TOWARDS AN OPTIMIZED MANAGEMENT OF ACCIDENTS

Nuclear safety has ever been one of the main research domains in EURATOM programs for decades, and accident prevention and mitigation have drawn much of the attention paid along the years in this framework. In the essence of this interest are designs of reliable systems, accurate methods to estimate risk and a permanent search for optimizing accident management. This is, respectively, the focus of PIACE, MUSA, and AMHYCO projects. A fully passive system for decay heat removal in off-nominal conditions based on the concept of isolation condenser is the subject of PIACE. A harmonized approach for the analysis of uncertainties and sensitivities associated with Severe Accidents (SA) and, particularly, with the Source Term to the environment, is the final aim of MUSA. And, finally, AMHYCO is exploring the potential for an enhancement of the management of combustible gas risk. Despite the projects diversity, they all will converge in the same outcome: an optimization of nuclear safety from better safety systems, risk estimating methods and in-accident guidelines. These projects have received funding from the H2020 EURATOM research and training programme under grant agreements 847715 (PIACE), 47441 (MUSA), 945057 (AMHYCO).


INTRODUCTION
Nuclear safety relies on the Defence-in-Depth approach, which identifies Nuclear Power Plant (NPP) conditions progressively degraded from nominal to postulated severely deteriorated in five levels, and articulate means (standards, engineering safety features and guidelines) to minimize risk to public and environment. For decades, Euratom has brought nuclear safety up as one of the key targets to do research into. With a broad scope, which ranges from testing more reliable and efficient systems to enhancing procedures, guidelines and measures on robust knowledge and know-how, research nuclear safety projects have been awarded under the HORIZON 2020 framework. PIACE, MUSA and AMHYCO are good examples of them.
The high standards set for NPPs gets its top in nuclear safety systems that are supposed to operate in harsh conditions. Their performance needs to be demonstrated before being implemented in any reactor design, particularly, if they are based on an innovative passive approach. The challenge is multiple. On one side, suitable facilities capable of setting representative initial and boundary conditions as well as being scalable to NPPs, should be built and/or adapted, if already existing. On the other, because the system should be properly modelled so that its effect is soundly assessed under any anticipated condition. PIACE integrates all these aspects for a specific proposal to remove decay heat based on the concept of isolation condenser.
Numerical simulation tools are widely used to assess the behaviour of NPPs during postulated accidents including Severe Accident (SA). In other words, they are a central element of the safety demonstration where the compliance of the main safety features of a NPP is checked against the actual safety requirements reflecting the state-of-the-art. In addition, the development and optimization of Accident Management (AM) measures aiming at preventing and mitigating the consequences of SA heavily rely on numerical simulations with SA codes. Since the SA tools predict important parameters such as the time of failure of safety barriers, on one hand, and the potential radiological Source Term (ST) to be released to the environment if the safety barriers fail, on the other hand, it is of paramount importance to be aware of and enhance their accuracy. To do so, MUSA brings uncertainty quantification into SA analyses and provides insights that might change the perception of the current precision of SA codes and highlight areas where further research would be more beneficial.
One of the priority areas of SA research has been risk of energetic combustion of H2 and CO, as it might jeopardize containment integrity. The Fukushima Daiichi accidents occurred in Japan in 2011 just strengthened the need to properly manage such risk and practically eliminate the consequential dynamic phenomena. By keeping the containment integrity, the potential release of significant amount of radioactive material to the environment would be prevented. Despite the numerous research projects and activities conducted in different frameworks, some of their outcomes have not been fully exploited in the form of actions to be taken for accident management, particularly in the ex-vessel phase of accidents when H2 and CO coexist in containment atmosphere. AMHYCO fills in this gap by using a comprehensive approach involving analytical methods, systems testing and SA management procedures.
Euratom has ever paid due attention to communication and dissemination of the research results and outcomes of the projects sponsored in its frame, and this aspect has been particularly emphasized in H2020. Consequently, in each of the above projects there is a specific Work Package (WP) dealing with these aspects. It is a common feature nowadays to host project websites for internal communication, as storing and share-points of the material produced or used as a background in the project. Newsletters, Linkedin and Facebook accounts are among the instruments used by these projects to reach the public. No less important, this WPs foster the education and training aspects by facilitating participation of young researchers and engineers in scientific and technical events as well as financially supporting their mobility all across of European laboratories.

PROJECT OVERVIEW
The PIACE (Passive Isolation Condenser) project will provide a significant contribution to the safety improvement of the present and future technology of nuclear reactors. The project will demonstrate the feasibility and reliability, shortening the time to market, of an innovative Decay Heat Removal (DHR) system, based on an isolation condenser with non-condensable gases, to manage the variable decay heat in passive way, verifying and validating the concept of a patented passively controlled safety system for decay heat removal [1]. The innovative concept has the important peculiarity to be completely passive, by making use of simple physical phenomena, limiting the use of energy sources or movement of mechanical parts only at the actuation stage (category D of [2]) without human intervention, and have the flexibility to be adapted both to the liquid metal and water cooled reactors .   The project has received funding from the Euratom research and training programme  2014-2018 under grant agreement No 847715. The total cost is in the order of 3 M€  with a human resources involvement in the order of 400 person-months, over a 3-year  period (2019-2022). It is an international collaboration which involves 11 European  partners, both research centres and private companies (ANN, EAI, ENEA, GEN  ENERGIJA, JSI, RATEN, SCK-CEN, SIET, SINTEC, TRACTEBEL and UPM), and is led by ENEA. In addition, an external advisory committee, specialized in different types of plant, as well as safety and licensing aspects, is supporting the project; and an exploitation manager is defined to guarantee the process of exploitation and dissemination of the results reducing the risks associated with a technological innovation.
The ultimate goal of the PIACE project is the finalization of specific designs of the innovative DHR concept, ready for industrial implementation on several reactor technologies ranging from currently operating plants (L-and H-WRs) to innovative energy systems (including FRs and ADSs).
Two main branches can be identified in the project, a design assessment, with the support of numerical codes, to analyse the applicability of the concept to the different reactor technologies of interest (namely, LFR, ADS, PWR, BWR and PHWR), and an experimental investigation to test the feasibility and the performance of the system during accident scenarios. The experimental tests will get advantage of the facility SIRIO located in Italy in the SIET laboratories [3]. The facility is conceived for feasibility testing on this new DHR system and it will be easily adaptable for the tests in PIACE. A simplified scheme of SIRIO facility along with a photo of its steam generator is reported in Figure 1. The system mainly consists of: Steam Generator (SG); Isolation Condenser (IC); Non-condensable Gas Tank (GT); and By-pass heat exchanger (HX) each equipped with isolation valves.
During the steady state condition, the water moves in two-phase flow between the steam generator and the bypass heat exchanger in natural circulation, while the branch of the isolation condenser is isolated and contains gas (nitrogen) at a lower pressure than the water side. When the SG pressure increases up to a pressure set point the opening of the upper IC valve (V1) occur. The steam enters in the IC, "pushes" the non-condensable gas in the gas tank and condenses along the IC tubes. After a certain delay with respect to (V1), the condensate isolation valve (V2) opens and a natural circulation occurs, providing heat transfer from the SG to the water of the IC pool. Then, the electrical heater power is modulated to simulate the decay heat, and when the power removal capacity of the IC exceeds the decay power, the pressure of the system (SG+IC) decreases and the non-condensable gas flows back from the tank into the IC tubes decreasing its condensing capacity. As a consequence, the pressure remains almost constant and the system is able to passively align the power removed from the isolation condenser with that absorbed by the steam generator.

PROJECT STRUCTURE
The project is structured into 6 distinct work packages (WP), as reported in Figure 2. The WP1 is devoted to the implementation of the innovative system to the different reactor technologies and to analyse the plant response for a reference full scale system, through modelling and analyses by system codes. The result consists in a first feasibility study that highlights the potential benefit of the safety system applied to a specific reactor technology. The WP1 is concluded.
In the WP2 the main phenomena controlling the system behaviour are identified. Then, a pre-test phase with numerical models is used to analyse the scalability of the system to be tested on the SIRIO facility, once the facility is adapted to reproduce the reactor technology under investigation. Finally, the test matrices are defined, for each reactor technology, to identify the minimum set of experimental runs necessary to catch the controlling phenomena. The WP2 is concluded.
The WP3 is devoted to the experimental campaigns on the SIRIO facility. The first run planned is relevant for LFR technology. Further tests will be then performed, considering other two reference cases, with an upgrade of SIRIO. The WP3 is ongoing.
The selection of the two reference cases is concluded.. The test campaign for the LFR technology is ongoing and the preparation of the SIRIO facility for the two further reference tests is started. Delays are accumulated on these text campaigns due to the forced long inactivity of the laboratories caused by the pandemic situation of Covid-19. An extension of the project of six months is indeed requested.
The WP4 activities will be devoted to the comparison of the experimental data with the pre-test analyses to solve any difference or scaling distortions, as well as to validate the computational tools and provide "best practices" guidance to capture the main underlying phenomena. Moreover, a technical specification of the safety system will be developed for each of the nuclear technologies under study, covering aspects like, but not limited to, system functions, system criteria, interface requirements, system performances, validation basis. The activities outcome will represent a perspective improvement in terms of TRL and will bring the innovation closer-to-the-market. The preparation of the structure of the documentation is already done.
The WP5 deals with dissemination and training activities. In addition to the classic dissemination activities, workshops are planned during the project to inform young researchers and students of the issues of interest on passive safety systems.
As expected, one WP (WP6) deals with coordination activities related to both project management (quality plan, risk management, interface management, administrative matter, reporting) and technical coordination (project execution plan, technical interfaces, technical review). This can be also applied to MUSA (WP1) and AMHYCO (WP6).

PRELIMINARY RESULTS
The WP1 relates to the selection of enveloping accidental transients and the most representative results obtained through the system codes simulations, for each of the reference technologies, are reported in [4]. Just to illustrate the work done a couple of examples is succinctly described next:  On the liquid metal reactor technology, the activities are based on ALFRED design [5], a pool type structure with all the components installed inside the reactor vessel. The operating conditions of ALFRED are a thermal power of 200 MWth and a thermal cycle between 400 and 480°C. On the secondary side, the water enters the steam generators at 335°C and flows out as superheated steam at a temperature of 435°C and 175 bar. The passive DHR interfaces with the secondary system through the feed water and the steam line and each steam generator has its own safety system loop. The accidental transient studied in the PIACE project was taken from the list of accidental events developed during the LEADER project, [6]. The enveloping event that was identified for the numerical study is the Protected Loss Of Offsite Power (PLOOP) and accounts for the loss of electricity (resulting in loss of all pumps), containment isolation, SCRAM and actuation of the safety system. The transient was analysed by means of the system code RELAP5-3D, [7]. Two main phases of the transient are identified. At first, the safety system operates at maximum power while maintaining the non-condensable gases trapped inside the tank, with a total power around 4 MW. Over time, the temperatures of the reactor coolant system decrease and, once much of the heat capacity has been removed, non-condensable gases migrate from the tank to the isolation condenser and the power balance is achieved. In parallel, the temperatures of the reactor coolant system stabilize to a value greater than 350°C, about 20°C above the freezing temperature. The system maintains this safe condition for the entire 24-hour period studied in this case. The results obtained indicate an increase in the grace time against reactor coolant freezing (which would be otherwise reached in approximately 10 hours) and temperatures are kept within the safety conditions. Figure 3 shows the trend of power absorbed by the coolant and removed by the isolation condenser. Figure 3. IC power removal by IC in ALFRED during a PLOOP As said above, ADS transients and accidents based on MYRRHA design [8] have been also analysed.
 On the water-cooled reactor technology, the considered plant, in the case of PWRs, was a two-loop Pressurized Water Reactor (PWR) with thermal power of 2000 MWth. The enveloping scenario assumed is a station blackout, where the electrically powered components are not available and only passive components and systems remain functional. The proposed conceptual design of the Passive Isolation Condenser (PIC) consists of cylindrical headers and vertical tubes. The PIC was modelled as being part of a separate, closed loop that includes a heater, which simulated the heat flow due to the decay power. The transient involving the functioning of the PIC was simulated with the system code RELAP5/MOD3.3, [9]. A parametric analysis was performed, with different values of the nitrogen tank volume and different initial decay power, highlighting their impact on transient behaviour and most suitable parameter selection for accident mitigation. Figure 4 shows the trend of power removed by the isolation condenser with different nitrogen tank volume. In addition, transient and accident analyses have been also conducted in BWR (particularly, ESBWR) [10] and PHWR (particularly CANDU 600) [11].
The scalability of the systems under investigation has been analysed and the representativeness on the SIRIO facility delineated, in terms of upgrading needs related to instrumentation, layout, volumes, main components and logical controls. In addition, the test matrix of each reactor technology relevant for the experimental characterization on SIRIO facility has been set ( Table 1). The study provided the basis for the selection of the experimental campaigns of the two additional reference cases.

PROJECT OVERVIEW
The MUSA (Management and Uncertainties of Severe Accidents) project aims to quantify the uncertainty in SA codes when modelling accident scenarios for the prediction of the radiological ST [12]. Numerical simulation tools are a central element of the safety demonstration where the compliance of the main safety features of a NPP is checked against the actual safety requirements reflecting the state-of-the-art. In addition, the development and optimization of AM measures aiming at preventing and mitigating the consequences of SA heavily rely on numerical simulations with SA codes such as ASTEC, AC2, MAAP, MELCOR, etc. However, despite statistical tools are available worldwide, few investigations have been focused on SA and UQ. Therefore, MUSA goes beyond the state-of-the-art regarding the predictive capability of SA analysis codes by combining them with the best available UQ tools. By doing so, not only the prediction of the timing for the failure of safety barriers and of the radiological ST in case of a SA in a NPP will be possible, but also the quantification of the uncertainty bands of selected analysis results, considering any relevant source of uncertainty, will be provided. A schematics of the uncertainty propagation is given in Figure 5 [13].
MUSA is not restricted to reactor accidents, which include GEN II and GEN III designs (both PWR and BWR ones). Spent Fuel Pool (SFP) accidents are also addressed. In addition to including AM in the uncertainty analyses, MUSA intends to develop some innovative AM strategies for SFP accidents. It is worth noting that the project has strong links with the communities dealing with Probabilistic Safety Assessment (PSA) level 2, emergency response, environmental consequence analysis, and AM, all of which are undertaking deep reviews since the Fukushima Daiichi accident.

PROJECT STRUCTURE
As shown in Figure 6, the technical WPs (i.e., WP2-WP6) distribute in two blocks (WP1 and WP7 deals with coordination and dissemination, respectively). The first one, including WP2 and WP3, is meant to prepare everything necessary to conduct the second block, which can be referred to as the "application WP block" (i.e., WP4, WP5 and WP6). The "application block" represents roughly two thirds of the total workforce anticipated in MUSA. This would entail both uncertainties in the existing models and uncertainties due to the lack of specific models in the codes. A preliminary "knowledge-based matrix" containing the selected variables, parameters and models and its uncertainty ranges has been set. To do so a list of FOMs was previously developed consistent with the focus of MUSA on ST.
WP3 (Review of Uncertainty Quantification Methodologies, RUQM) aims at reviewing and assessing methodologies and codes used for uncertainty quantification and sensitivity analyses and their applicability for the analysis of SA. In particular, the strengths and weaknesses of each methodology/code to be applied have been identified and by the end of the project a report on guidelines for the best use of UaSA (Uncertainty and Sensitivity Analysis) codes/methods in the SA domain is planned to be delivered. Among the uncertainty quantification tools, partners are using SUSA, DAKOTA, URANIE, RAVEN, and data assimilation tools (NEMM, MOCABA); additionally, the coupling and post-processing are requiring PYTHON scripting.
WP4 (Application of UQ Methods against Integral Experiments, AUQMIE) was outlined to get some experience and insights into the application of the RUQM methodologies against internationally recognised integral ST-experiments. Since its onset, it has become a drill for other application WPs (WP2 and WP3) and an opportunity to provide some early feedback to RUQM. The test PHEBUS FPT-1 was selected for this purpose. The work has already been practically finished and the main outcomes gathered in [14]; Figure 7 shows an example of the results obtained. WP5 (Uncertainty Quantification in Analysis and Management of Reactor Accidents, UQAMRA) aims at demonstrating the applicability and the level of readiness of uncertainty assessment in a broad range of set-ups. In addition to uncertainty bands affecting ST estimates, two other major outputs are to be produced: a best-practice protocol for applying UaSA methods to SA codes (as a feedback to WP3), and an identification of areas where further research would efficiently reduce ST estimates. Four sub-groups have been set up according to the reactor technology (PWR Gen II; PWR Gen II; VVER/CANDU; BWR). The work is undergoing and preliminary insights into BEPU application to SA have been already gained [15].
WP6 (Innovative Management of SFP Accidents, IMSFP) follows the same orientation as WP5 with respect to RUQM methodologies application to SFP SA scenarios, but a special emphasis has been placed on AM mitigation measures. The same scenario (a loss-of-cooling accident inspired in the SFP of Unit 4 of the Fukushima Dai-ichi NPP) is being simulated by all the partners. Given the differences with WP5, FOMs have been somewhat adapted. Preliminary results of the UaSA exercise are being obtained and major observations have been already delivered [13].

PRELIMINARY RESULTS
Once over MUSA half-life time, some preliminary results, particularly from the application packages (WP4-WP6), are worth noting:  Identification and characterization of the input uncertain parameters has resulted/is resulting a key and challenging task. On one side, the huge complexity of a SA scenario involves a tight phenomenological interlink that, even when just focused on a few FOMs, the number of potentially influencing input variables subject to uncertainty is easily over hundreds or even further. On the other, the lack of characterization (lower and upper bounds, probability density function) data for many of those variables make their definition uncertain by itself.  A review is ongoing to identify if input parameters are lacking. But the input uncertainty quantification, which has been identified as an important step in the UQ phase [13], is a long task which is out of the scope this project.  Highly demanding computation costs have been found in all the application WPs, although a noticeable variability among partners have been reported (from a few hours to several days per single run). The use of multi-core processors or PC clusters might help to overcome this challenge (particularly, with UQ exercises over 100 runs).  Code crashes look practically unavoidable when addressing "full-scope" scenarios for the first time, and there is not a systematic way to handle it. Extending the number of runs with respect to the minimum required by the Wilks' formula to ensure having enough computations is an option, but does not guarantee a proper FOM distribution without biases. All this said, some codes appear to be more resilient to failures than others.  Systematization of SA code/UQ-tool coupling as well as post-processing through scripting is an essential step to better accommodate the long time required by this application and to address the best way the analysis of the huge numerical data resulting.  Application of regression techniques entails no minor challenges. At this time, it seems unlikely to can define an optimum correlation coefficient and what is thought to be essential is to be aware of what the chosen coefficient means. This is an area where several project partners are working on. Using physical understanding of the scenario to figure out the rationale behind the significance/non-significance of any sampled variable.

PROJECT OVERVIEW
The project AMHYCO (Towards and Enhanced Accident Management of the Hydrogen/CO Combustion Risk) will consider practical issues to further reduce (as much as possible) the threat posed by combustion of gases generated during accidents on containment integrity [16] . Specifically, it will improve the SAMGs for both in-vessel and ex-vessel phases using numerical and experimental results. It will also experimentally study the phenomena that are difficult to predict numerically (such as H2/CO/H20 distribution and combustion). A third goal will be to improve the predictability of the numerical tools used for explosion hazard evaluation The project has received funding from the Euratom research and training programme 2019-2020 under grant agreement No 945057. The total cost is in the order of 4 M€ with a human resources involvement in the order of 490 person-months, over a 4-year period (2020-2024). It is an international collaboration which involves 10 European partners, both research centres and private companies (CIEMAT, CNL, CNRS, ENERGORISK, FRAMATOME, FZJ, IJS, IRSN, LGI, NRG, RUHR U., UPM), and is led by UPM. In addition, an external advisory committee and an end user group have been set up to count with some external consultancy and to ensure the optimum exploitation and dissemination of the results.
An appropriate management of the associated risk to combustible gases is paramount to avoid the potential release of radioactive material to the environment due to the containment loss of integrity. Severe Accident Management Guidelines (SAMGs) have not been updated according to the knowledge gained in the last decade and the database extension resulting from recent and ongoing projects. The AMHYCO project intends to respond to practical questions, such as the right timing and mode for actuation of containment safety systems (i.e. FCVS, sprays, fan coolers) to reduce as much as feasible the threat posed to containment integrity. To do so, all the available tools to enhance the present status (i.e. LP, 3D and CFD codes, together with experimentation and the best use of engineering judgement) are to be applied.
The scope of this project is outlined by the most common reactor technology in the EU: PWRs. The three existing designs are being addressed: PWR-Western type (PWR-W), PWR-KWU type (PWR-K) and PWR-VVER type (PWR-V). When feasible, diversity in reactor size (power) is being also considered.

PROJECT STRUCTURE
In order to meet the above objectives, the project proposes a working method based on five different work packages (WPs), as illustrated in Figure 8, which are at the same time successive and interlinked. The WP1 aims at reviewing the status of existing methodologies and practices related to gas combustion risk management in the late phase of severe accidents. As most of the mitigation strategies adopted in European countries to prevent this risk are based on use of PARs, a critical assessment of the PAR behaviour in the late phases of severe accidents is performed. Similarly, a survey of the available experimental data and engineering combustion models is provided. The lessons learned will help identifying knowledge gaps related to hydrogen (H2) and carbon monoxide (CO) recombination and combustion. In addition, a review of the safety equipment qualification criteria and gas monitoring systems, as well as the existing SAMGs is established with the aim to pave the way towards the SAMGs enhancement.
The intention of the WP2 is to identify accident sequences in which H2 and CO combustion risk might jeopardize containment integrity. Given the current European nuclear reactor fleet, the three PWR reactor designs have been modelled (PWR-W; PWR-K; PWR-V) and the most relevant sequences have been chosen according to a set of agreed criteria: high molar fractions of combustible gases (H2 + CO) in control volumes which conditions are within flammability limits; large total mass of combustible gas (H2 + CO) within the containment; and, fast combustible gas (H2 + CO) release rates. Besides, this WP includes a specific task to produce "Generic Containments (GC)" databases based on containment descriptions available and/or reachable by the AMHYCO team [17]; from the 3D description (CAD file), lumped-type nodalizations are agreed among partners to be later used in WP4.
The objective of WP3 is to experimentally investigate phenomena related to hydrogen and carbon monoxide combustion and mitigation, which are still not covered by existing numerical models. The focus is set on combustion of hydrogen/carbon monoxide/steam mixtures as well as on the operational behaviour of PARs under boundary conditions representative of the late phase of severe accidents. Hence, following the conclusions of the review performed in WP1 and the scenarios selected in WP2, WP3 will provide specific experimental data to assess and enhance the numerical tools used in WP4.
The most penalizing sequences identified in WP2 for the PWR containment designs will be further analysed within WP4. The focus of the analysis will be to investigate the late phase containment response in terms of containment pressure build-up, H2/CO combustion risk, efficiency and options/timing of individual mitigation measures (in particular PARs, fan cooler or spray systems and FCVS), and equipment and instrumentation surveillance. For this purpose, the sequences for each plant design will be generalized and harmonized, as far as possible in WP2 to enable comparability and the RCS is simply represented by source terms (energy and mass; fission products will be neglected).
Finally, WP5 aims to identify room for improvement of SAMGs with respect to the management of combustible gases within the containment. The basis for this evaluation are the experimental and computational efforts carried out in the abovementioned WPs. Thereby, the insights from the calculations including the late phase with CO present in the containment will be transferred into SAMG considerations.
Recommendations are foreseen for all large-dry PWRs containments explored in the project.

PRELIMINARY RESULTS
For the time being, the results have come from the insights gained from WP1 reviews, the calculations conducted in the frame of WP2 and the GC databases built-up and the early phase of experimentation concerning PARs performance and combustion of combustible gas mixtures. All of them may be summarized as follows:  Additional experimental and theoretical investigations are needed to fill the knowledge gaps regarding the PARs behaviour in representative conditions of severe accidents late phases. Given that the different nature of existing models (from engineering correlations to CFD models accounting for phenomena as thermal radiation, superficial and bulk chemical reactions on the surface and in the gas) were validated on experiments dealing with H2, data with gas combustible mixtures are necessary to extend the domain of the models validation. WP3 will supply the data for such purpose and the resulting enhanced equations will be implemented in safety codes to simulate the containment severe accident scenarios in WP4.  The requirements for combustion risk management and mitigation measures address only in-vessel conditions and aiming to preserve containment integrity; their extension to ex-vessel conditions needs to be established. In addition, most times they have a qualitative nature. This focus on the in-vessel phase extends to monitoring systems, which are not designed to measure CO. Finally, just few SAMGs instruct on how to use other safety systems (i.e., sprays, FCVS) to handle the combustion risk in the late phases.  H2/CO combustion in representative conditions has been poorly investigated.
Few data are available in the open literature and, consequently, the application of existing modelling is restricted to a limited validation. Part of the work underway in WP3 is dealing with this issue and will provide data on laminar and turbulent flame speed, and flammability limits and flame acceleration criteria. These data are expected to contribute to enhancement of the existing engineering correlations and combustion models.  A number of accident sequences have been modelled for each reactor type addressed ( Figure 9). The accident selection has been fundamentally based on the resulting Shapiro diagrams, samples of which for PWR-W are shown in Figure 10. In the case of PWR-KWU reactors, the sequence selected has been an SBLOCA (80 cm2). Despite the diversity of reactor size in PWR-W (i.e., 700 MW, 1000 MW and 1300 MW), it was agreed that the selected sequences were a SBLOCA (120 cm2) and a double-ended guillotine break (LBLOCA). As for VVER reactors, the decision is still to be made. and in starvation conditions (O2/N2=0.11) for an initial pressure ranging from 1 to 3.5 bar were determined in the ENACEFF facility. The effect of CO2 and H2Ov is being currently investigated [18,19] .  Scoping tests were performed with different catalysts (platinum-and palladiumbased) to define the boundary conditions of more complex experiments to be conducted in the REKO-3 facility [16]. Specific focus was set on identifying the conditions for catalyst poisoning which could be of high significance for latephase PAR operation. For Pt-based catalysts, the catalyst poisoning temperature (i.e. the threshold value of the catalyst temperature below which the catalyst gets poisoned) was shown to be independent of the gas temperature, while Pd-based catalysts were found to be less sensitive to poisoning at elevated gas temperature.

FINAL REMARKS
This paper illustrates the ongoing support from Euratom to further strengthen nuclear safety of NPPs through investigating innovative reliable safety systems (PIACE project), enhancing predictive analysis of SA scenarios (MUSA project), and optimizing guidelines to manage combustible gas risk in containment (AMHYCO project).
The PIACE project is focusing on the designs of the innovative DHR concept, ready for industrial implementation on several reactor technologies ranging from currently operating plants (L-and H-WRs) to innovative energy systems (including FRs and ADSs). The innovation lies in the ability of the system to control passively the power exchanged with the heat sink by using non-condensible gases. The activity developed so far showed: the applicability of the system to the different reactor technologies; the scalability of the system to be tested on the SIRIO facility; the necessary testing of each reactor technology; and the selection of the reactor technologies as reference cases for the test campaign. The experimental campaign is ongoing.
The MUSA project is facing the challenge of bringing uncertainty and sensitivity analysis into the SA simulations as a means to better assess the potential source term to the environment, encompassing both Gen. II and Gen. III reactors and SFP accidents. So far, a database on input deck uncertainties has been built and applied to the simplified scenario of PHEBUS-FPT1 and the first challenges and difficulties of a systematic application of UaSA into SA analyses has been experienced and ways to overcome them are being proposed. Currently, the practical application to reactor and SFP scenarios is underway and the lessons learned from the FPT1 training experience are being tested in these "full-scope" simulations. From these exercises feedback on BEPU methodologies application in SA analyses are being settled.
The AMHYCO project is addressing practical issues to further reduce (as much as possible) the threat posed by combustion of gases generated during accidents on containment integrity. Specifically, it will propose improvements the SAMGs for both in-vessel and ex-vessel phases using numerical and experimental results. It will also experimentally study the phenomena that are difficult to predict numerically (such as H2/CO/H20 combustion and recombination). The progress done so far has paved the road to the full containment simulations and SAMGs feedback that will take place in the second phase of the project; in addition, experimental investigations on PARs performance and gas combustion under conditions barely explored earlier are ongoing.
Finally, despite the harsh COVID conditions that have prevailed since early 2020, the three projects are deeply commited to young researchers' and engineers' education and training through powerful mobility programs that are already being executed.