results of some of these projects have been taken into account by the different vendors in the designs of some systems (eg emergency and building condensers of the SWR-1000, isolation condenser and passive containment cooler of the ESBWR) while the data banks have contributed to validate some numerical models in the existing thermal-hydraulic lumped parameter codes (some of them developed 20 years ago). More research is needed to address properly the numerical simulation of many of the above mentioned phenomena.
The rest of the projects of the INNO cluster have provided valuable and strategic recommendations on issues such as: an R&D database on safety-related innovative nuclear reactor technology (SINTER) using an interactive telecommunication platform; industrial R&D needs for next generation reactors (MICA); and the potential of gas cooled high temperature ractors (INNOHTR).
THE NEXT PROGRAMME
The Fifth EURATOM framework programme, 1998-2002, is aimed at improving both the competitiveness and the public acceptance of the nuclear option. In this programme the field of interest will encompass more and more the human and organisational aspects of safety culture.
The emphasis in this series will be on practical aspects of operational safety (in particular the man-machine interface and organisational factors). This will be considered in close relation with current pressures on the industry including market deregulation, new international agreements, public and political pressure, and the reduction of research budgets.