Conference Agenda

Slovenia has been a party to the Treaty on the Non-Proliferation of Nuclear Weapons (NPT) since 1992, and in 2000 the Additional Protocol to the Safeguards Agreement with the International Atomic Energy Agency (IAEA) entered into force.

The Additional Protocol aims to enable the IAEA to provide assurance about both declared and possible undeclared activities and to get a more complete picture of a State’s overall nuclear programme. Under the Additional Protocol, the IAEA is granted expanded rights of access to information and sites.

As set out e.g. in the Resolution on Nuclear and Radiation Safety in the Republic of Slovenia (for the period 2024 - 2033), Slovenia remains firmly committed to the NPT and to the Safeguards Agreement, including the Additional Protocol, and ensures that nuclear safeguards inspections are conducted smoothly and promptly, when required. Under the Euratom Treaty, the European Commission (Euratom) has special competence for the supervision of nuclear material, which is aligned with the IAEA's competence.

A few reports are presented more in details, namely under Article 2.a.(i) – a general description of and information specifying the location of nuclear fuel cycle-related research and development activities not involving nuclear material; Article 2.a.(iii) – site declarations (general descriptions of each building on each site); 2.a.(ix) – export-related information regarding specified equipment and non-nuclear material (listed in Annex II); and Article 2.a.(x) – general plans for the succeeding ten-year period relevant to the development of the nuclear fuel cycle.

A substantial decrease in the number of the IAEA’s inspection, using the possibility of “complementary access” in Slovenia (after 2020, and pandemics with covid-19) may have different reasons, some of them being the optimisation of the IAEA’s resources, using novel technologies like satellite images, remote data transmission, mature reporting of changes to the sites and R&D activities from the national perspective, transparency etc.

The article gives an overview of domestic experiences, piled through a quarter of century of reporting and interaction with the IAEA (some fewer years with Euratom) as well as highlighting some newest arrangements – in particular the SLA (State-level approach), concluded by the IAEA in 2025.

The global energy transition has renewed interest in nuclear power, particularly through the development of Small Modular Reactors (SMRs) and advanced reactor concepts. These next-generation systems aim to achieve enhanced sustainability through the reprocessing of spent fuel and the closing of the nuclear fuel cycle While these strategies are intended to significantly reduce the environmental footprint of nuclear energy, their effectiveness and economic viability at scale remain to be demonstrated through sustained commercial operation. In light of these developments, this report provides a comprehensive overview of advanced technologies and fuel cycle strategies, establishing a comparative framework with mature thermal reactor systems.

The analysis focuses on Boiling Water Reactors (BWRs) as established representatives of thermal reactor technology, alongside Sodium-cooled Fast Reactors (SFRs) and Lead-cooled Fast Reactors (LFRs), which represent advanced next-generation systems. The study details their historical evolution, fundamental design, component layout, and operational characteristics, including safety and control mechanisms.

The comprehensive comparative analysis is conducted at two distinct levels. Initially, the study compares thermal and fast reactors, utilizing the Pressurized Water Reactor (PWR) as the primary benchmark due to its status as the most mature and widely deployed technology worldwide. While thermal reactors, such as the PWR and BWR, offer technological maturity and operational stability, fast reactors enable superior fuel utilization and a significant reduction in long-lived radioactive waste. This is achieved through their fast neutron spectrum, which—facilitated by the absence of a neutron moderator—enables the breeding of new fissile material and the transmutation of long-lived actinides into shorter-lived or stable isotopes. Building on this foundation, the report provides a targeted technical comparison between specific designs, evaluating the trade-offs between the BWR’s simplified single-loop direct cycle and the PWR’s high-pressure dual-loop stability. Furthermore, it contrasts the excellent heat transfer and technological maturity of sodium-cooled fast reactors (SFRs) against the superior chemical inertness and passive safety potential of lead-cooled designs (LFRs), while simultaneously addressing the critical challenges of material corrosion and coolant density inherent to these systems.

The final section emphasizes the role of spent nuclear fuel reprocessing and the closing of the nuclear fuel cycle. The findings conclude that while thermal reactors will remain the backbone of global nuclear capacity due to their proven reliability, SFR and LFR technologies—integrated with advanced fuel cycles— could represent a promising pathway for enhancing the long-term sustainability of nuclear energy.

This paper presents the design and experimental verification of a special fuel assembly for the LR-0 research reactor, combining an outer hexagonal shape corresponding to the VVER-1000 geometry with an inner square lattice of fuel rods. This concept enabled the creation of a unique mixed core region with two fuel geometry types while preserving compatibility with the existing reactor configuration. The research therefore provided an experimental platform for investigating the neutronic behaviour of a non-standard fuel arrangement.

A functional prototype of the assembly was manufactured and experimentally tested in the reference LR-0 core. The measurements focused on determining key neutronic parameters and on power distribution mapping. The obtained results were compared with calculations performed using suitable neutronic codes. The main outcome of the work is a reference benchmark for code validation together with expanded domestic know-how in the design and assessment of dual-geometry fuel assemblies.

Small Modular Reactors (SMRs) are emerging as flexible low-carbon power sources capable of complementing intermittent renewable energy systems, making Daily Load-Following Operation (DLFO) an increasingly important operational requirement. Previous studies have primarily focused on load-following control strategies for large Pressurized Water Reactors (PWRs). However, the Innovative Small Modular Reactor (i-SMR) adopts soluble boron-free(SBF) operation and possesses unique core characteristics, including a short active core height and relatively high control rod worth. Therefore, optimization of control rod operation for i-SMRs is required.

In SBF cores, reactivity control is achieved using fixed burnable absorbers and control rods without soluble boron, resulting in increased dependence on control rod operation during power maneuvering.

Since control rod movement directly affects neutron flux and power distributions, the i-SMR employs a control rod operating strategy in which control rod banks are partially inserted and sequentially inserted or withdrawn according to reactor power level.

Therefore, the position and overlap of each control rod bank must be carefully designed to minimize power distribution perturbations and xenon transient effects while maintaining thermal design margins. Control rod overlap refers to an operating condition in which adjacent control rod banks are simultaneously partially inserted within a certain axial region of the core. Because control rod overlap directly affects neutron flux distribution, it strongly influences the behavior of the peaking factors.

The reference i-SMR analyzed in this study has a thermal power of 520 MWth and consists of 69 fuel assemblies arranged in a 17 × 17 lattice configuration with an active core height of 240 cm. The maximum U-235 enrichment is 4.95 w/o. To suppress excess reactivity and local power peaking throughout the fuel cycle, the fuel assemblies employ Gd2O3-UO2 burnable absorber rods containing enriched gadolinium. The regulating control rod system consists of 28 Ag-In-Cd absorber rods divided into four control rod banks—R4, R3, R2, and R1—according to insertion sequence. During base-load and load-following operation, banks R4, R3, and R2 are primarily utilized for reactivity control and axial power distribution control.

Three-dimensional core depletion and transient simulations were performed using the ASTRA code developed by KEPCO Nuclear Fuel. ASTRA is a three-dimensional nuclear design and core depletion code for PWR analysis based on advanced reactor physics methodologies.

The reference DLFO scenario consists of the following operational sequence: (1) steady-state operation at 100% rated power, (2) power reduction from 100% to 20% according to a prescribed ramp rate, (3) low-power hold operation at 20%, (4) power increase from 20% to 100%, and (5) return to steady-state full-power operation. The ramp rate is selected within design and regulatory limits, and operating conditions with significant xenon transient effects are considered to provide conservative operating conditions. The analysis is conducted under representative beginning of Cycle (BOC), Middle of Cycle (MOC), and End of Cycle (EOC) conditions to comprehensively evaluate peaking factor behavior throughout the entire operating cycle.

To optimize control rod overlap, representative overlap conditions were defined by varying the overlap between adjacent control rod banks from 40% to 60%. For each overlap condition, three-dimensional core transient calculations were performed to comparatively evaluate the transient behavior of Fr and Fq as well as xenon transient effect during load-following operation. Particular emphasis was placed on the instantaneous behavior and maximum deviation of the peaking factors during power maneuvering. In addition, RMS fluctuations were evaluated to quantitatively assess transient stability throughout the entire operational sequence.

Furthermore, the effects of power distribution distortion and axial power distortion caused by sequential bank operation were investigated to evaluate the core response characteristics associated with different overlap conditions. Insufficient control rod overlap may cause abrupt peaking factor variations, whereas excessive overlap may intensify axial power asymmetry due to simultaneous insertion of multiple control rod banks.

Based on these analyses, this study aims to identify the optimal control rod overlap condition that minimizes transient variations in Fr and Fq while maintaining thermal design margins and stable load-following performance. The results of this study are expected to contribute to the optimization of control rod operating and core design strategies for SBF-based i-SMR load-following operation.

A modular multiphysics framework was developed for coupled neutronic and thermal-hydraulic analysis of pressurized water reactor cores. The framework combines continuous-energy Monte Carlo neutron transport calculations performed with Serpent 2 and a simplified assembly-wise thermal-hydraulic solver derived from the CTEMP module of the CORD-2 system, enabling the treatment of reactivity feedback due to changes in fuel temperature and coolant density. In the present work, the coupling methodology is validated through analysis of the first Krško NPP operating cycle. The calculated results are compared with plant measurements and with results obtained using the CORD-2 code system, employing the ENDF/B-VII.1 nuclear data library.

The design of demonstration fusion reactors must incorporate robust safety systems capable of mitigating potential accident scenarios. One of the postulated accidents in such reactors is an ingress-of-coolant event (ICE), in which high-pressure, high-temperature water is released into the vacuum vessel due to a rupture in the cooling tubes of plasma-facing components. To investigate this scenario, an integrated ICE experimental facility (JAERI, Japan) was constructed.

The purpose of this study is to simulate the ICE event using the ATHLET code, which is part of the AC² software package developed by GRS, Germany. The simulation results will be compared against experimental data.

Accurate evaluation of effective delayed neutron fraction, and the effective neutron generation time is essential for reactor kinetics, safety assessment, and transient analysis. This work utilizes the recently implemented iterated fission probability (IFP) methodology in OpenMC to calculate these kinetics parameters for the TEPLATOR reactor model. TEPLATOR is a heavy-water moderated and cooled pressure-channel small modular heat-only reactor. The objective of this study is to calculate adjoint-weighted kinetics parameters for selected reactor configurations and to compare the results with reference Serpent calculations.

The TEPLATOR model considered in this study includes a detailed representation of the heterogeneous core. It accounts for pressure channels, VVER-440 fuel assemblies, moderator and coolant regions, and the full control rod system. Kinetic parameters are evaluated for three reactor operating states: (1) all control rods withdrawn, (2) all control rods inserted, (3) a critical configuration defined by the specific positions of the regulating, compensation, and emergency rods. Similar Serpent calculations are performed for the same reactor conditions to perform a consistent code-to-code comparison.

A good level of agreement between OpenMC and Serpent is obtained for all TEPLATOR configurations in this work. This will help assess the applicability of OpenMC IFP capability to a detailed heavy-water reactor model. The results will provide a useful basis for future reactor kinetics studies, safety evaluations, and transient analyses of the TEPLATOR concept.

The Information Centre is part of the Nuclear Training Centre at the Jožef Stefan Institute and informs the visitors about nuclear power and nuclear technology, about radioactivity, about Krško Nuclear Power Plant and about energy in general.

Our main target population are the schoolchildren from the last grades of elementary school and from high school (ages 13-18). We attract some 7000 visitors per year and offer them live lectures on nuclear technologies (fission and fusion), a lecture about use of radiation in medicine, industry and science and a lecture on stable isotopes. A general lecture about energy and an energy workshop is often performed for younger visitors. The visit includes a demonstration of radioactivity, a tour of our permanent exhibition and an optional tour of the TRIGA research reactor.

Since 1993 we monitor the opinion trends by polling some 1000 youngsters by 10 questions that remain unchanged for several years. This enables us to follow the trends in the basic knowledge of energy issues among youngsters and their attitude towards nuclear energy.

Sustainability of nuclear energy is strongly based on the deployment of innovative fast reactors. In this frame, most promising systems employ sodium or lead as coolant. Prototypes and demonstration plants of these reactors are foreseen to be loaded with (U,Pu)O₂ mixed oxide fuel with concentrations of plutonium up to 30 wt.%. This design choice should accelerate licensing procedures thanks to the availability of operating experience gained in reactors such as Phenix. Thermal conductivity is one of the key properties of MOX fuel. Measurements of MOX thermal conductivity have been published since the 1960s. However, research efforts are still ongoing to investigate the effect of high concentrations of plutonium and addition of minor actinides. Moreover, an extension of deviation from stoichiometry domain and the effect of burn-up have been topics of recent measurements and investigations. Several correlations of MOX thermal conductivity are available in the literature. Each correlation tends to model the effect of main input parameters (temperature, plutonium content, deviation from stoichiometry, and burn-up) based on the experimental dataset used for its development. This leads to differences in modelling. This paper presents preliminary results obtained using a machine learning approach whose indications are compared with the outcomes of correlations published recently or widely applied in the analysis of MOX fuel performance.

Delivering smarter, faster, and more flexible nuclear solutions starts now !
Additive manufacturing technology is already supporting large size reactors operating fleet, new build and AR/SMRs programs.

Which assets and benefit does it bring ?

- Qualified process compliant with nuclear codes & standards (ASME, RCCM), for high quality, unit or in series production with cost & lead time competitiveness

- Improved performance through complex geometries not possible with conventional methods

- Strengthened supply chain resilience for safety-critical components

At Framatome, our Advanced and Additive Manufacturing center has already filled the gap between concept and industrial applications and deliveries.

Rapid, reliable, and compliant with the highest standards of the nuclear industry as constant priorities, Framatome has unique capacity to produce nuclear-grade parts from a few millimeters to several meters, and from a few kilograms up to 25 tons.

By integrating R&D, engineering, qualification, and manufacturing as one stop shop, we help customers and partners to accelerate innovation and bring safer, more efficient solutions to new build projects and long term operations.

In today's world, small modular reactors, especially microreactors as a subcategory, are favoured as emerging technological advancements. The aim is to design a multifunctional microreactor that can be used for electricity production and load-following, as well as for radioisotope production, irradiation, and research purposes when electricity is not needed.

As part of the analysis of the properties of a multifunctional UZrH fueled microreactor, we studied the impact of different methods of short-term reactivity control. We investigated the efficiency and feasibility of control rods, drums, and plates, within the limited space available. This constraint results from the microreactor’s small size, as it is designed to be transported as a whole by truck. Therefore, it is important to compare the advantages and disadvantages of these control methods. The aim of the analysis is to determine the most favourable solution for the designed core, which will provide sufficient negative reactivity input for safe reactor shutdown.

Calculations were performed for three different core configurations in which all fuel elements were identical, using the same type of fuel (fresh fuel with homogeneous enrichment) and without any burnable absorbers. The models were prepared using the open-source Monte Carlo transport code OpenMC, and the calculations were performed using the ENDF/B-VIII.1 cross-section library. We evaluated the performance of different absorbers in the control rods and materials in the drums. Then, we analyzed the impact of individual control methods on the distribution of neutron flux and fission rate in the core, which provided better insight into their behaviour and operation. Finally, we determined which method was most suitable for further calculations and analyses of the core under consideration.

Magnox swarf, the magnesium alloy cladding residue from spent nuclear fuel decanning, constitutes approximately 10,000 tonnes of material stored at Sellafield, UK in the Magnox Swarf Storage Silo). Primarily composed of magnesium with minor aluminium and trace elements, the swarf is heavily contaminated with fission products (e.g., caesium-137, strontium-90), activation products, and adherent actinides/uranium residues. Traditional management involves direct encapsulation, contributing to long-term disposal burdens. This proof-of-concept study explores a hydrometallurgical route to recover high-purity magnesium while concentrating radionuclides into a minimized waste stream.

The process begins with dissolution of Magnox swarf in hydrochloric acid to convert metals to soluble chlorides. Subsequent stepwise pH adjustment using a base (e.g., NaOH or ammonia) exploits solubility differences: magnesium hydroxide (Mg(OH)₂) precipitates selectively in the alkaline range (typically pH 9–11), while many fission product hydroxides exhibit different precipitation behaviours. Vacuum filtration isolates the Mg(OH)₂ solid, which is then redissolved in fresh hydrochloric acid. The cycle of dissolution, controlled precipitation, and filtration is repeated iteratively. Each radioactive supernatant or filtrate is recycled back into the preceding dissolution step to progressively concentrate contaminants and minimize secondary waste volumes. Purification continues until the recovered magnesium hydroxide shows no measurable radioactivity above background using gamma spectroscopy and gross alpha/beta counting.

Laboratory-scale experiments utilise non-radioactive Magnox alloy simulants for process optimisation. Key parameters investigated include acid concentration, pH ramp rate, temperature, agitation, and the effect of co-dissolved species (Al, Fe, trace actinides) on precipitation efficiency and purity. Characterisation employs ICP-MS/OES for elemental analysis, XRD for phase identification, SEM-EDS for morphology, and radiometric techniques to track decontamination factors at each stage.

Preliminary thermodynamic modelling and small-scale tests indicate high magnesium recovery (>90% per cycle) with effective partitioning of key gamma-emitters into the aqueous waste stream. Multiple cycles are expected to achieve decontamination factors sufficient for the magnesium product to be considered non-radioactive or suitable for clearance/recycling as industrial-grade Mg(OH)₂ or converted metal. Challenges such as hydrogen generation during initial dissolution, amphoteric behaviour of aluminium, and handling of actinide traces are addressed through process controls and reagent selection.

This approach offers a potential pathway to volume reduction and resource recovery from legacy nuclear waste, aligning with circular economy principles in decommissioning. Successful demonstration could inform larger-scale treatment of Magnox wastes, reducing disposal liabilities and recovering magnesium for non-nuclear applications. Results from the proof-of-concept, including mass balances, decontamination efficiencies, and economic considerations, will be presented.

Key words: SMR, cogeneration, district heating, process heat, hydrogen, industrial decarbonisation

Abstract: Small Modular Reactors (SMRs) are increasingly recognised not only as low-carbon electricity generation technologies but as flexible, multi-functional energy sources capable of delivering heat, hydrogen, and a range of industrial energy services. Their capability to operate in cogeneration mode and to supply thermal energy across a wide temperature spectrum enables applications that extend well beyond conventional electricity production.

This paper presents an overview of non-electric applications of SMR technologies, with a particular focus on those most relevant to the Slovenian energy system. The analysis demonstrates that SMRs can serve as a technically robust platform for decarbonising energy-intensive sectors and integrating low-carbon heat into existing infrastructure.

A key challenge in modern decarbonisation pathways is that, although electricity generation can increasingly rely on renewable and nuclear sources, a significant portion of final energy demand remains tied to heat, steam, and chemical energy carriers. In Slovenia, such demand is largely concentrated in district heating systems and industrial processes, which are still heavily dependent on fossil fuels. SMRs offer a viable solution by enabling simultaneous production of electricity and usable thermal energy from a single low-carbon source.

The fundamental enabler of these applications is nuclear cogeneration. By concurrently generating electricity and heat, SMRs achieve higher overall energy utilisation compared to electricity-only systems. Controlled extraction of thermal power allows for substitution of fossil-based boilers and furnaces while maintaining stable electricity output, contributing to both emissions reduction and improved system efficiency.

District heating represents one of the most mature and near-term opportunities. SMRs are capable of supplying hot water or low-pressure steam in the range of approximately 80–150 °C, which aligns well with the operational requirements of district heating networks. In Slovenia, where heat demand is concentrated in urban areas with existing or planned district heating infrastructure, this application offers strong potential for early deployment.

Industrial process heat is another high-impact application area. A substantial share of industrial energy consumption is associated with steam generation and process heat for operations such as chemical reactions, drying, distillation, and material processing. SMRs can provide thermal energy ranging from low-temperature heat to high-temperature steam exceeding 500 °C in advanced designs. Industries with continuous heat demand—such as chemical processing, pulp and paper, food and beverage production, pharmaceuticals, and certain metallurgical processes—are identified as key candidates in the Slovenian context.

Hydrogen production is also of strategic importance. SMRs can support hydrogen generation via low-temperature electrolysis using nuclear electricity, as well as through high-temperature electrolysis and thermochemical processes utilising nuclear heat, offering improved efficiencies. In Slovenia, hydrogen is primarily considered for domestic industrial use rather than export-oriented production. Additional applications, such as desalination, are technically feasible but currently of limited relevance in the national context.

Building on the identified potential of these applications, GEN energija is currently preparing a comprehensive feasibility study for the deployment of SMRs in Slovenia. The study will cover a broad range of technical, economic, and strategic aspects of SMR implementation, and will also include a dedicated assessment of non-electric applications, such as heat production, hydrogen generation, district heating, and selected industrial uses. It aims to evaluate how these applications contribute to the overall feasibility and economic viability of SMR projects, providing a clear and data-driven basis for strategic decision-making and long-term energy planning.

Overall, the findings confirm that the primary value of SMRs in Slovenia lies in their ability to provide reliable, low-carbon heat and cogeneration solutions for urban and industrial clusters. This significantly enhances their role beyond electricity generation and positions SMRs as a key enabler of deep decarbonisation within the national energy system.

Accurate prediction of critical heat flux (CHF) is essential for defining safe operating margins in nuclear reactor thermal-hydraulics. However, conventional CHF correlations often show limited generalization because of the nonlinear dependence of CHF on pressure, mass flux, flow quality, inlet subcooling, and channel geometry. This paper aims to develop a physics-informed machine learning framework for CHF prediction using a consolidated experimental database of more than 24,000 boiling heat-transfer cases.

The main objectives are to construct physics-guided input features, evaluate advanced regression models, and design leakage-aware validation strategies that preserve independence among experimental campaigns. The study also introduces a hybrid residual-learning approach, where machine learning is used to model deviations from a physics-based CHF baseline rather than replacing thermal-hydraulic knowledge entirely. Model interpretability is considered through SHAP analysis to identify the relative importance of governing parameters.

The proposed framework is intended to improve generalization, physical consistency, and transparency in CHF prediction, supporting future applications in reactor safety analysis, advanced reactor design, and digital-twin development.

In hypothetical nuclear severe accidents, the coolability of a debris bed formed after core meltdown is essential. Insufficient cooling can compromise containment integrity. Failure to ensure effective cooling can so potentially lead to the release of radioactive materials, posing a risk to the environment.

To analyse this process, numerous experiments were conducted, e.g. PEARL and PRELUDE at ASNR, France, DEBRIS and FLOAT at IKE, University of Stuttgart, Germany, and POMECO at KTH, Sweden, focusing on the investigation of debris bed quenching under different flooding conditions.

The water flow analysis indicated that quench front evolution and cooling dynamics under top-flooding conditions occur in two distinct stages. In the first phase, water flows downward through the debris bed, forming preferential channels either near the wall or in the central region. The second phase begins when the quench front reaches the bottom and a stable water pool forms, supplied by the preferential channels. The water level then rises until the entire debris bed is quenched.

An analytical model will be developed to qualitatively describe the debris bed quenching process by considering the two phases separately. In the first step, the model will predict the quenching velocity and the duration of the initial phase, during which water penetrates downward through the debris bed. From the estimated fraction of the debris bed quenched in this phase, the quenching velocity and duration of the second phase will be determined, which will allow evaluation of the total quenching time. The model predictions will be compared with the experimental results.

Thermal fatigue is an important degradation mechanism in nuclear power plant components, where cyclic temperature variations induce stresses that can initiate and propagate cracks. In pressurized water reactor (PWR) mixing tees, accurate prediction of crack growth remains challenging due to the combined influence of thermal loading, stress conditions, and environmental effects. Although several estimation models have been developed, detailed comparisons between them are relatively limited. This study presents a numerical framework for analysing through-wall crack growth of a semi-elliptical surface crack under sinusoidal thermal loading. The approach combines transient heat conduction analysis, thermo-mechanical stress evaluation, and Paris-law-based crack growth modelling, while comparing the ASME and JSME assessment procedures. The results indicate that thermal load amplitude and loading frequency are the dominant parameters affecting fatigue behaviour. Elevated amplitudes significantly accelerate crack growth, leading to reductions in component lifetime by several orders of magnitude. The most critical conditions arise at intermediate loading frequencies, where the combined effects of steep thermal gradients and cumulative cyclic damage are most pronounced. Variations in crack geometry are also studied, although their effect is shown to be considerably smaller in comparison to operating conditions. Both assessment procedures predict consistent overall trends, although the ASME model yields systematically more conservative lifetime estimates, while the JSME model demonstrates greater sensitivity to mean stress and environmental effects. Overall, the developed framework successfully captures the key mechanisms governing thermal fatigue crack growth and provides a robust and computationally efficient tool for estimating the service lifetime of components subjected to thermal cyclic loading.

GEN energija is leading the project for the construction of a second unit at the Krško Nuclear Power Plant, referred to as JEK2. The project is currently in the early phase of spatial planning, while several supporting technical studies and analyses are being prepared. In 2025, GEN energija carried out technical feasibility studies for the potential deployment of AP1000 and EPR designs at the proposed JEK2 location. As part of technical dialogues with potential vendors, information was also obtained regarding the transport of largest and heaviest components.

The transport of large and heavy components represents one of the key logistical challenges in the development of the new nuclear power plant. GEN energija has therefore carried out several studies related to the transport of components for JEK2. An initial feasibility study identified and assessed possible transport corridors, while in 2025 and 2026 a more detailed Preliminary Transport Study (PTS) for JEK2 was performed. The PTS addressed maritime and inland waterway ports, road and rail transport options, infrastructure constraints, legal and permitting requirements, technical feasibility, risks and the broader social acceptability of infrastructure adaptations, with particular focus on the identification and evaluation of the optimal transport route to the planned JEK2 site.

Eight potential road transport routes to Krško were analysed. These included routes from the seaports of Koper, Trieste (IT) and Rijeka (CRO), and from the inland ports of Budapest (HU), Bratislava (SK), Linz (AT) and Vukovar (CRO). The analysis considered route length, number of overpasses, tunnels, bridges and other structures, available clearance profile, load-bearing limitations, permitting complexity and the feasibility of infrastructure modifications. Railway transport was also considered but was found to be unsuitable for all critical components due to clearance profile limitations (height and width limitations), although it may remain applicable for certain heavy components such as transformers.

The comparison showed that routes involving international transit or highway corridors include significant technical and administrative limitations. These include numerous tunnels and overpasses, large numbers of bridges requiring structural assessment, limited clearance profiles, uncertain permitting in several jurisdictions and limited potential for permanent infrastructure adaptation. Inland waterway routes also introduce additional risks related to transshipment and water-level variability.

Based on the comparative assessment, the optimal transport route was identified as the route from the Port of Koper to Krško via Ljubljana, Celje and Bizeljsko, mainly using main, regional and local roads while largely entirely avoiding highways and railways. The route is located only in Slovenia. It is approximately 274 km long and contains no tunnels. It has the lowest number of fixed vertical obstacles among the assessed routes and offers the greatest potential for increasing the transport clearance profile through targeted infrastructure adaptations. The existing route allows exceptional transports with a total height of up to approximately 4.70 m, while further adaptations and bypass arrangements could increase the available height of profile to approximately 7.10 m, and in more extensive infrastructure scenarios even higher. The preliminary transport study identified 112 critical points or sections along the optimal route. The critical points were divide into several categories, which include vertical clearance restrictions, width restrictions, load-bearing restraints, special infrastructure constraints and bridges requiring load-bearing assessment.

The selected route also offers advantages from a legal and administrative perspective, as it lies entirely within Slovenia. This provides a unified and predictable regulatory framework, enables direct coordination with national authorities and allows infrastructure modifications to be justified as also being in the public interest. Such modifications could also provide long-term benefits for the operation and maintenance of the future nuclear facility.

Further work will require detailed geodetic surveys, structural assessments of bridges, optimisation of transport configurations, 2D and 3D transport simulations, definition of required infrastructure measures, permitting and implementation planning.

The study concludes that the Koper–Ljubljana–Celje–Bizeljsko–Krško route is the most suitable basis for further development of the transport route for JEK2 critical components. Early planning, coordination with suppliers and infrastructure owners, and timely preparation of technical and permitting documentation will be essential to reduce project risks and ensure the safe and feasible delivery of critical components to the JEK2 site.

Slovenia is preparing for the JEK2 unit alongside the operating Krško NPP. In this regulated environment, every engineering, licensing, and security decision must trace back to specific clauses of binding standards and internal documents such as design basis records, technical specifications, and operating procedures. Engineers currently experiment with general-purpose LLM-based tools, but these hallucinate, blur successive revisions of the same standard, and route confidential queries through cloud APIs incompatible with air-gapped environments.

Our long-term objective is a nuclear-tailored, on-premise LLM assistant operating on internal documents that cannot be shared externally. We base it on retrieval-augmented generation (RAG): the LLM is constrained to ground every response in citations from a controlled document set. RAG is not a replacement for the LLM, but the adaptation that makes a generic model trustworthy enough for safety-critical use. Because operational documents cannot be made public, this paper develops and validates the method on a publicly redistributable corpus of international nuclear and cyber-security standards, so the architecture, evaluation criteria, and failure modes can be discussed openly and reproduced.

There are three contributions. First, a reference architecture for an on-premise, air-gapped, LLM- and language-agnostic RAG assistant. It defines document ingestion and section-aware splitting, a triple-retriever stage combining BM25, the nuclear-domain sparse encoder FERMI, and a general dense encoder, followed by fusion, reranking, and citation-constrained generation. The full stack runs on average hardware using open-source components.

Second, four evaluation criteria absent from standard RAG benchmarks: correct document and section attribution, preservation of the prescriptive shall/should distinction, correct refusal when retrieved evidence is insufficient, tested with deliberate traps mixed with legitimate questions and sensitivity to successive revisions of the same standard.

Third, an empirical test of whether FERMI, which was developed by Atomic Canyon with Oak Ridge National Laboratory, trained on the U.S. NRC ADAMS corpus, and in production at Diablo Canyon as part of the Neutron Enterprise platform, generalizes from U.S. regulatory documents to an international corpus drawn from IAEA Safety Standards, the IAEA Nuclear Security Series, NIST publications, and selected IEC, EPRI, and NRC NUREG materials.

Evaluation is planned over a 60-item ground-truth set balanced across the four criteria, run as a controlled matrix of six retriever configurations and two open-weight LLMs (Llama-3.1 and Phi-3.5) with three random seeds each. No results are reported in this submission; the full paper will present the experiments and an honest discussion of cases where the approach underperforms. The intended contribution is methodological: the reference architecture and test harness together provide a reproducible starting point adaptable to other safety and privacy-critical document collections.

Neutron moderation is central to reactor physics. Standard analytic treatments assume static, unbound nuclei and ignore molecular binding and crystal lattice effects. These approximations yield closed-form solutions but deviate from the behavior of real materials. To address two complementary needs, a controlled reference for validating neutron transport codes (deterministic, stochastic, and AI/ML-based) and an interactive teaching tool for moderation physics, we developed VERITAS[1] (Visual Explorer for Real vs. Idealized neutron Transport and Analytic Solutions). The user varies fundamental physical inputs (target nuclear mass, source neutron energy, and the functional form of the cross-sections) and directly observes their effect on slowing-down behavior, in parallel with analytic theory and Monte Carlo simulation.

VERITAS is built upon the PYDECS[2] Python module, which allows custom nuclides with user-defined scattering, absorption, and fission cross-sections. For real materials, ENDF/B-VIII.1 data are used for common moderators and reflectors (light and heavy water, beryllium, graphite, and lead). Supported interaction mechanisms include slowing-down kinetics, thermal upscattering, absorption, and leakage, evaluated in user-defined geometries and source configurations.

Computed observables include neutron energy spectra, mean free paths, average logarithmic energy decrement, and time-dependent population decay. These quantify the deviation introduced by theoretical simplifications. OpenMC is currently used as the stochastic transport solver, but any other transport code can be validated against the analytic reference solutions in the same way.

PYDECS extends VERITAS to applications beyond simplified theory, including scattering-step convolution, diffusion and slowing-down times, and neutron transmission through resonance traps across a range of atomic mass ratios. It thus serves both as a pedagogical aid and as a practical entry point to Monte Carlo methods in nuclear engineering.

[1] Downloadable at https://www.rcp.ijs.si/veritas/.

[2] Based on the IDECS Fortran code by L. Plevnik.

The JEK2 project is one of the key strategic energy investments of the Republic of Slovenia. Its purpose is to strengthen long-term energy security, reduce greenhouse gas emissions, and support the transition to a low-carbon electricity system. As renewable energy sources continue to grow and fossil fuels are gradually phased out, nuclear energy remains an important stable source of electricity generation that ensures reliability of supply and stability of the power system. JEK2 is therefore planned as a long-term national infrastructure project that requires coordination of technical, regulatory, organizational, and financial activities.

The current project phase is mainly focused on preparing the conditions for the Final Investment Decision (FID), planned for the end of 2028. The project timeline covers the period from 2026 until the expected start of commercial operation in 2041 and includes several interconnected phases such as spatial planning, nuclear licensing, supplier selection, development of the financial and business model, preparatory works, construction, and transition to operation. Due to the high level of dependency between activities, the schedule is not treated as a simple sequence of tasks, but as an integrated project timeline connecting technical, regulatory, and organizational processes.

An important milestone is the decision to start the preparation of the National Spatial Plan (DPN), adopted on 17 February 2026. This milestone marked the transition of the project into a more intensive spatial planning phase and enabled the preparation of expert studies, environmental reports, and variant analyses. The process also includes environmental impact assessment, cooperation with national authorities, public participation, and cross-border environmental procedures. Adoption of the DPN is expected in the second half of 2028 and represents one of the key conditions for approval of the FID.

Another important part of the project is the preparation of safety documentation and the nuclear licensing process. Activities began with the preparation of the Site Safety Analysis Report (SSAR), which examines geological, seismic, hydrological, climatological, and other site characteristics. The purpose of these analyses is to define the project input conditions for further engineering and technical specifications. In later phases, the Preliminary Safety Analysis Report (PSAR) and the Final Safety Analysis Report (FSAR) will also be prepared as part of the licensing process for construction and operation. Nuclear licensing activities are closely connected with regulatory requirements and have a direct impact on the overall project schedule and critical path.

Supplier selection is also a key project activity. The project is currently leaning toward the use of proven NOAK (nth-of-a-kind) technology because this approach provides greater predictability in design, licensing, and construction and reduces project and investment risks. In contrast, the use of FOAK (first-of-a-kind) technology would represent significantly higher regulatory, technical, and financial risks due to limited references and a greater possibility of delays and cost overruns. The Technical and Financial Study (TFS), completed in 2025, provides the basis for supplier selection and tender preparation. The Final Investment Decision is expected at the end of 2028, while the contract with the selected supplier is planned for 2029.

An additional important element is the development of the financial and business model. Activities include economic analyses, preparation of the financing structure, and coordination with the European Commission within state aid notification procedures. The project foresees a combination of equity and debt financing and cooperation with domestic and international financial institutions. Stable financing conditions and alignment between technical and financial planning are essential for maintaining the planned project schedule.

Because of the complexity of the project, systematic risk management is essential. The main risks include delays in spatial planning, regulatory procedures, safety documentation, financing, and supply chains. For this reason, the project is gradually developing an integrated project schedule in Primavera P6, which enables the connection of all major project activities, identification of the critical path, and continuous adjustment of the project timeline according to project development. The use of Primavera P6 also supports scenario analysis, assessment of schedule risks, and more efficient coordination between project stakeholders. Such an approach improves transparency in project management and supports better control of project complexity.

Current analysis shows that achieving the Final Investment Decision in 2028 is feasible, but only with effective management of regulatory, organizational, and financial processes, as well as timely decision-making. The JEK2 project therefore represents an example of a modern approach to managing large infrastructure projects through integrated planning and coordination of complex systems.

Due to neutron irradiation, the mechanical properties of crystalline material undergo significant changes compared to those in their unirradiated condition. Experimental tests in irradiated austenitic stainless steels and Zircaloy have demonstrated a loss of toughness and ductility, an increase in yield stress, and a reduction in the work-hardening regime. Accordingly, this study introduces a strain gradient crystal plasticity model to capture the irradiation-induced effects on the mechanical behavior of crystalline materials.

In particular, the formation of shear bands—such as slip bands and kink bands—is investigated as two primary modes of localized plastic deformation in irradiated crystals. The results show that irradiation leads to highly localized deformation patterns concentrated within the clear channels, whereas the deformation in unirradiated crystals remains nearly homogeneous. Furthermore, the influence of grain boundaries on the deformation behavior of polycrystalline materials is examined through the incorporation of higher-order traction tensors within the proposed strain gradient framework. Finally, both the classical and generalized balance laws are solved using a Fast Fourier Transform (FFT)-based homogenization method, which provides an efficient computational approach. These balance laws are explicitly coupled within the FFT framework, ensuring an accurate and consistent numerical implementation.

The JEK2 project represents a major strategic investment in low-carbon electricity generation and long-term energy security. Given the technical complexity of nuclear installations, their extended operational lifetime, and the stringent regulatory and societal expectations associated with nuclear technology, the development of a robust and comprehensive management framework is of fundamental importance. To this end, a structured Quality Management Manual (QMM) will be established for JEK2, integrating internationally recognized nuclear, quality, environmental, and occupational health and safety standards into a coherent management system.

The QMM will be designed as an Integrated Management System (IMS), consolidating the requirements of ISO 19443, ASME NQA-1, IAEA GSR Part 2, ISO 14001, and ISO 45001. The JEK2 framework incorporates these standards into a unified governance and process architecture. This integration minimizes duplication, enhances clarity of responsibilities, and ensures consistent implementation across all lifecycle phases of the project.

At the core of the management system lies the principle that nuclear safety is the overriding priority. In accordance with IAEA GSR Part 2, leadership commitment and safety culture are treated as foundational elements. The QMM will define clear roles, responsibilities, and authorities at all organizational levels, ensuring that decision-making processes are transparent, evidence-based, and demonstrably aligned with safety objectives.

From a quality assurance perspective, ISO 19443 and ASME NQA-1 provide the structural backbone of the system. The QMM will establish a graded approach. This approach will enable proportionality without compromising nuclear safety. Full traceability of materials and components, strict control of design documentation and independent verification and validation activities will ensure compliance with both regulatory and technical requirements.

Risk management will be treated as a continuous and systematic activity. The QMM will incorporate structured methodologies for identifying, assessing, and mitigating technical, organizational, and environmental risks. Formal change management procedures shall be established to evaluate the potential safety, quality, and environmental implications of design modifications, supplier changes, or organizational adjustments. Nonconformity control, corrective and preventive actions, and root cause analysis are integral components of the improvement cycle. This will ensure that deviations are not only corrected but also analysed for underlying systemic causes.

Environmental management, in alignment with ISO 14001, will be integrated into both strategic planning and operational control.

Occupational health and safety management, structured in accordance with ISO 45001, will be embedded within daily operational practices and contractor oversight. Hazard identification, risk assessment, competence management, and worker participation mechanisms will be systematically implemented.

Continuous improvement and organizational learning constitute an overarching principle of the JEK2 management system. Internal audits, management reviews, self-assessments, and benchmarking against international nuclear projects will be conducted to evaluate system effectiveness.

The structural integrity of reactor pressure vessels (RPV) shall be ensured during their entire lifetime, including the design operation and long-term operation (LTO) stages. As the RPV is irradiated with neutrons during operation, its base and weld metals get embrittled with a subsequent increase of yield strength, loss of toughness and increase of the ductile-to-brittle transition temperature (DBTT). As the RPVs lifetimes are extended, the margins to brittle-fracture initiation of a postulated crack thus reduce, and it may become challenging to prove their safe operation during an eventual pressurized thermal shock (PTS) event.

The warm pre-stress (WPS) effect has been long known and proven experimentally as the phenomenon by which ferritic low-alloy steels employed as base and weld metals in RPVs show an apparent increase of fracture toughness when they are preloaded at higher temperatures above DBTT in the ductile upper shelf region. Hardening, crack front blunting and compressive residual stresses ahead of the crack tip are the reasons for such behaviour. A structural integrity assessment of the RPV under PTS requires deterministic and/or probabilistic analyses to prove sufficient margin against brittle fracture. While application of the WPS effect has a direct positive impact on the assessment outcome, it is not fully accepted in the regulations of all countries. It is however recognized that a cracked body preloaded in the upper shelf region (at stress intensity factor K_I denoted as K_WPS) will not propagate in brittle fracture under subsequent constant or decreasing loading even if it reaches the brittle region with K_I higher than fracture toughness (K_IC), unless there is a reloading. And even after reloading, brittle fracture can occur at K_I higher than K_IC.

The expert group on mechanical systems (EG2) of the European Technical Safety Organisations Network (ETSON) has been developing a report on Comparison of Rules and Practices for Warm prestress application. In the report, several member countries have discussed their regulatory requirements, current practices, collected experimental data on WPS and performed benchmark calculations with national WPS models and compared the results against experiments. This paper presents the calculations performed with the WPS models available in the FAVOR (Fracture Analysis of Vessels – Oak Ridge) computer code employing representative load paths from the ETSON experimental data collection as inputs. Two approaches are followed to compare the probabilistic FAVOR results with the experimental K_I at fracture (K_FRAC): (i) evaluation of the probability of crack initiation when the transient reaches K_FRAC, and (ii) definition of the FAVOR K_FRAC as the K_I when the probability of initiation is equal to 0.5. The results show that the deterministic WPS models in FAVOR are typically conservative while the probabilistic model is slightly non-conservative. However, under the LUCF load path (Load-Unload-Cool-Fracture) where experimental K_FRAC is typically lower than K_WPS, the deterministic-baseline and the probabilistic models clearly behave non-conservatively.

Multiphase flow and natural convection in passive residual heat removal systems remain among the key modelling challenges in the safety analysis of Small Modular Reactors (SMRs). Safety condensers (SACO), widely proposed in advanced light water reactor concepts, rely on coupled phase-change and buoyancy-driven heat transfer phenomena that are difficult to resolve using conventional thermal-hydraulic system codes. In particular, pool-side phase change and thermal stratification remain important sources of modelling uncertainty.

The first stage of a computational fluid dynamics (CFD) investigation of multiphase natural convection under SACO-relevant thermal-hydraulic conditions is presented. The objective is evaluation of the applicability of multiphase CFD approaches for representation of buoyancy-driven multiphase flow structures, thermal stratification, and local heat transfer phenomena in large water pools with submerged condensers, representative of passive residual heat removal systems. The analysis focuses on the pool-side domain and is based on selected experiments from the Primary Coolant Loop (PKL) facility, providing representative benchmark conditions for safety condenser operation.

The presented work represents an initial step toward CFD-informed improvement of SACO modelling and establishes the baseline for future comparison with system thermal-hydraulic codes. Furthermore, the findings support assessment of the feasibility of CFD–system code coupling for SACO-like passive safety systems.

Accurate numerical prediction of thermal stratification and swirl flow penetration in externally cooled dead-end pipes poses significant challenges due to the complex interplay of buoyancy-driven forces, high turbulence intensities, and rotational flow phenomena. Numerical simulations of dead-leg configurations show that the prediction of turbulent swirl interacting with buoyancy-driven natural circulation is highly sensitive to CFD model settings, including boundary conditions, mesh resolution, turbulence models, and numerical methods such as pressure discretization schemes and solver formulations.

This study investigates the influence of domain size, outlet boundary conditions, and pressure interpolation schemes on the numerical prediction of flow and thermal characteristics within a dead-leg configuration representative of safety injection (SI) pipes in some 2-loop Westinghouse pressurized water reactors (PWRs) with direct vessel injection. Steady-state and transient Reynolds-Averaged Navier–Stokes (RANS) simulations were performed using the ANSYS Fluent code. The primary objective is to assess the sensitivity of key flow parameters — including swirl penetration depth, temperature distribution, and thermal stratification gradients within the dead-leg — to different simulation setups. This assessment is particularly relevant because not all numerical options available in the CPU solver are currently supported by the Fluent GPU solver.

To ensure mesh-independent solutions, a mesh sensitivity study was conducted using three meshes ranging from approximately 2 million (coarse) to 20 million (fine) cells. Results show that both domain size and outlet boundary condition selection have a measurable influence on the predicted thermal-hydraulic behavior of the dead-leg. Thermal stratification, quantified by the vertical temperature difference at three axial locations, varies significantly across cases, with the strongest stratification gradients occurring deeper within the pipe. The results further indicate that the domain size (i.e. outlet of the main branch flow region) influences the buoyancy-driven flow structure within the dead-leg, despite being located relatively far from the tee junction.

Lead-cooled fast reactors (LFRs) and lead-cooled small modular reactor concepts impose demanding requirements on structural integrity assessment and in-service inspection. The use of liquid lead or lead-bismuth eutectic as coolant, elevated operating temperatures, limited component accessibility and complex internal geometries create challenging conditions for ultrasonic non-destructive evaluation (NDE). Physics-based numerical modelling can therefore provide an important tool for early assessment of inspection feasibility, selection of ultrasonic parameters and interpretation of inspection signals before full experimental qualification is available.

This paper presents a physics-based modelling study of ultrasonic wave propagation relevant to NDE of LFR components. The simulations were performed using the k-Wave framework and focused on four aspects important for future inspection development: numerical resolution and computational cost, reflection coefficient estimation at acoustic interfaces, ultrasonic attenuation in heavy liquid metal media and multi-angle defect imaging.

The first part of the study evaluates the influence of spatial and temporal discretization on computational performance. A two-dimensional model with a 25 mm × 15 mm domain was used, including a 10 mm ultrasonic source and a simplified plate-like reflector. When the spatial step was reduced from 0.4 mm to 0.025 mm, the total number of grid points, including the perfectly matched layer, increased from approximately 8.0 × 10³ to 6.7 × 10⁵, while the simulation time increased from about 1.8 s to 274 s. A separate study at fixed spatial resolution showed that changing the Courant–Friedrichs–Lewy number from 0.1 to 0.9 increased the time step from 1.69 ns to 15.25 ns and reduced the simulation time from 101 s to 11 s. These results quantify the practical trade-off between numerical accuracy and computational efficiency for ultrasonic modelling of reactor-relevant inspection cases.

The second part of the study investigates reflection coefficient estimation. The theoretical reflection coefficient for the selected material interface was 0.429, based on acoustic impedance contrast. Simulated values depended strongly on excitation length, time step and signal processing method. For a 0.6 mm plate and four-cycle toneburst excitation, a CFL value of 0.1 gave reflection coefficients of 0.420 using Hilbert-envelope evaluation and 0.434 using absolute-amplitude evaluation, both close to the theoretical value. For a 10 mm plate, where overlapping multiple reflections are reduced, averaged four-cycle results at CFL = 0.3 gave values of 0.426 and 0.434, again showing good agreement with theory. In contrast, one-cycle excitation and coarser temporal discretization produced larger deviations, demonstrating the importance of pulse design and signal processing in quantitative ultrasonic simulations.

The third part addresses attenuation in liquid lead and lead-bismuth eutectic. A power-law attenuation model was implemented and compared with analytical estimates based on Stokes’ formulation for Newtonian fluids. For liquid lead at 637 K, the estimated Stokes attenuation coefficient was approximately 9.9 × 10⁻¹⁶ Np/(Hz² m), while for lead-bismuth eutectic at 673 K it was approximately 9.6 × 10⁻¹⁶ Np/(Hz² m). However, comparison with available experimental data for lead-bismuth eutectic indicates an attenuation coefficient of approximately 9.2 × 10⁻¹⁵ Np/(Hz² m), about one order of magnitude higher than the Stokes-based estimate. This suggests that viscosity-based analytical attenuation alone may underpredict ultrasonic losses in heavy liquid metal environments and that experimentally supported material models will be required for reliable LFR inspection simulations.

Finally, the study explores multi-angle ultrasonic imaging of simplified defect geometries, including triangular, semicircular and rectangular wall features. Normal-incidence inspection alone was found to be insufficient for reconstructing the true defect shape, especially when reflections were scattered away from the receiver. To address this, simulated A-scans were acquired from multiple angular positions and mapped into a common B-scan coordinate system. The combined multi-angle images provided a more complete representation of defect geometry and demonstrated the potential of angle-diverse ultrasonic inspection strategies for complex LFR component features.

The presented results establish a preliminary modelling basis for ultrasonic NDE development in lead-cooled reactor applications. By linking numerical wave propagation, reflection validation, attenuation assessment and multi-angle imaging, the work supports future design of inspection concepts for LFR and lead-cooled SMR components. Further work will focus on improved heavy liquid metal material models, experimentally validated attenuation data, more realistic component geometries and integration with practical ultrasonic probe design.

Due to increases in available computational power, the fidelity typically used in neutronics simulations has been steadily increasing. In particular, Monte Carlo codes allow the use of extremely detailed geometries, which can approximate real-world geometries to essentially any degree of accuracy. However, one of the bottlenecks of such studies is the procedure for preparing geometrical models for use in Monte Carlo codes such as MCNP or OpenMC, as models generally still need to be processed, usually manually by removing details that do not affect results at the chosen fidelity but add unnecessary complexity, to ensure compatibility. Furthermore, the size and complexity of the model necessitate compartmentalisation into smaller sub-models, as well as strict adherence to common modelling practices and numbering schemes to ensure that the model can be maintained.

In this paper, we discuss lessons learned and our current strategies in geometry preparation for neutron transport simulation models, developed through supporting the development of some of the major future fusion and fusion-relevant machines such as DEMO, VNS, and IFMIF-DONES. We also discuss some potential pitfalls of increased model complexity and provide an outlook on future directions for these processes.

Variance reduction (VR) methods are crucial for various analyses, particularly in fusion research, where radiation shielding and deep penetration problems are often computationally intensive to solve using Monte Carlo particle transport codes. The aim of VR is to increase computational efficiency without introducing bias. As VR implementations in OpenMC are still relatively new, this study evaluates the performance of weight windows (WW) and survival biasing on a test case.

The ICSBEP (International Criticality Safety Benchmark Evaluation Project) labyrinth geometry was used as a simple and well-understood benchmark case to validate implementations of VR techniques in OpenMC. Neutron flux and statistical uncertainties were examined at ten distinct tally positions along the labyrinth as well as across the entire geometry using mesh tallies, while computational efficiency was quantified using the figure of merit (FOM).

Multiple methodologies for both global and tally-specific WW implementation were investigated, including WW for global optimisation generated in OpenMC via the Magic and Random Ray methods, as well as MCNP WW converted into OpenMC. Various tests were performed to determine appropriate values for different parameters, which need to be set when generating and using WW generated in OpenMC. Based on these results, the identified parameters served as a basis for the final simulation setup. Furthermore, the impact of survival biasing and the ‘statistical weight’ cutoff parameter was investigated both with and without WW, while using different optimization schemes.

The tests and analyses from this work quantify the effectiveness of VR implementations in OpenMC and can be used as a guideline on how to set up various VR parameters for OpenMC simulations.

In gamma-ray spectroscopy education, learners often treat tabulated photon emission energies and intensities as if they directly represent the spectrum measured by a high-purity germanium detector (HPGe). This simplification can obscure several essential aspects of detector response, including photon interaction mechanisms, detector geometry, energy-dependent full-energy peak efficiency, Compton scattering, escape-peak formation, background contributions, and counting statistics. To support teaching of these concepts, an interactive Application has been developed for generating HPGe spectra from user-defined gamma-ray emissions using forward modelling with a precomputed detector response matrix.

The method is based on response-matrix folding. A detector response matrix (R) is calculated in advance with the MCNP code for a specified HPGe detector and measurement configuration. The matrix is discretised in incident photon energy and detector output channel. Each column of R represents the expected spectrum over detector channels for monoenergetic photons within one incident-energy bin, as represented by the Monte Carlo model. This response includes contributions such as full-energy deposition, Compton continua, and escape structures according to the simulated geometry and physics treatment. For a selected source definition, an incident photon vector (x) is assembled from gamma-ray energies and emission probabilities. The photon intensities may be scaled by activity and live time, allowing users to examine how source strength and acquisition duration affect the expected spectrum. The detector spectrum is then calculated by the linear forward model (y = R x).

The application provides two main input methods. In the first, users manually enter gamma-ray energies and emission probabilities, which is useful for exploring idealised or simplified source cases. In the second, users select radionuclides from an isotope emission library. The interface also supports mixtures of multiple radionuclides, either through individual isotope activities or through fractional composition combined with a total activity. Manual gamma lines may be added to isotope-library selections when needed for demonstration purposes. Additional options include flat or file-based background addition, optional Poisson sampling to illustrate counting statistics, interactive linear and logarithmic plotting, gamma-line and escape-peak annotations, diagnostic summaries of signal and background counts, and export of both the folded detector spectrum and the incident photon distribution.

The tool is intended primarily as a teaching and visualisation aid rather than as a replacement for full detector simulation or experimental calibration. Its purpose is to make the mapping between emitted photons and measured HPGe spectra visible and adjustable in real time.

By allowing users to vary source composition, intensity, background, and statistical treatment while observing the resulting detector response, the application provides a compact platform for teaching forward-folding concepts and for developing intuition about HPGe spectra in nuclear engineering education.

The spent fuel pool (SFP), where spent nuclear fuel is being stored prior to being moved to long-term storage, is an important part of any nuclear power plant (NPP). Failure of cooling of the spent fuel elements may lead to an accident with disruptive consequences for the entire nuclear facility.

SFPs of existing NPPs were (probably) designed using simplified (that is, volume-averaged and one-dimensional flow) descriptions. So far, the lack of accidents has proved that the cooling was suitably designed, thus such methods have been successful. However, the detailed flow and temperature conditions in SFPs remain mostly unknown.

One possibility to obtain such insights would be detailed measurements of flow velocity and temperature in either actual SFPs (which would be impractical) or suitable experimental facilities (which would be costly and would probably not justify the costs). However, the development of Computational Fluid Dynamics (CFD) makes it possible to simulate the conditions in SPFs on the local instantaneous scale. Although the results of such simulations are still uncertain, they can provide useful insights on the cooling of fuel elements. Specifically, simulations can provide an assessment of actual cooling safety margins. This is a valuable incentive for the development of such methods.

In the proposed work, flow and temperature conditions in a model of a rack in an SFP were simulated on the local instantaneous scale using a CFD code. The rack contains four fuel elements, arranged in a square lattice. The dimensions of the rack and of the fuel elements are based on an international benchmark exercise (in which system codes devised for reactor cores were used) that was organised within the NUGENIA association in 2015-2016. As the emphasis is on the conditions in the coolant surrounding the fuel elements, the elements were modelled as porous media. The simulated velocity and temperature fields in the coolant surrounding the elements are presented and analysed. Because of the unavoidable uncertainties, parametric simulations were also performed to ensure that no “cliff-edge” effects were present.

In a pressurized water reactor (PWR), the upward coolant flow through the reactor core should be as homogeneous as possible. This is all the more difficult to achieve due to, first, the abrupt change of flow direction of the coolant after the downward flow through the downcomer of the reactor pressure vessel (RPV) and, second, the many different structures in the lower plenum of the RPV. The flow (allegedly) becomes suitably homogeneous after flowing through suitable arrangements of circular openings of appropriate size in plates below the reactor core, specifically the lower support forging and the lower core plate. However, due to some arrangement of support as well as other systems, some openings of the lower support forging are filled up. Namely (for whatever reason), the design of the lower support forging is such that some openings are used for support structures instead of locating those structures between the openings. A consequence of this might be that the flow through the reactor core is not as homogeneous as it could be without these asymmetric arrangements of openings through which the coolant actually flows.

To investigate the possible asymmetric flow caused by such arrangements, the flow in the lower part of a pressurized water reactor vessel lower plenum between the inlet of the lower support forging and some upward distance from the lower core plate was simulated, using a local instantaneous description. Only part of the RPV horizontal cross-section was considered. Different arrangements, symmetric as well as asymmetric, of openings in the lower support forging were considered. The dimensions and arrangements of the openings were based on a Western-type PWR.

The results showed that the asymmetry of the arrangement of openings in the lower support forging does not significantly affect the homogeneity of the velocity field above the lower core plate. This supports the suitability of the adopted designs in such reactors.

The zirconium-based cladding with protective chromium coating (Zr/Cr cladding) is one of the most promising perspective advanced tolerant fuel (ATF) cladding candidates for application in commercial nuclear power plants (NPPs) throughout the world. This cladding material has excellent characteristics of corrosion and oxidation resistance compared to zirconium as well as good mechanical robustness both for the NPP normal operation temperatures and high-temperatures conditions. Also, this ATF concept does not require major modifications to existing core design (the evolutionary way of nuclear energy development). Currently, it is known from numerous experimental data that in the temperature range close to upper limit of design-basis accident (T=1200C) and higher there is a considerable worsening of Zr/Cr cladding protective properties. In particular, a role of Zr-Cr inter diffusion and the Zr-Cr eutectic reaction with subsequent influence on degradation of protective properties are revealed. The severe accident integral experiment CODEX-ATF was conducted on the 23rd August 2023 at HUN-REN Centre for Energy Research, Institute for Atomic Energy Research, Budapest, Hungary. The test bundle contained different fuel rods simulators with both ATF Cr-Coated abd conventional Zr Claddings in order to make it convenient to compare high temperature behaviour of ATF and non-ATF claddings. The experiment CODEX-ATF was one of the first integral severe accident tests (the maximum temperature T=1660C) with ATF Zr/Cr claddings. This is why the experimental data obtained are exclusively valuable. The new advanced models of high-temperature Zr/Cr cladding oxidation were developed in the paper. In particular, the model was based on simultaneous solution of oxygen, chromium and zirconium diffusion equations in different layers of the cladding including inter metallic layer ZrCr2. A very important role of Zr outward diffusion and Cr inward diffusion to the interface between chromium and zirconium resulting to severe degradation of protective properties were taken into account. The models were implemented to newly developed severe accident computer running code. The output parameters of the numerical calculation (the temperature history, the temperature axial profile, the hydrogen production etc.) we're compared with experimental data. The comparison between calculation and experimental results showed a good predictive ability of the models and the numerical code developed in the paper.

Convective boiling is one of the most efficient heat transfer mechanisms and is widely applied in energy conversion and thermal management systems, including nuclear engineering, process technology, refrigeration and electronics cooling. Its use is, however, limited by the occurrence of boiling crisis and the associated critical heat flux, where a rapid deterioration of heat transfer may lead to excessive thermal loads and component damage. Reliable prediction of boiling flows remains challenging, particularly because two-phase CFD models still have limited capability to reproduce local vapour distribution, bubble size and near-wall heat transfer over a broad range of operating conditions.

This contribution presents an overview of a two-year bilateral research project between the Jožef Stefan Institute (JSI) and the French Alternative Energies and Atomic Energy Commission (CEA), focused on experimental investigation and modelling of bubble dynamics in convective boiling.
Experiments were performed in the THELMA laboratory at JSI using a transparent annular test section, where boiling refrigerant R245fa flows through a 2 mm thick channel between an inner water-heated copper tube and an outer borosilicate glass tube. Thermocouple measurements in the copper tube provide axial temperature distributions with inferred heat fluxes, while simultaneous high-speed recordings enable visual observation of boiling refrigerant. Investigated operating conditions consider vertical and horizontal configurations at a system pressure of 2.6 bar and refrigerant inlet subcooling of 14 K, with water-side heating powers of 0.1-0.9 kW and refrigerant mass fluxes of 150-600 kg/m²s. Neural-network-based bubble detection was applied to high-speed recordings to obtain distribution of bubble size and vapour volume fraction.
In parallel, selected operating conditions were analysed with two-fluid CFD simulations in OpenFOAM (JSI) and TrioCFD (CEA). Preliminary simulation results show strong sensitivity to wall boiling closure-model choices and in general void fraction is under-predicted compared to present experimental estimates. The final project stage will use these findings to guide model improvement regarding wall-boiling closures and bubble dynamics in the bulk flow.

The leakage on the SI-53 safety injection line at the Krško Nuclear Power Plant (NPP) in October 2023 motivated a detailed investigation of thermo-hydraulic phenomena in a horizontal dead-end safety injection pipe. The direct cause analysis indicated that cyclic thermal loading has contributed to the crack growth in the SI-53 line. Since the SI-53 line is directly connected to the downcomer region, the unsteady flow in downcomer may represent an important source of instabilities that drive thermal loading in the affected pipe section.

The objective of this work was to numerically investigate flow instabilities in the downcomer during normal operating conditions. Computational Fluid Dynamics (CFD) simulations were performed in OpenFOAM v10 using an isothermal and incompressible transient solver with a hybrid Detached Eddy Simulation (DES) approach. The numerical model reproduced turbulent flow in the entire domain of the downcomer, including both inlet legs, and captured the interaction between the downcomer flow and the dead-leg entrances of the two safety injection lines. The simulations revealed a highly non-uniform and inherently unsteady flow, with instabilities originating from flow separation regions and persisting throughout the entire streamwise extent of the downcomer. Special attention was devoted to an unsteady swirling structure penetrating into the SI-53 line and inducing disturbances to the recirculating flow inside the dead-leg section.

Within the SEAKNOT European project (2022-2026), a Phenomena Identification and Ranking Table (PIRT) was prepared to determine the research priorities concerning phenomena that are expected to occur during severe accidents in light-water reactor nuclear power plants.

The following features were ranked as either high, medium or low:

— phenomena knowledge (that is, existence of relevant experimental data);

— theoretical knowledge (that is, suitable models having been developed and validated);

— safety relevance (that is, influence of the considered phenomena on adverse accident consequences).

For containment phenomena, in the first step, rankings concerning the three features were submitted by individual organizations and collected. Then, major discrepancies between rankings, when existing, were discussed and some individual rankings were voluntarily changed. In this way, a consensus was reached and “resulting” rankings were obtained.

The second step was to combine, according to prescribed rules, the resulting rankings of “phenomena knowledge” and “theoretical knowledge” into a single resulting ranking (again, as either high, medium or low) referred to as “knowledge”.

The final step was to combine, again according to prescribed rules, the resulting rankings of “knowledge” and “safety relevance” into final rankings of the phenomena (again, as either high, medium or low) that determine the research priorities.

In the field of containment phenomena, seven contributing organisations submitted rankings for the PIRT. After applying the described procedure, the following phenomena were ranked as having the highest priority for further research, both for the so-called "in-vessel" as well as for the so-called "ex-vessel" accident phase:

— heat transfer from structures to atmosphere causing buoyant flow in the containment atmosphere;

— hydrogen combustion in the containment atmosphere;

— spray mass and energy exchange resulting in depressurization of the containment atmosphere.

All three considered phenomena were assigned a high safety relevance. Also, for all the three phenomena, the experimental knowledge was considered as high, whereas the theoretical modelling was considered either as lacking or existing models not having been sufficiently validated for specific severe accident conditions.

In the proposed paper, the rationales for such rankings for the three phenomena are considered in detail (including a review of the most relevant open literature and other available information) to confirm that these three phenomena that came standing out following the PIRT preparation should indeed be assigned high research priorities. Such results are of great importance, as they may influence or even determine future research programmes and projects in the field of containment phenomena during severe accidents.

Small Inventory Member States (SIMS) face unique challenges in radioactive waste management (RWM). Although they are required to comply with the same regulatory and safety expectations as countries with larger radioactive waste inventories, they often operate with substantially more limited financial, technical and human resources. The EURAD-2 Strategic Study (StSt) on Alternative RWM STRAtegies (ASTRA) addressed SIMS RWM challenges by investigating alternative waste management approaches and identifying opportunities for cooperation, knowledge sharing and stakeholder engagement.

The aim of ASTRA was to analyse management strategies for small and heterogeneous waste streams, investigate shared approaches of pre-disposal and disposal activities, and evaluate strategic issues that are often insufficiently addressed in existing programmes, including governance, maintenance of specialist competencies, and long-term safety. An additional objective was to strengthen interaction between countries with small inventories (SIMS) and those with more advanced programmes and larger inventories (Large Inventory Member States – LIMS), enabling exchange of practical experience and lessons learned.

The StSt is based on a combination of literature review, expert consultation, workshops, Live Discussion Forums (LDFs) with SIMS in tandem with LIMS representatives, and three dedicated surveys developed within the tasks. The surveys addressed (i) management strategies for small amounts of radioactive waste, (ii) opportunities and barriers for shared waste management solutions, and (iii) strategic aspects relevant to long-term implementation. Civil Society (CS) representatives contributed to the development of the surveys and participated throughout the discussions by bringing up societal perspectives, transparency, and public participation considerations.

Part of the StSt focused solely on the management of waste containing naturally occurring radioactive materials (NORM), uranium-, thorium- and radium-bearing wastes, as well as depleted uranium (DU). This subtask consolidated international experience on lifecycle management, identified gaps in disposal readiness across Member States, and highlighted the need for harmonised classification approaches, improved safety case methodologies, and enhanced technical and regulatory coherence for NORM and DU waste streams. Of special interest for CS participants was the management of uranium mine and mill tailings.

The results demonstrate that many challenges faced by SIMS are common across Europe, regardless of differences in national programmes. Waste characterisation and separation-sorting at source were consistently identified as fundamental prerequisites for effective waste management. Participants highlighted the need for disposal-oriented waste management approaches, improved decision-support tools and access to practical guidance for managing waste streams that do not readily fit existing waste acceptance criteria.

Knowledge management emerged as one of the most important enabling factors. Survey responses and workshop discussions revealed strong support for structured exchange mechanisms that facilitate transfer of experience between more advanced and less advanced programmes. In particular, participants recognised the value of permanent problem-oriented discussion platforms that allow regulators, waste management organisations, technical support organisations, research entities, CS members and other stakeholders to exchange practical solutions and lessons learned. Therefore, in response to the identified need for enhanced knowledge transfer, LDFs were developed, and their long-term continuation is recommended to ensure a structured knowledge transfer mechanism.

Shared solutions were identified as a potentially important component of future waste management strategies for SIMS. Respondents recognised benefits related to improved cost-effectiveness, more efficient use of limited resources, and enhanced access to specialised expertise and infrastructure. However, significant challenges remain, including governance arrangements, allocation of long-term responsibilities and liabilities, regulatory compatibility, transport requirements, and public participation on a level playing field. Related survey findings indicate that successful implementation of shared approaches will require not only technical development but also institutional, political and societal frameworks capable of supporting long-term cooperation.

The involvement of CS provided valuable additional perspectives. Discussions highlighted the importance of transparency, meaningful stakeholder engagement, and early public participation in decision-making processes. CS representatives emphasised that trust in both national and multinational waste management solutions depends on clear accountability, independent oversight and effective participation including communication of risks, uncertainties, and responsibilities. The identified factors were considered particularly relevant for shared solutions where responsibilities may extend across national borders and multiple generations.

Based on the surveys and workshops, a dual approach (national implementation pathways and shared, multinational solutions) is proposed for strengthening waste management resilience in SIMS: (i) the national pathway focuses on independent development of waste management strategies, including characterisation, treatment technologies and decision-support tools and (ii) the shared pathway builds on cross-border cooperation, enabling access to specialised expertise, infrastructure and coordinated solutions. Together, the complementary approaches provide flexibility in addressing the diverse constraints faced by SIMS.

The outcomes developed in cooperation with CS representatives provided input for future EURAD activities and broader European discussions on RWM. StSt findings show that sustained cooperation, knowledge sharing, and stakeholder involvement are essential elements for developing resilient and proportionate waste management strategies, particularly for Member States managing relatively small radioactive waste inventories.

Acknowledgement. The results presented in this paper have been generated within the European Partnership on Radioactive Waste Management (EURAD-2). EURAD-2 has received funding from the European Union’s Euratom Research and Training Programme (EURATOM) under grant agreement No 101166718.

Passive nuclear track detectors permanently record the trajectories of incident particles within the detector material, yet the performance of their automated analysis still offers room for improvement. This work, carried out as part of the EU‑funded project PANDORA, presents a comprehensive processing pipeline for images acquired from PADC solid‑state nuclear track detectors (SSNTDs) using the TASLImage scanning system. First, two segmentation approaches are evaluated: a classical Otsu‑thresholding method and a modern transformer‑based Cellpose‑SAM model. Experimental results show that Cellpose‑SAM provides substantially better agreement with reference measurements and successfully separates overlapping tracks that conventional thresholding cannot reliably distinguish. Following segmentation, deterministic track parameters are extracted via principal component analysis (PCA) and compared with predictions from deep learning regression models fine‑tuned on a dataset obtained during acceptance testing at the Slovak Metrology Institute. For deterministic geometric parameters, the regression models achieve accuracy comparable to PCA‑based measurements (R² > 0.999). In contrast, shape‑related characteristics remain challenging to predict (R² < 0.32), indicating their inherently stochastic nature. Overall, the results demonstrate the promise of combining AI‑based segmentation with deep learning regression to advance automated analysis of nuclear track detector images.

The design of technological systems for the second block of the Krško Nuclear Power Plant (JEK2) requires a precise definition of the physicochemical and biological properties of the source water. Since the Sava River will serve as the primary source for makeup water for the cooling tower and other processes, the establishment of a robust water pretreatment system is crucial for heat transfer efficiency, energy yield, and operational safety.

The main engineering challenge when using water from the Sava River is preventing corrosion of structural materials, mineral deposition (scaling), and biological fouling. To ensure the uninterrupted operation of the cooling tower fills and piping, it is necessary to provide softened water free of total suspended solids (TSS) and microbiological contamination. Following additional purification, a portion of this water is also used in the primary and secondary circuits.

Technological Solutions and Treatment Stages

To achieve the required specifications, a multi-stage purification process is considered:

• Mechanical Pretreatment: Automatic bar screens, drum screens, and sand filters at the intake remove branches, leaves, sand, and larger suspended solids, protecting pumps and piping from damage.
• Coagulation and Clarification: Using lamella clarifiers and coagulants, fine colloidal particles combine into flocs and settle. Ultrafiltration (UF) then completely removes microorganisms, algae, and turbidity.
• Chemical Softening: By adding hydrated lime in combination with cationic exchangers, calcium and magnesium are precipitated, which lowers hardness, stabilizes pH, and prevents scale formation and corrosion.
• Advanced Demineralization: Two-stage reverse osmosis (RO) removes over 99% of dissolved salts, while electrodeionization (EDI) and mixed-bed ion exchange columns finally purify the water to an ultra-pure grade.

The choice of the optimal technological configuration is directly dependent on a detailed understanding of the Sava River's dynamics. Since natural watercourses exhibit high variability, the analysis must be conducted continuously across all seasons, under various weather conditions, temperature regimes, and extreme water levels (droughts and flood waves).

Main Parameter Groups for Water Quality Monitoring

For the precise dimensioning of the water treatment plant, each sample must include an analysis of the following parameter groups:

• Sensory and Field Measurements: Perceptible water properties (color, odor, foam, sewage) and field measurements (water/air temperature, pH, turbidity, water level, dissolved oxygen, redox potential).
• General Physicochemical Parameters: Total dissolved solids (TDS) and total suspended solids (TSS), total hardness, alkalinity, and conductivity.
• Ionic Composition and Inorganic Parameters: Key cations and anions, specific impurities (silica, phosphorus, bound nitrogen, dissolved gases), and metal analysis (iron, manganese) in filtered and unfiltered water.
• Biological and Microbiological Parameters: The presence of bacteria (risk of aerosol emissions from cooling tower) and acute toxicity tests (e.g., for water fleas/Daphnia).

Characterization of Technological Sludge

The byproduct of the coagulation, softening, and filtration processes will be technological sludge. As part of the design phase, it is necessary to accurately analyze the content of heavy metals and hazardous substances that could be present in specific concentrations within the technological sludge.

Only systematically and long-term collected input data will enable the correct choice of technology, optimization of chemical consumption, and an objective comparison of technical solutions for the safe and economically sustainable operation of JEK2.

The long-term operation of nuclear power plants requires continuous assessment of neutron irradiation effects on safety-significant reactor components. Among these, the reactor pressure vessel (RPV) is particularly important because prolonged exposure to fast neutrons may lead to material embrittlement and degradation of mechanical properties. Accurate determination of neutron flux and irradiation levels at the RPV is therefore essential for evaluating vessel integrity and supporting plant lifetime extension programs.

This study investigated burnup-dependent variations in fast neutron irradiation of the reactor pressure vessel at the Krško Nuclear Power Plant (NPP Krško, NEK). Neutron transport calculations were performed using the Monte Carlo N-Particle (MCNP) code. A simplified MCNP model of the NEK containment building was enhanced with a detailed geometrical and material description of the reactor core, a detailed fixed neutron source, and variance reduction parameters generated using the ADVANTG code.

The validated model for the 25^th reference fuel cycle was used to calculate the fast neutron flux (E > 1 MeV) within the reactor pressure vessel. To obtain detailed spatial information, the vessel region surrounding the active core was subdivided into voxels, which enabled 3d neutron flux mapping. The analysis focused on multiple burnup states within a representative fuel cycle, including beginning-of-cycle (BOC), intermediate, and end-of-cycle (EOC) conditions. In addition to characterizing in-cycle irradiation variations, the study assessed the applicability of using averaged BOC and EOC conditions to represent a complete fuel cycle.

The developed methodology provides a high-fidelity framework for assessing fast neutron irradiation of the NEK reactor pressure vessel and contributes to improved understanding of burnup-dependent irradiation effects relevant to aging management and long-term plant operation.

Predicting the structural lifetime of fusion reactor components requires a fundamental understanding of radiation-induced defect evolution in reduced-activation steels and model iron (Fe) systems. This study investigates defect formation and thermal evolution in EUROFER97 and Fe single crystals as a function of displacement per atom (dpa). We utilize a complementary experimental approach combining Rutherford Backscattering Spectroscopy in channelling configuration (RBS-C) with Nuclear Reaction Analysis (NRA) for deuterium (D) retention measurements, and a comparison with atomistic simulations.

Fe (100) single crystals were irradiated with MeV ions at temperatures from cryogenic conditions to 290 K. At 290 K, RBS-C measurements revealed no detectable increase in structural disorder, indicating high interstitial mobility and rapid dynamic recombination. However, D retention experiments indicate the presence of residual vacancies (~0.04 at. %), demonstrating that vacancy-type defects persist even when interstitial-related disorder is not observed by RBS-C. At irradiation temperatures below –50 °C, a clear increase in the RBS-C disorder signal with increasing damage dose is measured, consistent with enhanced defect stabilization at low temperatures. Comparison with Molecular Dynamics (MD) and Object-Based Kinetic Monte Carlo (OKMC) simulations, extending the timescales, indicates that small C15 interstitial clusters dominate at low doses and produce minimal channelling contrast, explaining the weak RBS-C response.

For EUROFER samples irradiated up to 0.6 dpa at temperatures between 300 K and 700 K, D retention measurements show a strong temperature dependence of vacancy concentration. The D concentration decreases with increasing irradiation temperature, reaching levels comparable to unirradiated material at 700 K. Comparison with pure Fe and Fe-Cr model alloys demonstrates that alloying elements (Cr, C) enhance vacancy stabilization and deuterium retention. Careful consideration of post-irradiation D outgassing is necessary for accurate quantification.

These findings provide quantitative insight into vacancy and interstitial dynamics in fusion-relevant steels. They showcase the highly complementary sensitivities of RBS-C and D retention techniques while delivering robust experimental benchmarks to validate predictive multiscale defect channelling and modelling.

Reliable operation of medium-voltage motors is essential for maintaining availability and safety of equipment in nuclear power plants. Although these machines are designed with significant operating margins, gradual degradation caused by bearing wear, insulation ageing, cooling system deterioration and mechanical imbalance can still occur during long-term operation. Early-stage degradation typically manifests as subtle changes in temperatures, vibrations or electrical parameters that may remain undetected by conventional monitoring systems based on fixed alarm thresholds. This study focuses on the analysis of different simulated fault scenarios and the evaluation of threshold selection methods in a regression-based neural network fault detection system. A feed-forward neural network architecture is employed and remains unchanged throughout the study to isolate the influence of threshold definition on detection performance. A historical dataset containing 18 months of operational measurements was used, including winding temperatures, bearing temperatures, and process-related operating parameters. Prior to model training, the dataset was subjected to preprocessing steps such as data validation, removal of corrupted records, handling of missing values and normalization. The neural network model was trained exclusively on data representing normal operating conditions, enabling it to learn the relationships between input process variables and expected motor behaviour. During operation, deviations between predicted and measured values (residuals) are used as indicators of abnormal conditions. Different simulated fault scenarios were introduced to represent various types and intensities of degradation. Based on these residuals, multiple threshold-setting approaches for alarm triggering were defined and compared, including both fixed and statistically derived thresholds. The performance of each approach was evaluated in terms of detection sensitivity, robustness, and false alarm rate. The results highlight the significant impact of threshold selection on the effectiveness of neural-network-based fault detection systems. While the neural network model provides a stable baseline for anomaly detection, the choice of threshold method largely determines detection accuracy and reliability. The findings contribute to improved design of alarm strategies in predictive maintenance systems, enabling more reliable and timely fault detection in critical rotating machines.

The deployment of light water small modular reactors (LW-SMRs) and advanced modular reactors (AMRs) is increasingly presented as a pathway to flexible, low-carbon and potentially cost-effective nuclear energy. However, their safe and sustainable implementation depends not only on reactor design, licensing and economics, but also on early legislative preparedness, clearly allocated institutional responsibilities and credible radioactive waste management (RWM) arrangements across the full fuel cycle. In the frame of EURAD-2 FORSAFF (EURAD-2 Partnership, Grant Agreement 101166718), key knowledge gaps are identified for the characterisation, treatment, conditioning, storage, transport, reprocessing and disposal of waste streams arising from LW-SMRs and AMRs.

The analysis shows that, although LW-SMRs build on comparatively mature light-water technology, their modularity, decentralised deployment models and potentially novel spent-fuel and waste inventories may challenge existing waste management infrastructures, waste acceptance criteria and regulatory assumptions. For AMRs, uncertainties are more substantial, including limited operational experience, unconventional coolants and fuels, chemically reactive or irradiated materials, activation products, tritium and carbon-14 inventories, and the long-term behaviour of novel waste forms. These uncertainties have direct implications for repository safety cases, transport and storage requirements, treatment and conditioning strategies, and the feasibility of reprocessing, recycling or fuel take-back options.

The paper therefore argues that RWM provisions must be integrated from the earliest stages of reactor development, energy-policy formation and licensing, rather than treated as a downstream technical issue. Particular emphasis is placed on adaptive legislation and regulatory guidance covering technology-specific waste routes, cross-border transport, ownership and liability models, emergency preparedness, funding mechanisms, and the compatibility of SMR/AMR waste streams with existing and future disposal facilities. Such provisions are necessary to avoid orphan waste, late-stage licensing conflicts, fragmented national solutions and loss of public confidence.

The Slovenian situation provides a relevant national case for this discussion. Slovenia already has nuclear experience through the Krško nuclear power plant and radioactive waste management institutions, but future decisions on new nuclear capacity, including potential SMR/AMR-related pathways, would take place in a governance context where stakeholders (in the frame of ECOSENS project Grant Agreement 101060920) have reported weak implementation of participatory practices, limited transparency, insufficiently visible use of public comments in policy decisions, and a need for stronger legislative and administrative support for participation. The stakeholder survey indicates that Slovenian respondents considered radioactive waste management comparatively more participatory than several other nuclear-policy domains, while still highlighting gaps in information quality, emergency preparedness, public dialogue, access to independent expertise and trust-building mechanisms.

For Slovenia, this implies that SMR/AMR deployment cannot be assessed solely through technical feasibility or energy-system modelling. It should also be examined through a national governance framework that links nuclear legislation, environmental assessment, RWM policy, spatial planning, emergency preparedness, financing and stakeholder engagement. Early clarification of responsibilities among government, regulators, technical support organisations, waste management organisations, operators, vendors, local communities and civil society would be essential. Equally important would be the creation of transparent engagement processes in which stakeholders can scrutinise assumptions on waste inventories, transport routes, storage options, disposal compatibility, long-term liabilities, costs and intergenerational fairness.

By connecting FORSAFF technical findings with the Slovenian stakeholder-engagement evidence from ECOSENS, the paper proposes a more integrated agenda for SMR/AMR governance: representative waste inventories and technology-specific waste routes should be developed in parallel with adaptive regulation, accessible information, independent review and structured public participation. This approach would strengthen regulatory preparedness, reduce long-term uncertainty and cost, improve the legitimacy of decisions on future nuclear technologies, and support public confidence in the safe and sustainable management of radioactive waste.

This paper presents the Slovenian National Strategy for Research and Development in the Field of Peaceful Uses of Nuclear Energy and Ionizing Radiation. Formally adopted by the Government of the Republic of Slovenia in December 2025 and currently in its implementation phase, the strategy provides a comprehensive, forward-looking roadmap.

It addresses not only nuclear and radiation safety but also a wide spectrum of complementary areas, including radioactive waste management, the decommissioning of nuclear facilities, emergency preparedness and response, physical protection, cybersecurity, medical applications of ionizing radiation and the development of specialized academic study programmes. The strategy establishes a framework to strategically define the research and development activities required to support the country's current and future nuclear programmes.

A key feature of the strategy is its holistic approach, which emphasizes strengthening both programmatic structures and the broader research community. Particular attention is dedicated to human resources—specifically education and training—as well as securing the stable, adequate, and independent funding essential for effective long-term research programmes.

Furthermore, dedicated working groups composed of recognized national experts are envisioned to prepare research programme proposals for the upcoming five- and ten-year periods.

Overall, this strategy represents a major step toward establishing a well-structured and sustainable nuclear research sector in Slovenia while fully meeting international commitments.

The scrape-off-layer (SOL) and the divertor form the exhaust system of the tokamak reactor. Particles and energy cross the last-closed-flux -surface (LCFS or the separtrix) via anomalous transport, which is mainly a result of coherent structures named blob-filaments penetrating from hot confined plasma into SOL region upstream and travelling along the field lines towards the divertor targers. The radial footprint of the heat flux can usually be described by a single exponential drop, however there is often a glimpse of the so-called narrow feature right at the strike point, where the heat flux is significantly higher than expected. There are several kinetic phenomena observable around this region, such as the discrepancy between the ion and the electron temperature and the highly negative floating potential, which results in negative, nonzero grounded current.

We have tried to explain these observations by employing a 1d3v fully-kinetic particle-in-cell code BIT1 for simulations of a SOL flux tube from one target to the other. The particle and energy source used was a time dependent injection mimicking the blob-filament injection. We were particularly interested in the radial losses of both particles species based on their dwelling time in the simulation domain and on the difference between the ion and the electron temperature forming due to particular parallel and radial transport features. The simulations performed describe the flux tube closest to the separatrix and the one next to it. The results showed that it is possible to get the discrepancy between the two temperatures using the described approach, which opens a new research approach to precise modelling of the SOL and will have to be validated in 2d3v kinetic simulations in the future.

Control of Special Processes is one of the 18 criteria found in 10CFR50 Appendix B - Quality Assurance Criteria for Nuclear Power Plants. Welding, brazing, non-destructive testing, coating and heat-treating are the main special processes that have to be controlled and results of which is highly dependent on the control of the process or the skill of operators and in which the specified quality cannot be readily determined by inspection or test of the process item. Almost all plants are in the process of extending plant operating licenses and special processes play a critical role in establishing conditions and providing reliable data for the technical justification to extend operations.

At Krško NPP the Quality Control Department provides Control of Special Processes as a part of Quality and Nuclear Oversight Division. In light of controlling special processes, the Quality Control Department focuses on the control of welding and heat-treating processes in quality-affecting activities, using non-destructive examinations (NDE) being implemented and applicable to our personnel.

This paper describes quality control of manufacturing of Radioactive Waste (RW) Containers at Krško NPP. Radioactive Waste Containers (i.e. drums) at Krško NPP are under ownership of Chemistry department. There are several types of RW containers and drums used for packing Solid Low and Intermediate Level Radioactive Waste (LILRW) such as: Standard 208 litre carbon steel drum, 320 litre carbon steel overpack drum, 200 l Stainless Steel heavy drum with biological shield (150 litre of usable volume), 200 l Stainless Steel heavy drum without biological shield, 869 litre tube type container (TTC), etc. Some containers are mass-produced, others are custom-made to be compatible with NPP Krško’s radioactive waste treatment systems. Each waste container has its own specification with criteria that must be met to ensure safe storage and transport. The waste containers used at the Krško NPP are designed and tested in accordance with IAEA Transport Regulations. According to these regulations, containers are divided into two groups, i.e. (1) Type A Package (all D and H type containers) and Industrial Package Type 2 (IP2 – tube-type containers T1 and T2). For this reason, QC tests and inspections are performed immediately after production and every other year for individual container types, stored in Radioactive Waste Storage Building (RWSB) in Krško NPP. Scope and frequency of inspection is defined in plant procedures. An inspection interval shall not exceed three years without technical justification.

The importance and role of quality control in the production and condition monitoring of RW containers is crucial for maintaining their function throughout their intended service life.

The Dispatchable Adaptive Reactor With Interchangeable componeNts (DARWIN) is a concept for a highly flexible nuclear reactor intended for both routine and emergency applications. Its modular design relies on interchangeable modules, each optimised to specific functions, e.g., pumping of flood water and the provision of clean, dispatchable electricity. Such flexible low-carbon generation is expected to play an important role in future electricity systems with a high share of intermittent renewable energy sources, replacing the load-following role currently fulfilled primarily by gas power plants.
The fast response module is intended for both steady state energy generation and operation in the load‑following mode. Power variations during load‑following induce several coupled effects within the reactor core. Fuel and moderator temperature profiles change directly with power and the concentrations and distributions of neutron poisons ¹³⁵Xe and ¹⁴⁹Sm evolve over time.
The significance of these effects depends strongly on reactor characteristics, including core size, fuel composition, moderator and coolant properties, as well as the operating temperatures and pressures of the primary system. Since the DARWIN core design is still in an early stage of development, it is important to identify which parameters and physical effects are most critical for the performance of the fast response module and for future power conversion strategies.
A literature review of existing research on load-following reactor operation was conducted, with emphasis on the influence of fuel type (UO₂, MOX, UZrH_X, UYH_X) and key kinetic parameters, including neutron generation time (Λ), delayed neutron fraction (β), coolant temperature coefficient of reactivity (α_c), and fuel temperature coefficient of reactivity (α_f). The relevance of these parameters was subsequently evaluated using a two-point reactor kinetics model tested against representative electricity production and demand scenarios.

Convective boiling experiments with high spatial and temporal resolution are essential for the development and validation of CFD closure models relevant to fusion divertor cooling applications. However, experimental investigations under fusion-relevant high heat flux conditions remain limited by the challenging operating environment, including high pressures, elevated temperatures, opaque heating surfaces and restricted optical accessibility. Existing experimental facilities generally provide limited visualization and diagnostic capabilities, preventing detailed observation of boiling phenomena and associated interfacial dynamics. To address these limitations, the high heat flux flow boiling visualization facility FEDORA (Fusion Experiment for Divertor Optimization Research Applications) has been designed and recently upgraded for investigations under conditions representative of several fusion divertor concepts, including DEMO, DTT and W7-X. Through appropriate dimensional analysis and the use of a surrogate working fluid, the facility enables the investigation of fusion-relevant boiling phenomena at significantly lower heating power while maintaining similarity of the governing thermal-hydraulic mechanisms.

The present work focuses on the development and implementation of a second-generation high-pressure test section and associated diagnostics, extending the capabilities of the previously presented FEDORA facility toward enhanced optical accessibility, advanced multiphysics measurements and improved operational flexibility. The upgraded test section was specifically designed for operation under thermal-hydraulic conditions relevant to multiple divertor concepts while enabling synchronized visualization and local measurements of boiling phenomena. The design incorporates replaceable visualization windows for high-speed optical observations, improved pressure and temperature measurement ports, an upgraded IR visualization system and a newly developed heater design. In addition, a micro-fibre optical probe has been integrated into the test section to enable local measurements of bubble size and velocity. Particular attention has been devoted to the electro-thermal design of the electrically heated section, where minimization of electrical contact resistance, mitigation of current crowding effects and suppression of localized hot spot formation were identified as critical challenges for reliable high heat flux operation. The three-dimensional geometry of the electrical contacts and current injection region was optimized using supporting numerical simulations to improve heat flux uniformity and reduce parasitic thermal non-uniformities within the heated surface.

The upgraded facility will provide detailed experimental data over a broad range of operating conditions for the development and validation of advanced theoretical and CFD boiling models relevant to fusion thermal management applications.

Accurate estimates of reactor internals and the pressure vessel neutron exposure are essential for ageing management and for interpreting surveillance and ex-core dosimetry in pressurized water reactors. This work presents a comprehensive development of cycle-dependent models for assessing cumulative neutronic effects over several Krško Nuclear Power Plant (NPP) cycles, as well as for validating several cycle-dependent ex-vessel neutron dosimetry (EVND) measurements.

Cycles 26–31 of Krško NPP were selected, since they were covered with the second EVND campaign. For each of these cycles, beginning-of-cycle (BOC) and end-of-cycle (EOC) core state MCNP models were obtained using an automated McCord subroutine (AutoMcCord). The methodology is based on CORD-2 deterministic nuclear design calculations, where neutron transport equation is coupled with the thermohydraulic feedback. These data serve as input for the determination of the MCNP generated source term. Self-consistency between power and temperatures in the Monte Carlo calculations is achieved through a specially developed in-house iteration process. Fuel assembly power distributions from CORD-2, the initial MCNP models, and the iterated MCNP models will be compared by cycle, burnup state, and core quadrants. The core power axial offset is used as a convergence parameter, because it effectively captures the differences between the initial and iterated MCNP solutions.

In addition, between-cycle and cycle-to-cycle variations in neutron spectra and neutron flux in the ex-vessel region are analyzed. Neutron fluxes and spectra are compared at representative ex-vessel locations for all prepared core models, with particular emphasis on differences between BOC and EOC states of different cycles, as well as on variations across multiple burnup points within one representative cycle.

The outcome of the study is a consistent set of validated and converged MCNP generated source models for Krško NPP cycles 26–31, together with ananalysis of how fuel assembly and axial power profiles evolve through the iteration procedure. The work aims to provide reliable, state-of-the-art NEK MCNP ex-vessel models for neutron embrittlement, dosimetry, and material activation studies and analyses.

Large-break loss-of-coolant accident (LBLOCA) is a design basis accident during which the core would completely uncover in the existing light water reactors. The deterministic safety analysis must demonstrate that acceptance criteria for emergency core cooling systems must be fulfilled, peak cladding temperature being the most known. The best estimate computer codes may be used for such analysis, provided that uncertainties are quantified. The main objective of the study was to evaluate the applicability of the TRAC/RELAP Advanced Computational Engine (TRACE) thermal-hydraulic computer code for an uncertainty and sensitivity analyses of a double-ended cold leg guillotine LBLOCA in a two-loop pressurized water reactor (PWR). For calculations the TRACE input deck, which was in the past first converted from verified and validated RELAP5 input deck of two-loop PWR, and then manually corrected, was used. It was assumed operation of one train of safety systems. For selecting the base LBLOCA scenario, the discharge coefficient of the breaks was varied between 0.2 and 1.0 by step 0.1. At the end the scenario with nominal discharge coefficient 1.0, giving one of the highest values of peak cladding temperature, was selected for uncertainty and sensitivity analysis.

For the uncertainty and sensitivity analysis, Symbolic Nuclear Analysis Package (SNAP) software and the uncertainty plugin using DAKOTA (Design Analysis Kit for Optimization and Terascale Applications) were used. Sixteen input uncertain parameters influencing the peak cladding temperature were selected based on the literature. These parameters were randomly sampled using DAKOTA in each of the 130 runs, performed by TRACE. Second DAKOTA run was performed using the variate and extracted figure of merit (FOM) values to obtain the statistical results and cumulative distribution functions for FOM. DAKOTA also calculates the response correlations for FOM. These indicate how FOM correlates to each model variable.

The results showed that the peak cladding temperature was obtained in the blowdown phase in the first few seconds, before the start of safety injection pumps. The sensitivity analysis showed that the most influential input uncertain parameters were initial reactor power, power peaking factor and decay heat.

The present study investigates the dielectric relaxation and charge-transport behavior of regenerated silk fibroin composites modified with pristine and neutron-irradiated TiO₂ nanoparticles. Silk fibroin is a promising biopolymer platform for flexible electronics, bio-dielectrics, and sensing technologies because of its film-forming ability, mechanical flexibility, environmental compatibility, and processability from aqueous solutions. However, pure fibroin is generally characterized by high electrical resistance and broad non-Debye relaxation behavior, which limits its direct application in functional dielectric and frequency-selective devices. To overcome this limitation, TiO₂ nanoparticles were incorporated into the fibroin matrix at low concentrations of 0.1, 0.5, and 1.0 wt%. In parallel, neutron irradiation was applied as a post-synthesis defect-engineering method for TiO₂ nanoparticles in order to evaluate whether radiation-induced defect states could provide additional control over the impedance response of the resulting composites.

Broadband impedance spectroscopy was performed over the frequency range of 20 Hz–10 MHz and the temperature interval of 300–420 K. The impedance spectra of pure fibroin revealed thermally activated non-Debye relaxation, expressed by frequency-dependent Z′ and Z″ dispersion, temperature-shifted loss peaks, and depressed Nyquist semicircles. These features indicate a distribution of relaxation times associated with segmental motion in the fibroin matrix, interfacial polarization, and charge accumulation effects. Upon incorporation of pristine TiO₂ nanoparticles, the real part of impedance decreased systematically, while the characteristic dispersion regions and Z″ loss peaks shifted toward higher frequencies. This behavior confirms that TiO₂ nanofillers enhance charge mobility, reduce effective interfacial barriers, and promote faster polarization dynamics within the fibroin matrix. The effect of TiO₂ loading was not strictly monotonic. The most efficient acceleration of relaxation was observed near 0.5 wt% TiO₂, suggesting improved filler–matrix interaction and more effective interfacial charge equilibration at this concentration. At 1.0 wt%, a partial reversal or broadening of the relaxation response was observed, which can be attributed to stronger Maxwell–Wagner–Sillars polarization, increased interfacial heterogeneity, and possible overlap of relaxation processes at higher filler content. These results demonstrate that even small amounts of TiO₂ nanoparticles can significantly modify the dielectric response of fibroin, but that the final behavior depends on a balance between improved transport pathways and enhanced interfacial polarization. A more pronounced modification was observed when neutron-irradiated TiO₂ nanoparticles were used as fillers. At identical TiO₂ concentrations, composites containing irradiated nanoparticles exhibited lower impedance, further right-shifted Z″ peaks, reduced Nyquist arc diameters, and generally shorter Cole–Cole relaxation times compared with composites containing pristine TiO₂. The strongest effect was observed at 1.0 wt%, where the relaxation time was reduced by approximately two times near 420 K. This improvement is attributed to irradiation-induced defect channels in TiO₂, including neutron-transmutation-generated vanadium dopants and Ti³⁺/oxygen-vacancy centers. These defects can increase carrier density, facilitate small-polaron hopping, reduce activation barriers, and lower interfacial resistance at the fibroin–TiO₂ boundary.

The impedance spectra were successfully interpreted using Cole–Cole relaxation analysis and an equivalent circuit model consisting of bulk and interfacial branches represented by resistance–constant phase element combinations. The depressed semicircles in the Nyquist plots confirm the non-Debye nature of the relaxation processes and indicate a broad distribution of relaxation times caused by structural and electrical heterogeneity in the polymer–nanofiller system. The decrease of resistance with increasing temperature, TiO₂ loading, and neutron modification supports a thermally activated transport mechanism controlled by both matrix dynamics and filler-induced interfacial effects. Overall, the study demonstrates that neutron irradiation of TiO₂ nanofillers is an effective, solvent-free, and materials-efficient route for tuning the dielectric dispersion and electrical relaxation behavior of fibroin-based nanocomposites. Unlike conventional chemical doping or high-filler-loading approaches, this method enables functional modification at very low nanoparticle concentrations while preserving the biopolymer processing route. The obtained results provide a useful physical framework for designing flexible bio-dielectric materials with controlled impedance response, tunable relaxation frequency, and improved charge-transport properties. Such defect-engineered fibroin–TiO₂ composites may be promising for applications in flexible electronics, impedance-based sensors, biointegrated dielectric layers, and frequency-selective functional components.

As the rise of weather dependent energy sources is causing fluctuations in electricity generation, more and more nuclear power plants are expected to transition to load following operation, during which the power can be changed multiple times a day. This introduces new challenges from materials aging perspective and from core physics perspective, as the power changes introduce imbalances in fission product concentrations. The changes in strong absorbers, such as xenon and iodine can, if not managed, cause loss of sufficient reactivity or non compliance with technical specifications, which is why operators rely on simulations to plan such power changes.

The Loadf program was developed at the Jožef Stefan Institute to simulate transient operation in the Krško Nuclear Power Plant during load following operation. It uses the diffusion code GNOMER for neutron transport including a thermal hydraulics module and a dedicated routine to track the xenon and iodine concentrations. The code has been verified on common operational transients, but was historically not used for uncertainty propagation studies. Uncertainties in the microscopic cross sections can be propagated to the macroscopic cross sections with which the nominal core conditions are calculated and serve as a starting point for the off-nominal conditions during a transient. One of the challenges that needs to be solved in this respect to describe the full effect of the xenon feedback effects is the perturbation of the reactivity coefficient library, which is used in the calculations. This library contains information about boron reactivity worth, temperature feedback coefficients, xenon reactivity and equilibrium concentrations as well as iodine decay constant and Xenon and Iodine fission yields from $^{235}$U. In addition to LoadF, the reactivity coefficient library is also used by the CORD-2 core design suite, but there the reactivity coefficients are not used as for primary thermal hydraulic treatment but solely for the acceleration of convergence of critical boron concentration.

In this work we concern ourselves with the xenon reactivity coefficients. The \texttt{RHOXE} parameter is the xenon integral worth, expressed in pcm as a function of the relative xenon concentration. The \texttt{DENXE} parameter is the equilibrium xenon number density, given as a function of the relative power, and sets the reference concentration toward which the transient xenon relates. The \texttt{RHOXB} parameter is a boron correction applied to the xenon worth. All three quantities are tabulated at various stages of fuel burn-up but calculated at a nominal set of conditions.

We present a method to generate these coefficients consistently with perturbed nuclear data. The required inputs for the transport calculations are produced using WICORD, a CORD-2 utility that generates WIMS inputs. The transport code WIMSD-5B is then used to estimate the equilibrium xenon concentration and its reactivity worth under the given conditions. By running this procedure with perturbed WIMS-D libraries, a corresponding set of perturbed reactivity coefficients is obtained. The perturbed coefficients, used together with the corresponding WIMS-D libraries, allow the xenon feedback to be perturbed consistently in load-following uncertainty propagation simulations. Thus, uncertainties due to reactivity coefficients can be assigned to the parameters calculated within the LoadF transient simulations.

Tungsten (W) is the primary candidate for plasma-facing components in the DEMO divertor. However, its intrinsic brittleness, recrystallization-induced degradation, and pronounced grain growth at elevated temperatures limit its performance under fusion-relevant thermal loads. To overcome these drawbacks, W–W₂C composites were developed in which tungsten sub-carbide (W₂C) particles are formed in situ at W grain boundaries during Field Assisted Sintering (FAST) [1].

The in situ formed W₂C inclusions effectively pin grain boundaries, suppressing abnormal grain growth even at temperatures up to 2000 °C. At the same time, carbide formation promotes densification and eliminates oxygen from the starting powder without hydrogen treatment. Compared to pure W, isotropic W–W₂C composites show refined microstructures, enhanced flexural strength at room and elevated temperatures, and improved long-term thermal stability. For low W₂C contents (2–4 wt%), the ductile-to-brittle transition temperature (DBTT) ranges between 200–400 °C, comparable to or lower than conventional isotropic tungsten grades. Thermal ageing above 1250 °C for up to seven days confirms the stability of the microstructure due to grain boundary pinning. Deuterium retention measurements further demonstrate acceptable trapping levels under fusion-relevant conditions [2].

Building on these findings, isotropic FAST-sintered W–W₂C composites were subjected to controlled high-temperature plastic deformation (compression and/or forging) to produce anisotropic microstructures with tailored grain morphology. Phase composition (XRD, EBSD), microstructure evolution, and mechanical properties were systematically evaluated before and after additional thermal treatment at 1600 °C for 24 h. Attention was given to verifying whether W₂C inclusions effectively inhibit grain growth also in anisotropic architectures.

The combination of high thermal stability, improved mechanical performance, reasonable thermal conductivity (>100 W/mK at 1000 °C), and controlled hydrogen isotope retention makes W–W₂C composites promising candidate materials for DEMO divertor applications.

References

[1] S. Novak et al, Mater. Sci. Eng. A, 2020, 772, 138666.

[2] P. Jenuš et al., Nucl. Mater., 2023, 581, 0–6.

Acknowledgement: This work has been carried out within the framework of the EUROfusion Consortium, funded by the European Union via the Euratom Research and Training Programme (Grant Agreement No 101052200 — EUROfusion). Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the European Commission. Neither the European Union nor the European Commission can be held responsible for them. This project has received funding from the Slovenian Research Agency (P2-0087, P2-0405, and P2-0050).

The micro and small modular reactors (MMR and SMR) represent one of the most promising technologies for ensuring a reliable, sustainable and low-carbon energy supply. Research and development activity are crucial to study the impacts of the introduction of MMR and SMR on operational capabilities, environmental impacts and compliance with safety standards. The focus is also on adapting the regulatory framework for safe and efficient implementation of these advanced systems in today’s energy systems.

The proposed paper will focus on implementing MMR and SMR in Slovenia. The purpose is to conduct an analysis of the Slovenian regulatory framework for ensuring radiation and nuclear safety that are currently in use. The analysis of the technological neutrality of the identified legislation from the perspective of the implementation and use of SMR/MMR will aim to find gaps, needed to be addressed from the perspective of the SMR and MMR introduction. The second aim is to identify areas where increased research and development efforts are needed in support of the introduction of SMRs and MMRs. A view of any necessary adjustments of regulatory standards towards the practical use of MMR and SMR will be included.

Since its commissioning in 1966, the 250 kW TRIGA research reactor at the Jožef Stefan Institute (JSI) has been operated with a strong commitment to safety and operational efficiency. For more than a decade, Safety Performance Indicators (SPIs) have been routinely monitored as an integral part of reactor oversight. These indicators encompass reactor operating hours, the number of samples irradiated, radiation exposure to operating personnel, and the quantity of radioactive gases discharged to the environment. This paper reviews the SPIs collected in 2025 and discusses their significance in supporting the reactor’s continued safe, dependable, and effective long-term operation.

In the field of research, we have continued most of the established research campaigns from previous years. Irradiation of various samples for CERN continued and more measurements were performed on the water activation loop setup (KATANA). We also continued the collaboration with the Faculty of Chemistry and Chemical Technology in the field of BNCT (Boron Neutron Capture Therapy). CEA researchers conducted several experiments, including the use of the new STARSHIP experimental device.

The STARSHIP device is a high-temperature irradiation system based on a tubular furnace design. It is designed to enable controlled irradiation experiments under elevated temperatures. The primary purpose of the STARSHIP high-temperature irradiation device is the irradiation of material samples at elevated temperatures. It is also used for testing neutron and gamma-ray detectors under controlled conditions. The system can achieve temperatures up to 900 °C and is intended to replicate environments relevant to Generation IV high-temperature reactors and fusion devices.

As part of the EUROLABS project, the internal components of the Tangential Channel of the TRIGA reactor were refurbished and a specially designed cadmium shield was installed. The introduction of a cadmium shield represents an important upgrade. Its main function is to modify the neutron field so that the flux of fast neutrons remains essentially unchanged, while the flux of thermal neutrons is significantly reduced. This helps to lower the activation of irradiated materials and simplifies the handling of samples after irradiation.

Within the framework of educational activities, we carried out numerous exercises performed by the students from the following universities: University of Ljubljana, Aix Marseille University, Politecnico di Milano, Chalmers University and King Fahd University of Petroleum, as well as doctoral students from the United Kingdom participating in a CDT (Centre for Doctoral Training) programme. We also conducted a Nuclear technology course for NEK and GEN Energija personnel. In March, an Open Day event was organised to educate the general public about TRIGA reactor operations, attracting approximately 800 visitors.

Computational Fluid Dynamics (CFD) simulations play an important role in predicting and optimizing the performance of fusion reactor divertors. However, their reliability depends on accurate experimental validation, especially under the extreme thermal-hydraulic conditions encountered in such systems. The FEDORA (Fusion Experiment for Divertor Optimization Research Applications) experiment is a flow boiling visualization experiment developed as part of the THELMA laboratory at JSI-R4 to investigate boiling behaviour under operating conditions representative of fusion applications. Monitoring of two-phase flow behaviour inside the test section is performed non-intrusively using a transparent window and high-speed imaging. Captured images are post-processed to extract detections, tracking and velocimetry calculations of individual bubbles.

The core part of this methodology is a newly developed bubble tracking velocimetry framework based on the Tryolabs Norfair library. The method processes sequences of recorded images with detected bubbles, including their positions, produced by the StarDist AI model, to generate individual bubble velocity data. As a result, distributions of bubble velocities and turbulent kinetic energy can be determined, as well as embedding the velocity information into the video recordings themselves. Bubble matching between consecutive frames is based on a combination of a custom distance function and the Kalman filter used for future position estimation. The algorithm is capable of identifying newly formed bubbles and re-identifying those that have been occluded for several frames. Using information about bubble radius and tracking duration, an additional filter was developed to remove falsely identified bubbles. The visualizations generated by the algorithm annotate each bubble with an ID and color-code the edges, enabling users to assess the quality of the data pipeline by inspecting the post-processed video recordings.

The output of the bubble tracking velocimetry framework will help us to extract additional bubble parameters in two-phase flows, specifically useful for validation of CFD simulations.

Experimental work in physics education is frequently constrained by equipment deficits, financial limitations, and restricted access to modern pedagogical tools. These resource constraints often deter educators from introducing novel experiments beyond the standard curriculum. This paper presents a cost-effective, inherently safe methodology for constructing a homemade ionizing radiation detector. The primary objective is to facilitate experimental laboratory work in primary and secondary education, thereby rendering nuclear physics more accessible to both teachers and students.

The operational mechanism of the detector relies on the ionization of air between two high-voltage metal electrodes. The passage of an ionizing particle induces localized air ionization, resulting in an observable increase in spark discharge and a characteristic acoustic response (crackling sound). This paper details the structural components, the fabrication protocol, and specific device modifications engineered to detect ambient ionizing radiation.

Furthermore, a pedagogical demonstration is presented utilizing everyday items as radiation sources. This approach eliminates the necessity of purchasing commercial radioactive sources, reducing institutional expenditures and minimizing administrative and regulatory requirements. Such demonstrations offer students a direct, empirical understanding of the physical phenomena associated with ionizing radiation, fostering a deeper integration of experimental practices in nuclear physics curricula.

The detector is currently undergoing optimization and testing. A primary technical challenge involves mitigating the effects of ambient atmospheric humidity on the reproducibility of experimental results. Nevertheless, preliminary data indicate that the device holds significant potential for implementation in classroom demonstrations as well as for student-led investigative and research activities.

Accurate prediction of boiling heat transfer in actively cooled fusion divertor targets remains a major challenge for computational fluid dynamics simulations. These components operate under extreme thermal-hydraulic conditions, with heat fluxes of the order of 10 MW/m², high liquid subcooling, and coolant velocities approaching 10 m/s. Under such conditions, conventional Eulerian two-fluid simulations based on the classical Rensselaer Polytechnic Institute wall heat-flux partitioning model can produce non-physical wall-temperature behaviour.

In this work, an MIT-based wall-boiling model is analysed under fusion-relevant conditions. The model is built from mechanistic closure relations describing observable boiling phenomena, such as bubble departure diameter, nucleation site density, departure frequency, bubble growth and waiting times, and the surface area influenced by sliding bubbles. These closures determine the partitioning of the imposed wall heat flux between liquid heating, evaporation, and bubble-induced sliding conduction. Because they are linked to measurable physical quantities, they also provide a direct connection between CFD model development and dedicated validation experiments, such as the FEDORA facility.

The results are used to identify which parts of the MIT wall-boiling framework control the predicted thermal response under highly subcooled, high-velocity conditions. This provides guidance for future model development and helps define which boiling quantities should be prioritized for direct experimental assessment.

In today’s energy sector, where nuclear power plays a key role in the transition to a low-carbon society, sustainable material management—covering both raw materials and waste—is becoming vital for its long-term public acceptance. The nuclear industry uses both circular and linear waste management models. However, the public is mostly familiar with the linear model. In this approach, materials used in nuclear facilities are treated strictly as radioactive waste and sent directly for final disposal after being processed and prepared. Instead of trying to get the most value out of these materials, the linear model sees them as an environmental burden, ignoring their potential for recycling or reuse. In contrast, the circular economy focuses on keeping resources in use for as long as possible, recovering and regenerating products and materials at the end of each life cycle. The goal is to maximize value while minimizing waste and conserving natural resources. This circular concept isn’t limited only to the nuclear fuel cycle—where fuel recycling is already commercially used in several countries—but also includes low- and intermediate-level materials. With the right management, these could be reused or recycled instead of being permanently disposed of as waste.

The principles of the circular economy in the nuclear industry are based on the definition provided by the International Atomic Energy Agency (IAEA), which characterizes this Circular Economy Model as an approach where resources are kept in use for as long as possible through reuse and recycling. In line with this approach, it is crucial that end-of-life components are preferentially reused if they remain technically suitable. When reuse is not feasible, recycling is implemented; in addition to minimizing waste disposal, this enables the preservation of material properties and maximizes their value in future cycles of use.

The very definition of radioactive waste, as provided by the International Atomic Energy Agency (IAEA) and summarized in Slovenian legislation within the Ionising Radiation Protection and Nuclear Safety Act (ZVISJV-1), implicitly encourages the implementation of circular economy principles. According to Slovenian nuclear legislation, radioactive waste is defined as radioactive substances resulting from radiation or nuclear activities for which no further use is foreseen. This definition highlights the need to consider whether any further use is possible for radioactive materials in the nuclear industry before they are classified as waste. Consequently, only radioactive substances that contain or are contaminated with radionuclides above clearance levels, and for which no further use is foreseen, are considered waste. Current radioactive waste (RAW) management practices already establish a foundation and present an opportunity for the future implementation of circular economy principles. Some of current practices in the nuclear industry are thus already aligned with circular economy principles—such as refurbishment and lifetime extension of existing power plants, reprocessing of spent fuel, and recycling of metals and other materials. Other examples include reusing components from permanently shut down reactors in operating facilities, as well as repurposing buildings instead of constructing new ones. Additionally, using material from demolished buildings as backfill during site remediation is in line with circular economy principles, provided that the buildings cannot be refurbished or repurposed.

Key opportunities for implementing reuse and recycling focus on:

- the reuse of infrastructure facilities and equipment,

- metal recycling,

- fuel reprocessing, and

- lifetime extension of nuclear facilities.

Globally, integrating the circular economy into the nuclear industry is already reducing final waste volumes and enabling more cost-effective material management. Beyond economic benefits, grounding decisions on scientific facts and demonstrable circularity strengthens public trust in nuclear energy’s role in sustainable development. However, implementing these principles remains challenging, as most legacy plants approaching decommissioning were not designed with material circularity in mind. Nevertheless, past experiences offer valuable insights for new reactor generations under construction, where circularity can be integrated during the design phase—either through modular configurations for easier dismantling or advanced materials that optimize decontamination. To achieve systemic progress, developing uniform guidelines for assessing circularity in the nuclear sector is essential, as such frameworks currently do not exist. Standardization would significantly facilitate implementation, ensure greater consistency in waste management, and enable better international comparability of national programs.

The paper will elaborate on how application of sustainable circular economy in radioactive waste management strategies is already integrating technical safety requirements based on developed IAEA standards with societal expectations, ethical considerations and governance standards. Some concrete examples of these principles in RAW management already applied in different European countries will be given.

The Krško Nuclear Power Plant (NEK) maintains high standards of safety and operational reliability, with the Professional Training Department playing a pivotal role in the systematic education and training of both employees and external contractors. In response to the increasing demand for modern, more practice-oriented training methodologies, the development of a new Theoretical and Practical Training Center (CTIPU) was initiated in 2025. The newly established training center facilitates enable the implementation of advanced simulation-based scenarios and hands-on training of NEK personnel and contract workers within specialized training environments. These settings enable participants to acquire practical experience under controlled and safe conditions. Such an approach is of fundamental importance in the nuclear industry, where high-quality training, closely aligned with real-world conditions, is essential for mitigating human error and reinforcing a robust safety culture. The aim of this article is to present the conceptual design and functional framework of the CTIPU training center, with particular emphasis on specialized training facilities for safe work at height, in depth, within confined spaces, as well as for load lifting and rigging and work on electrical equipment. Furthermore, the article seeks to assess the role and significance of practical training and simulation-based learning environments in ensuring a high level of safety performance and preparedness for complex and demanding working conditions in the nuclear power plant.

Keywords: Nuclear power plant, education, CTIPU, testing grounds, safety culture

Monte Carlo neutron transport has become the most widely used method for the computational
analysis of reactor and shielding problems in the nuclear fields. It offers high level of accuracy
as it enables the treatment of complex geometries and the continuous-energy neutron spectrum,
with the obvious downside of remaining computationally demanding. Monte Carlo simulations
are still commonly performed on CPU-based systems, while recent GPU-oriented development in
Monte Carlo neutron transport code OpenMC has introduced an experimental event-based trans-
port algorithm where parts of the computation are executed on the GPU. So far the OpenMC
GPU performance studies have been mostly done for high performance cluster (HPC) class GPU
systems, whereas many practical users operate smaller institutional GPU servers, workstations or
individual multi-GPU nodes for which the performance has not yet been studied.

In this work, we evaluate the GPU-accelerated OpenMC implementation from the ExaSMR de-
velopment branch on a range of node-scale CPU and NVIDIA GPU systems. The objective is not
to reproduce HPC scaling results, but to assess what performance can realistically be expected
on more accessible hardware. The tested GPU platforms include data-center NVIDIA H200 NVL
accelerators, several generations of professional NVIDIA RTX GPUs (Ampere, Ada and Blackwell)
and older or consumer-oriented devices. CPU reference systems include dual-socket AMD EPYC
and Intel Xeon nodes, as well as high-core-count workstation-class processors. Where available,
both single-GPU and dual-GPU MPI configurations were tested.

Testing was conducted on multiple reactor criticality and shielding geometrical models to test if
and how much the performance is case dependent. The chosen models were the existing OpenMC
GPU problems, which includes the fresh and depleted PWR pin-cell and the Hoogenboom-Martin
reactor computational benchmarks. In addition, two application-oriented models were added. To
extend the assessment beyond simplified or development-oriented cases, we added a full experi-
mentally validated JSI TRIGA research reactor model and a fixed-source ITER-related shielding
benchmark. These models introduce different geometry, material, source and transport character-
istics, and therefore provide a more realistic test of the GPU transport implementation for reactor
physics and shielding workflows.

Performance was evaluated primarily using the steady-state particle tracking rate, expressed as
particles per second, rather than only reporting relative speed-up factors. This allows a more
transparent comparison between hardware platforms and benchmark cases. Where possible, total
wall-clock time was separated from transport-only timing, since smaller cases can be dominated
by initialization, nuclear data loading and statepoint writing rather than neutron transport itself.
In addition to raw throughput, approximate particles-per-watt and particles-per-cost metrics were
considered to provide a practical estimate of energy and economic efficiency. The results show
that modern NVIDIA GPUs can provide substantial acceleration for sufficiently large OpenMC
workloads, especially when the particle transport phase dominates the total runtime. However, the
performance benefit is not uniform across all benchmarks or hardware generations. Small problems
and short runs can show limited apparent speed-up when initialization overhead is included, while
larger reactor and shielding models better expose the computational advantage of GPU execution.

This study provides a practical node-scale benchmark of GPU-accelerated OpenMC across a
broader range of NVIDIA hardware than is usually reported in exascale-focused studies. The
results are intended to help users estimate whether GPU acceleration is beneficial for their own re-
actor physics or shielding workloads, and to identify the configuration details that materially affect
performance. The work also highlights that meaningful GPU benchmarking should report not only
speed-up factors, but also absolute particle rates, initialization overhead, memory requirements,
power consumption and cost-related metrics.

The backward-facing step (BFS) is a canonical configuration commonly used to study flow separation and reattachment phenomena. Non-intrusive wall-temperature measurements provide valuable insight into the spatial evolution of separated flows. In BFS flow, wall heat transfer is governed by boundary-layer separation, recirculation, shear-layer development, and downstream flow reattachment following the sudden expansion. This study employs high-speed infrared (IR) thermography together with a data-driven approach to analyse flow structures and estimate the location and dynamics of the reattachment region on a heated foil surface.

The investigation is based on experimental temperature measurements performed at the THELMA Laboratory of the Reactor Engineering Division at the Jožef Stefan Institute using a heated backward-facing step facility. High-resolution infrared images were acquired at 100 Hz for several Reynolds numbers. To capture the thermal footprint associated with turbulent separated flow and convective cooling dynamics, a heated foil was installed downstream of the step. The resulting wall-temperature distributions recorded by the IR system indirectly reflect the separated-flow recirculation zone, turbulent mixing, and thermal transport mechanisms.

The recorded IR images and corresponding spatial temperature-gradient fields are analysed using a framework based on convolutional neural networks (CNNs). Unlike conventional approaches relying on point-wise sensors, the proposed framework processes full-field thermal maps to correlate turbulence-induced thermal fluctuations with the reattachment region. A dedicated preprocessing pipeline, including foil-region extraction, dynamic-range normalization, spatial cropping, and thermal-image standardization, is applied to improve feature extraction and data consistency.

Preliminary results indicate that characteristic thermal signatures associated with enhanced convective heat transfer near the reattachment location are embedded in both the minimum-temperature regions and the spatial thermal patterns. In addition, unsupervised deep latent representation learning is being investigated to identify coherent thermal manifolds and persistent flow states directly from infrared thermography data.

Flexible operation of VVER-type nuclear power plants has become a major topic in recent years. With the increasing share of renewable energy sources in the energy mix of individual countries, the pressure to regulate the output of nuclear power plants is growing. VVER-type nuclear power plants were not originally designed for this type of operation; they were intended to operate at a stable power level. Regular power reductions based on grid demands may, in the future, lead to reliability issues of certain components and the need for their more frequent replacement.

Flexible power operation introduces several risks to operation of NPPs. These risks arise because flexible operation disrupts steady-state chemical equilibria, altering temperature, pressure, and flow conditions throughout the steam/water cycle, which affects the effectiveness of corrosion control and deposition prevention measures.

Among the potential risks are, for example, higher carryover of contaminants into the steam cycle, which may lead to reduced steam purity and deposit formation. There is also an increased risk of impurity hideout and condenser tube leaks, both of which can negatively affect cycle chemistry.

In addition, flexible operation may enhance corrosion-product transport and flow-accelerated corrosion (FAC), particularly due to changing thermodynamic conditions. It may also increase the risk of failures in steam path components and turbines.

Furthermore, frequent load changes can lead to a loss of proper chemical dosing, resulting in suboptimal water chemistry control. Additional issues include challenges in cycle chemistry monitoring due to fluctuating parameters, as well as the need for adaptive water treatment strategies customized to flexible operation.

As most VVER-type nuclear power plants have only recently begun implementing flexible operation, there is currently very limited operational data available on its impact on long-term performance. For this reason, a questionnaire was distributed to eight VVER 1000 and VVER 440 power plants across Europe to collect information on how flexible operation is approached and implemented in their respective countries. All available information and data have been thoroughly monitored and evaluated in detail, including records from previous units operation at stable power levels throughout their campaigns.

Despite not being originally designed for load-following regimes, most plants have already implemented flexible operation, at least during part of their fuel campaigns. Current operational experience suggests that, so far, no significant adverse impacts on system performance or material integrity have been clearly identified. While most plants have implemented monitoring practices and conducted preliminary impact assessments, the lack of long-term operational data remains a key limitation.

Although no immediate changes in maintenance strategies have been observed, operators generally expect an increase in inspection frequency and maintenance costs in the future. This highlights the need for continued monitoring, data collection, and optimization of chemistry control and operational strategies to ensure safe and reliable long-term operation of VVER units under flexible regimes.

Nuclear research reactors support a wide range of applications, including education, medical isotope production, neutron beam experiments, material irradiation, and detector testing. The suitability of a reactor for a given application depends on several factors: available personnel, funding, auxiliary facilities, and, most importantly, the capabilities of the reactor design. The VERONICA project aims to develop a versatile, multi-purpose research reactor, so it is essential that the proposed design accommodates as broad a range of applications as possible. One promising approach is a reconfigurable reflector, where changes in reflector material composition can tailor the neutron flux level, energy spectrum, and spatial flux homogeneity to the requirements of different applications.

In this study the physical parameters of a possible reconfigurable reflector are analysed by developing an OpenMC model of the conceptual VERONICA multi-purpose reactor core and its reflector. Several cases are analyzed, varying the whole reflector composition by adjusting the amount of graphite, beryllium, and heavy water. A parametric study is conducted across the full composition phase-space of these three materials, treating the reflector as a homogenised region. In addition, a conceptual design of having only section of the whole reflector programmable is studied. For each case, key neutronic parameters are evaluated: the effective multiplication factor (k-eff), neutron flux spectrum at various locations within the core and reflector, the spatial homogeneity of the neutron flux within annular sections of the reflector, and estimated isotope production rates for selected medical and industrial radionuclides.

The results are mapped over the phase space of possible reflector compositions, highlighting the compositions most suitable for the thermal, epithermal, and fast irradiation applications. This study provides an initial neutronic characterisation of the VERONICA reactor concept and demonstrates the potential of having a programmable reflector as a practical tool for neutron spectrum management in a versatile research facility.

Accurate thermal neutron scattering data are essential for reliable neutron transport and criticality calculations in thermal reactor systems, particularly for moderator materials such as zirconium hydride (ZrHx​), where hydrogen is bound in a crystalline lattice. The thermal scattering law, S(α,β), is derived from the phonon density of states and depends on the vibrational and crystallographic properties of the material. In zirconium hydride, the lattice structure varies with hydrogen concentration, with the δ- and ε-phases being especially relevant for reactor applications.

In this work, thermal neutron scattering data for zirconium hydride were generated and compared using three nuclear data processing approaches: NJOY through its LEAPR module, NCrystal, and CINEL. Identical phonon densities of states, obtained from ab initio lattice dynamics calculations, were used as input in all cases in order to isolate the effects of the processing methodology and the physical approximations implemented in each code. The generated thermal scattering data were produced in ENDF-6 format and subsequently converted to ACE format for use in Monte Carlo neutron transport calculations.

The comparison shows very good agreement between NJOY, NCrystal, and CINEL for the inelastic scattering component for both hydrogen and zirconium in the considered zirconium hydride phases. The main differences are observed in the elastic scattering component, where NCrystal and CINEL explicitly include coherent elastic scattering, while this contribution is not treated in the standard NJOY(LEAPR) approach. Despite these differences at the cross-section level, their impact on selected ICSBEP criticality benchmark calculations was found to be limited. The benchmark response is dominated by hydrogen scattering, while the contribution associated with zirconium was small for the systems considered. Benchmark calculations performed with NCrystal and CINEL showed nearly identical results, indicating a high level of consistency between these two approaches.

In addition to the comparison of thermal scattering models, the influence of the ACE processing route was investigated by converting identical ENDF-6 thermal scattering data using both NJOY(ACER) and ACEMAKER. While the processed cross sections were generally consistent, non-negligible differences were observed in selected benchmark reactivity results. Further analysis showed that these discrepancies originate mainly from differences in the incident neutron energy grid and from the interpolation of secondary energy distributions in the ACE-formatted data. When the same incident energy grid was used in both processing systems, the benchmark results were reproduced very closely, confirming that the observed differences are numerical in origin rather than due to the underlying physical scattering model.

The results demonstrate that NJOY, NCrystal, and CINEL provide mutually consistent thermal scattering data for zirconium hydride when the same phonon input is used, with differences mainly confined to the coherent elastic channel. For the investigated criticality benchmarks, these differences have a limited effect on integral results. However, the study also highlights that the ACE processing stage can introduce significant variations in Monte Carlo calculations if the incident energy grid and interpolation of secondary energy distributions are not represented with sufficient accuracy. This emphasizes the importance of both physically complete thermal scattering models and robust nuclear data processing procedures for reliable reactor physics simulations.

Thermal fatigue is an important degradation mechanism in many engineering systems and components, including nuclear power plant piping systems. The mechanisms driving cyclic thermal behaviour depend strongly on the geometry and flow conditions of the system. Thermal cycling may occur, for example, in tee junctions where fluids at different temperatures mix, as well as in externally cooled dead-end pipes that are thermohydraulically connected to the hot flow in the main branch. The latter configuration has proven particularly challenging for computational fluid dynamics (CFD) simulations.

In 2023, a leak occurred at the butt weld of the safety injection line SI-53 at the Nuclear Power Plant Krško. Investigations identified thermal cycling as the primary cause of the failure. Previous analyses, including CFD predictions, have provided useful insight into the flow behavior of the dead-leg.

The presented work aims to perform a preliminary investigation of the SI-53 leak using OpenFOAM in a high-performance computing (HPC) environment. Due to its open-source and expandable nature, it is a desirable program for future developments using this geometry, as the system has shown to be sensitive to a wide array of parameters. The goal is therefore to perform both steady state and transient simulations, to assess whether the Reynolds-averaged Navier-Stokes (RANS) simulations in OpenFOAM yield meaningful results, qualitatively comparable to results from ANSYS Fluent commercial code.

Existing results indicate the presence of two dominant flow structures in the dead leg; near the tee junction the flow develops a pronounced turbulent swirl, whereas the long, uninsulated section of the pipe develops a natural convection loop. The analysis presented here is focusing on the flow characteristics of the swirl, the interaction region between it and the natural convection loop, as well as the temperature on specific points outside the dead-end pipe wall. Numerical results are additionally compared with external wall temperature measurements provided by NPP Krško.

Components of nuclear power plants, especially the piping, can be subjected to different types of cyclic loads, the amplitude and frequency of which can vary significantly. These loads can initiate and propagate cracks, which can lead to a compromised structural integrity and thus jeopardize the safe operating conditions of the nuclear power plant. For this reason, modelling and prediction of fatigue crack growth is necessary for a better understanding of this phenomenon, and the phase-field method (PFM) is chosen for this endeavour. By regularizing the Griffith theory of brittle fracture, PFM smears the discrete crack into a continuous one and accumulation of fatigue damage through time is done with a fatigue degradation function. Variable amplitude loads (VALs) are a complex type of loading, which may vary from the typical constant amplitude load by a single cycle of higher amplitude called an overload, up to a spectrum-type with random-like changes of amplitudes and frequencies.

In this work, an elasto-plastic material response with a combined isotropic and kinematic cyclic hardening has been implemented in the PFM. Additionally, a strain energy decomposition of the elastic strain energy density is developed to remove damage accumulation during compressive strains. The effect on crack growth of overloads during constant amplitude loading and a series of blocks of cycles with different amplitudes at a constant frequency are studied. The simulation results incorporate comparisons of Paris law behaviours, crack length versus number of cycles, and stress and strain distributions near the crack of the considered fracture mechanics (test) specimen.

The European research reactor fleet is aging and contracting. The average age of the European research reactor population is approximately 60 years, and the number of operational facilities is well below the level needed to support qualification of fuels and materials for power plant life extension, deployment of small modular and Generation IV reactors, fusion device development, and stable radioisotope supply. Closing this gap is now a strategic priority at national, European and international levels.

The VERONICA project (Versatile European Reactor for Neutron Irradiation and Nuclear Research), led by the Jožef Stefan Institute (JSI, Slovenia) in bilateral cooperation with the French Alternative Energies and Atomic Energy Commission (CEA), proposes a new versatile European research reactor facility located in Slovenia. The reference concept hosts two independent reactors within a single facility: a Zero Power Reactor (ZPR) and a Multi-Purpose Research Reactor (MPRR) of 5–20 MW thermal power, together with hot cells, post-irradiation examination (PIE) capability and flexible diagnostic laboratories. The two reactors are designed to serve fuel and material testing, neutron transport and multiphysics benchmark experiments, neutron beam experiments (scattering and imaging), neutron transmutation doping of silicon, support to SMR/AMR development, fusion neutronics R&D, medical and industrial radioisotope production, and training of nuclear professionals.

This paper documents how VERONICA addresses national, European and international strategic objectives.

At the Slovenian level, the project is in line with several national strategy documents. The Research Infrastructure Development Plan 2030 (NRRI 2030) identifies the Reactor Infrastructure Centre and the need to upgrade or replace the TRIGA Mark II reactor as a precondition for further work on low-carbon energy technologies. The Integrated National Energy and Climate Plan (NEPN 2024, p. 202) sets out concrete operational steps for modernising the national research nuclear infrastructure, including technology assessment, siting, design, construction and commissioning of a new research reactor. The Resolution on the Long-Term Peaceful Use of Nuclear Energy in Slovenia (Official Gazette RS, No. 39/2024) calls for verification of the technical and economic feasibility of new research reactor builds and emphasises R&D on advanced, small modular and research reactors. The Resolution on Nuclear and Radiation Safety 2024–2033 (ReJSV24–33, Official Gazette RS, No. 122/2023), under objective 12, requires stable conditions for nuclear research, education and training. The R&D Strategy for the Safe Use of Nuclear Energy and Other Sources of Ionising Radiation 2025–2035 (Government of Slovenia, 25 November 2025) operationalises this commitment, listing JEK2, SMRs and advanced reactor technologies, decommissioning, radioactive waste management, and medical and industrial radiation applications as priority R&D areas, alongside the strengthening of research infrastructure and competence.

At the European level, VERONICA aligns with the Nuclear Illustrative Programme PINC (2025), which recognises new research reactor infrastructure as a strategic requirement for technological sovereignty, plant safety and long-term workforce competence. The Draghi report on European competitiveness (2024) identifies nuclear energy as a key technology for decarbonisation and competitiveness. The European Parliament briefing on Nuclear Investment Needs (2024) estimates around EUR 241 billion in nuclear capacity by 2050, with research and testing infrastructures listed as enabling assets. The SNETP European Strategy on Research and Technology Infrastructures (2025) classifies such facilities as strategic for scientific excellence, industrial competitiveness and the energy transition. VERONICA is positioned as a horizontal infrastructure enabler within the IPCEI (Important Projects of Common European Interest) on nuclear technologies, supporting SMRs/AMRs, radioisotopes, fusion R&D and fuel cycle work, in line with the EU Green Deal, Net-Zero Industry Act and Euratom objectives.

At the international level, VERONICA addresses needs identified by the OECD NEA in its assessment Research and Test Facilities Required in Nuclear Science and Technology (2009), which highlighted the need for new capacity for fuel and material qualification and support to Generation IV systems. The IAEA framework for research reactor strategic planning and the IAEA Rays of Hope initiative for medical applications are addressed through VERONICA’s neutron irradiation services, training role and radioisotope production capability. VERONICA also contributes to UN SDGs 3 (Health), 4 (Education), 5 (Gender Equality), 7 (Affordable and Clean Energy), 9 (Industry, Innovation and Infrastructure), and 17 (Partnerships). The H2020-Euratom TOURR project (Towards Optimised Use of Research Reactors in Europe, grant 945269) documents the decline of the European research reactor fleet and provides the coordination context that VERONICA answers.

The JSI-CEA cooperation provides a complementary base: CEA brings operational experience from over thirty research reactors and the Jules Horowitz Reactor (JHR) under construction. The CEA’s involvement is justified by the prospect of additional irradiation facilities in the near future, which will enable the effective pursuit of high-impact R&D activities. JSI brings sixty years of TRIGA Mark II operation and a user base in fuel and material research, neutron metrology, radioisotope work and training. The conceptual design is complete, a candidate site within the Jožef Stefan Institute Reactor Centre identified, and the feasibility study is scheduled for Q3 2026. Cost-benefit analysis and business model development are carried out with the Institute for Economic Research (IER, Slovenia).

VERONICA is therefore not a purely Slovenian project. It is a coherent response to a structural gap affecting the European nuclear ecosystem, aligned with national, EU and international strategic priorities.

As demands for safe, long-term, and flexible operation of nuclear power plants (NPP) continue to grow, enhancing monitoring processes and modernizing systems have become a priority. One of the main pillars of the assessment of secondary circuit water chemistry have been Hideout Return (HOR) calculations at many NPPs.

Regular HOR evaluation and trend analysis across previous campaigns can detect anomalies or deterioration of the chemistry regime. Impurities originating, for example, from the main condenser leakages or water treatment systems can deposit and concentrate within steam generator crevices. They may be subsequently detected more efficiently in HOR samples than in feedwater or blowdown. These assessments support targeted inspections and enhance secondary water chemistry control.

Soluble species, that are present in bulk water of SGs can be concentrated in restricted geometries (tube to tube sheet or tube to support crevice). During normal operation, restricted flow and local boiling inside crevices can cause salts and corrosion products concentrate beyond their solubility limits and subsequently cause localized corrosion. The tube concentrated solutions can also form solid deposits and significantly change the local crevice chemistry. HOR describes the release of these accumulated impurities back into the bulk water. The process starts during power reduction below about 50% power and continues during shutdown. Experimental data indicate that this phenomenon begins at power levels of approximately 70 %.

The concentration process in these tight geometries is limited by thermodynamic conditions. In crevices with restricted flow, the solution concentrates until its boiling point matches the local tube wall temperature. At this point, the cavity becomes filled with a highly concentrated liquid. Over time, the crevice may contain a concentrated solution and solid precipitates formed when solubility limits were exceeded. The process is self-limiting, and concentration factors are usually around 10⁴, as shown in studies. Overall long-term condition, operation and SG cleanliness are significant factors that affect the overall life span of a SG. Appropriate blowdown and sludge removal during campaign—not only during shutdown—further help limit impurity concentration by enhancing crevice flushing.

During unit shutdown, the accumulated impurities begin to return to the bulk water. As temperature decreases and boiling stops, water enters the crevices, dilutes the concentrated solutions, and carries impurities into the blowdown. By analysing these impurities together with shutdown parameters, the high temperature crevice pH(t) can be calculated. A near neutral pH(t) is generally optimal for steam generator operation. The key species monitored for HOR analysis include Na⁺, K⁺, Ca²⁺, Mg²⁺, Cl⁻, SO₄²⁻ and SiO₂, with optional evaluation of NO₃⁻, F⁻, Fe and TOC. During HOR, fast moving monovalent ions (e.g. Na⁺, Cl⁻) wash out first because they are weakly bound, more soluble and enter the bulk water much sooner than other species. Multivalent ions such as Ca²⁺ and SO₄²⁻ return more slowly, as their solubility increases later in the cooldown and they remain trapped in deposits until gas bubbles fully disappear.

HOR assessment provides insight into the concentration of impurities within crevices and helps identify conditions that may lead to local corrosion, tube degradation or deposit formation. Early detection of abnormal hideout behaviour supports targeted inspections and operational measures such as controlled blowdown or cooldown procedures. These actions help reduce the risk of costly repairs and unplanned outages. Consequently, they mitigate corrosion risks and contribute to extending the service life of components. Evaluating the secondary side chemical regime and its operational history is essential for understanding the crevice chemistry environment and for an effective chemistry control.

The Divertor Tokamak Test (DTT) facility aims to investigate innovative approaches for managing the power exhaust challenges associated with future fusion power plants. While several divertor concepts will be explored under conditions relevant to DEMO, the first divertor concept follows an ITER-like design with the plasma facing units (PFUs) composed of W monoblocks attached to a CuCrZr pipe and XM-19 supports via a Cu interlayer and brazing, respectively. The PFUs are designed to accommodate strike points at varying positions, corresponding to the different equilibria of candidate plasma scenarios. The cassette assembly is composed of 9 PFUs that are mounted in series onto 3 stainless steel backplates, which act as the interface with the cassette frame. Three distinct target regions are identified as the inner-vertical, the central-dome and the outer targets, the latter consisting of an outer-vertical target and an outer-horizontal target.

Previous studies on the structural integrity of PFUs under both normal and off-normal operating conditions identified electromagnetic (EM) loads as the source of higher (local) stresses in the CuCrZr pipe and supports, and highlighted specific PFU regions where additional supports could improve the preliminary design. As the DTT divertor PFU design is further developed, additional analyses of the full assembly with updated component geometries subjected to the anticipated EM loads are required to confirm its structural adequacy. With this objective, a finite element (FE) model of the latest divertor PFU assembly design has been developed. Connections between PFUs and backplates are represented through realistic interactions that replicate the mechanical behaviour of hinge and hinge/slot supports via kinematic couplings. Similarly, the interactions at the backplate/cassette interface are modelled through kinematic couplings and contact conditions representing the mounting screws and cassette frame, respectively. The EM loads arising from a slow unmitigated downward vertical displacement event (DVDE) with a 40 ms current quench time are applied in the structural analysis using a discrete vector field as input in the ABAQUS computer code. The results provide the overall displacements, identify the most stressed regions and provide a preliminary verification against the codified rules of the Structural Design Criteria for In-vessel Components (SDC-IC).

The preparation of the financing model for the JEK2 project focused on identifying an economically sustainable, financially resilient and institutionally feasible framework for the implementation of one of the largest strategic infrastructure and energy investments in the Republic of Slovenia. Due to the capital-intensive nature of nuclear projects, long investment cycles and the strategic importance of JEK2 for long-term energy security and decarbonisation, the analysis concentrated on financing structures capable of ensuring low capital costs, stable long-term cash flows and an appropriate allocation of risks between the state, the project company and potential investors.

A central organisational element of the proposed structure is the establishment of a dedicated Special Purpose Vehicle (SPV), which would serve as the legal and financial entity responsible for all permits, contractual arrangements, financing flows, construction activities and future operation of JEK2. Such a structure enables a clear separation of the project from other activities within the GEN Group, improves transparency of financial flows and project costs, facilitates risk management and supports compliance with European Union state aid rules.

Several financing models were analysed, ranging from fully state-based financing to fully private financing structures. The analysed options included direct state financing through on-lending to the SPV, a hybrid model combining on-lending with selective direct SPV financing, SPV bond issuance backed by state guarantees, and fully private financing. The analysis demonstrated that a purely private financing model would be difficult to implement in practice due to the high cost of capital, elevated investor return requirements and full exposure to electricity market risks. As a result, the hybrid financing structure referred to as Model 1a emerged as the most balanced and economically viable solution. Model 1a represents a hybrid sovereign financing structure combining state on-lending with the possibility of selective direct financing by the SPV

Under Model 1a, the majority of financing would be secured through sovereign borrowing by the Republic of Slovenia, with the proceeds transferred to the SPV in the form of on-lending. This structure allows the project to benefit from Slovenia’s sovereign credit rating and established access to international capital markets, thereby significantly reducing borrowing costs compared to stand-alone project financing. At the same time, the model preserves a degree of flexibility by allowing the SPV, under predefined conditions, to raise financing directly from the market in later stages of the project when construction risks are lower and the SPV has established a credible financing history.

The cost of financing represents one of the key determinants of the project’s long-term economic viability. For investment modelling and financial projections, a conservative nominal cost of debt between 4 % and 4.5 % was assumed. Under the reference scenario for a 1,100 MWe nuclear unit, cumulative investment costs are estimated at approximately EUR 14.5–14.8 billion in current prices and EUR 11.5–11.8 billion in constant prices. Under the larger 1,650 MWe scenario, cumulative investment costs are estimated at approximately EUR 21.3–21.8 billion in current prices and EUR 16.9–17.3 billion in constant prices. These estimates include interest during construction and a long-term inflation rate of 2 %.

The comparative economic analysis confirmed that financing structures with a higher share of debt financing and lower financing costs result in significantly lower overall electricity generation costs. Conversely, models relying more heavily on private equity financing substantially increase the weighted average cost of capital due to the higher return expectations of private investors. Consequently, optimisation of the debt-to-equity structure and efficient management of financing costs during the construction phase are essential for ensuring the competitiveness of electricity generated by JEK2.

An important component of the proposed business and financing framework is revenue stabilisation. Given the long operational lifetime of nuclear facilities, high upfront capital expenditures and exposure to volatile electricity market prices, the project assumes the implementation of long-term revenue stabilisation mechanisms, particularly a two-way Contract for Difference (CfD). Such mechanisms provide predictable revenues by compensating the project when market prices fall below the agreed strike price, while requiring the return of excess revenues when market prices exceed the agreed threshold. Additional instruments, such as long-term Power Purchase Agreements (PPAs), may further support revenue stability and improve bankability.

From a public finance perspective, the accounting and statistical treatment of the financing structure under ESA 2010 rules will be of critical importance. The final classification of the SPV and the treatment of on-lending arrangements will determine the impact of the project on general government debt and deficit indicators. Therefore, early coordination with the European Commission and Eurostat will be necessary to ensure compliance with EU fiscal and state aid rules.

Overall, the analysis concludes that the hybrid sovereign financing model with on-lending to the SPV and the possibility of selective direct SPV financing represents the most suitable solution for the implementation of JEK2. The model combines the advantages of sovereign financing in the highest-risk phases of the project with the flexibility of gradual market participation by the SPV. It supports lower financing costs, enables phased investment dynamics, preserves strategic state control, facilitates the use of ESG and green financing instruments and establishes a financially sustainable framework for the JEK2 project.

This paper provides a historical overview of start‑up measurements performed by the Jožef Stefan Institute (JSI) team at the Krško Nuclear Power Plant (NEK). It reviews the evolution of measurement activities, with emphasis on the development of software, hardware, measurement instrumentation, computing platforms, data acquisition, and data processing methods employed over time. For shutdown and control bank worth measurements, a new "Rod Insertion Method" was developed at JSI. NEK was the first commercial plant where this method was successfully applied, and it is based on the functional capability of the program DMR043 – also developed at JSI.

The paper also documents the organizational aspects of the work, including the roles of JSI teams, collaboration with plant staff, and the transfer of knowledge across generations of scientists and engineers. The quality assurance program applied to the measurements is also described.

A short analysis of key measurement results is presented, focusing on their agreement with design predictions.

Finally, the paper provides references to the complete set of relevant reports and supporting documentation generated throughout the history of the measurement campaigns.

The study of natural convection can provide more detailed knowledge of a passive, reliable mode of heat transport, which is commonly used also in small and micro modular reactor designs. Relying on passive systems for cooling the reactor core, based on natural convection, reduces the need for powered systems. This can simplify systems, including safety systems, and decrease the failure risk tied to active components in a nuclear power plant. However, the efficiency of such systems has to be demonstrated, and computational fluid dynamics simulations can provide an insightful view of the phenomenon. Detailed modelling of the whole reactor primary circuit for such simulations can be of high computational cost. Therefore, a potential option is to model the reactor core as a porous medium.

The phenomenon will be modelled with the computational fluid dynamics code MC3D (ASNR, France), which is commonly used in fuel-coolant interaction modelling. The code is also validated for the modelling of debris bed as a high-temperature porous medium. A simplified JSI TRIGA reactor geometry will be used, with the active core being modelled as a porous region with a set volumetric power. The objective of our research will be to obtain the start-up of coolant natural convection in the reactor pool. The simulation results will be compared with the results from the past experimental campaign.

Due to its small size and a relatively small number of fuel rods (≈50), the TRIGA reactor can be studied using detailed CFD, but we adopt a simplified approach. Such simulations can be especially useful for modelling large reactors with tens of thousands of fuel rods, where detailed calculations aren’t viable without time-demanding calculations by powerful supercomputers.

The transition to a low-carbon energy system in Europe requires not only new generation capacities but also large-scale energy storage solutions to compensate for the intermittency of renewable sources. Slovenia alone has an estimated storage need of around 2 TWh. While hydrogen is frequently discussed as a storage medium, its low volumetric energy density and technical handling challenges remain significant drawbacks. Antimatter, by contrast, possesses the highest known energy density of any substance; approximately 180 × 10⁶ MJ/g for annihilation of a particle and antiparticle with 100% conversion of the combined rest mass into energy. The paper presents a comprehensive assessment of the current state of antimatter production, storage, and controlled energy release, and presents a conceptual design for an antimatter power plant.

Antimatter production on Earth relies on high-energy particle accelerators. At CERN, protons are accelerated to 26 GeV and directed into an iridium target, producing approximately one antiproton-proton pair per 250,000 collisions. The resulting antiprotons are decelerated in the Antiproton Decelerator and in the ELENA ring to 1.5% of the speed of light, before being directed into experimental traps. Current production efficiency, expressed as the ratio of energy stored in the antimatter to the energy consumed in its creation, is approximately 10⁻⁹ at CERN. Studies based on Fermilab data indicate that optimizing beam acceleration and improving antiproton collection could increase this figure to around 3.9 × 10⁻⁵ which is still far from breakeven but represents several orders of magnitude of improvement. Positrons can be obtained more efficiently from radioactive isotopes such as ²²Na (half-life 2.6 years) or ¹⁸F (half-life 110 minutes). Fluorine yields up to 3.4 × 10¹⁶ slowed positrons per second per gram of source material.

The storage of charged antiparticles is constrained by the Brillouin limit, which caps the achievable plasma density in a magnetic trap at n_B = B²/(2µ₀mc²). For a 6 T Penning-Malmberg trap, this corresponds to 2 × 10¹² antiprotons/cm³ or 3.7 × 10¹⁵ positrons/cm³. Density of 0.01×10⁸ antiprotons per cm³ was achieved in ALPHA experiment at CERN. Technical feasibility of long lifetime was demonstrated by the BASE experiment at CERN which stored antiprotons for over one year.

To overcome the Brillouin density limit for charged particles, we consider neutral antihydrogen, formed by combining antiprotons and positrons via three-body recombination or radiative recombination inside a Penning trap. The ALPHA experiment at CERN can routinely produce approximately 50,000 antihydrogen atoms every four minutes and has demonstrated storage of antihydrogen for 1,000 seconds in an inhomogeneous magnetic trap. A more ambitious storage option is to form a Bose-Einstein condensate (BEC) of antihydrogen. Hydrogen BEC has already been achieved at densities up to 4.8 × 10¹⁵ cm⁻³ at 50 μK, corresponding to an energy density of 208.9 kWh/dm³. That is more than ten orders of magnitude higher than what is currently achieved in antihydrogen traps. The main challenge is developing helium-free cooling techniques, since liquid helium would annihilate antihydrogen on contact.

Energy release through matter-antimatter annihilation produces a characteristic mix of particles. Proton-antiproton annihilation yields on average three charged and two neutral pions, which decay into gamma rays (up to 800 MeV), muons, neutrinos, electrons, and positrons. We calculate, using exponential attenuation with mass attenuation coefficients from the NIST XCOM database, that approximately 27 cm of lead is sufficient to absorb 99.99% of the gamma-ray energy for the relevant photon energies (511 keV to 800 MeV). Charged pion transport and secondary particle effects will be refined in future Monte Carlo simulations using OpenMC or MCNP. The captured thermal energy can be converted to electricity via a combined-cycle gas turbine (Brayton + Rankine), achieving efficiencies up to 64% as demonstrated by Siemens Energy.

We describe a conceptual antimatter power plant integrating all required subsystems: a proton accelerator complex, an antiproton decelerator, Penning-Malmberg traps for antiproton and positron storage, an antihydrogen synthesis apparatus based on the ALPHA design, a thick-walled annihilation chamber with lead or tungsten shielding, and a combined-cycle electricity generation system.

Several critical problems remain. Production efficiency is nine orders of magnitude below breakeven, and accumulating enough antiprotons for 1 TWh would take around 10¹² years. Storage densities are far below theoretical limits, antihydrogen BEC has not been demonstrated and the hydrogen BEC lifetime of only 15 seconds is far too short for grid-scale use. Neutrinos carry away up to half of annihilation energy and adequate shielding and conversion systems do not yet exist at scale.

Despite these challenges, the topic carries substantial importance beyond the distant prospect of an antimatter power plant. The accelerator and trap technologies developed in this pursuit directly advance nuclear and particle physics infrastructure, plasma confinement techniques, and cryogenic engineering. Byproducts of the same production chain, such as ¹⁸F, can also be used for PET imaging. With sustained research and targeted technological development, antimatter could one day become a viable component of high-performance energy systems. Until then, it remains a powerful concept driving innovation at the edge of physics and engineering.

Radiation surveillance is evolving from manual, high-risk procedures to advanced, data-driven systems. A key trend is the transition from static, human-operated ground measurements to dynamic, autonomous Internet of Things (IoT) solutions. In this context, Unmanned Aerial Vehicles (UAVs) now serve as intelligent tools within a collaborative IoT ecosystem. Integrating UAVs with IoT enables multi-level monitoring, mitigates human risk, and expands access to hazardous environments. The STU team addresses this evolution with the RaSens IoT-based radiation monitoring device, developed through the STU-GUARD and STU-RMS projects. RaSens modules operate from UAVs or as independently deployed nodes, facilitating real-time monitoring, airborne radioactivity mapping, and networked data sharing. Additionally, RaSens functions as a laboratory gamma counter to monitor staff or experiments. Such a system should be fast, cheap, reliable, flexible, and must withstand a relatively high level of radiation without malfunctions or false responses. This paper summarizes laboratory test results from two research projects, focusing on node sensitivity, linearity, and resistance to neutron and gamma radiation. Tests were performed at the Laboratory of Neutron Applications at STU, the Czech Technical University in Prague, the Slovak Institute of Metrology, and in field environments. A sample application scenario involving a radiation threat is also presented.

Trustworthy multiphysics (MP) modelling is now paramount across the nuclear sector, as both the Long-Term Operation (LTO) of existing plants and the design of emerging systems (SMRs, Gen IV) require accurate understanding of complex coupled physical phenomena. Currently, precise and reliable modelling of these systems face significant limitations due to approximate physical models, discretization strategies, and a severe lack of high-fidelity experimental benchmark data. The OECD Nuclear Energy Agency (NEA) Expert Group on Multi-physics Experimental Data (EGMUP) has explicitly highlighted this gap, emphasizing the urgent need for experimental data of pronounced coupled phenomena captured with high-quality measurement systems.

To address this global challenge and bridge the observed discrepancies between simulations and experiments-which can reach up to 1500 pcm in reactivity and 10-15% in absolute neutron reaction rates - a comprehensive instrumentation framework is being developed at the Jožef Stefan Institute (JSI) TRIGA Mark II research reactor. Operating under the synergistic efforts of the EURATOM EVEREST, national CHAMP, and bilateral OPTI-Bench projects, this research aims to deploy a cohesive, multi-modal measurement platform capable of mapping core and pool parameters with sub-centimeter and sub-second resolution.

The experimental framework integrates multiple state-of-the-art measurement modalities to capture a complete multiphysics profile:

• Neutronic characterization: To provide a full suite of multiphysics measurements, the platform is augmented by established, high-fidelity neutronic diagnostic capabilities. Absolute Fission rates and neutron flux distributions are precisely quantified throughout the core using the Libera MONACO 3 system, CEA miniature fission chambers, activation foils, and advanced scintillator-fiber detectors developed at Ecole Polytechnique Federale de Lausanne.
• Thecmocouple Profiling: Distributed thermocouple arrays are strategically positioned within the reactor pool and directly inside the core. These provide critical baseline coolant temperature profiles. However, to overcome the inherent limitations of thermocouples in harsh radiative environments and improve spatial resolution, the transition to distributed optical fiber sensors, specifically utilizing Bragg optical gratings is also being actively investigated. These robust optical profiles will serve to establish precise thermal gradients and cross-validate the platform's emerging non-invasive measurement methodologies.
• Ultrasonic temperature tomography: A focal point of this development is the implementation of ultrasonic tomography for non-invasive, precise water temperature mapping. Initial simulation analyses have already been presented, utilizing full-wave propagation (k-Wave) and travel-time inversion algorithms (pyGIMLi) to demonstrate the feasibility of extracting time-of-flight (TOF) data across the 2-meter diameter tank. Based on these computational studies, the physical setup utilizes a ring of up to 128 ultrasonic transducers mounted on a movable structure at the periphery of the reactor tank. A transducer frequency of 40 kHz was specifically selected, as earlier analyses proved it offers an optimal balance—minimizing signal attenuation across the massive tank while providing sufficient amplitude and temporal resolution to clearly detect minor temperature gradients in the coolant.
• Volumetric reactor power measurement: The instrumentation suite also incorporates several high-precision water level measurements to continuously monitor the pool's surface level. By accurately tracking the volumetric expansion of the coolant, the reactor's overall thermal power can be independently deduced and correlated with other operational parameters.
• Laser projected Schlieren imaging: Fluid dynamics and natural convection patterns are characterized using optical techniques, specifically Background Oriented Schlieren (BOS) imaging. In this setup, a laser projects a speckle grid that acts as a high-spatial-frequency noise background. As density and temperature variations alter the water refraction index, the coolant flow is deduced by computationally tracking the minute changes in the distances between individual dots within the projected grid. This enables the non-intrusive, high-resolution mapping of complex water flow patterns.
• Inductive fuel measurements: To assess changes within the fuel itself, the platform incorporates advanced inductive measurement techniques. The initial goal is to assess the burn-up correlation response, with future implementations aiming to discern the fuel temperature distribution during active reactor operation.

The proposed experimental setup moves beyond traditional, low-resolution monitoring to produce open-access, benchmark-quality datasets. These high-fidelity measurements will ultimately reduce conservativeness in safety margins and enable the rigorous validation of the coupled neutronics and thermal-hydraulic codes essential for the design, licensing, and lifetime extension of the next generation of nuclear reactors.

Across Europe, substantial investments are being made in research and development of advanced reactor technologies. These promise improved safety characteristics, more efficient use of nuclear fuel and greater flexibility for use in electric and non-electric applications. Yet, success of these technologies will also depend on how they are introduced, explained, and discussed beyond the nuclear community. For emerging reactor concepts such as lead-cooled systems, little evidence exists on how citizens interpret these technologies, which factors drive their support or concern, and whether seemingly minor differences in technology descriptions can influence public attitudes. Addressing these questions is becoming increasingly important as several European countries move from reactor research and development towards demonstration and deployment.

Within the European ANSELMUS project, an interdisciplinary research task investigates the societal and ethical dimensions of advanced lead-cooled nuclear technologies. The work combines sociotechnical integration activities in research laboratories, structured dialogues with citizens and comparative studies of public perceptions to generate insights that can inform future deployment strategies.

This contribution will present, for the first time, initial results from a large-scale comparative survey conducted in Belgium and Romania, two countries actively engaged in research and development on advanced reactor and SMR. The survey explores citizens' perceptions, attitudes, and support of lead-cooled reactor technology, as well as the factors that shape these views.

A particular innovation of the study is the investigation of how technology framing influences public perceptions. Using a CAWI (Computer Assisted Web Interviewing)-based survey experiment with representative population samples, participants were randomly exposed to different descriptions of the technology, presenting it either as a "new" reactor technology, an "advanced" reactor technology, or without a specific framing. This design allows examining how nuclear terminology and communication choices influence perceptions of risk and benefit, emotional responses, trust in responsible actors, and willingness to support future deployment.

The first results provide novel insights into the relationship between communication, trust, perceived risks and benefits, and public support for advanced nuclear technologies. They demonstrate that societal responses are shaped not only by assessments of technical performance but also by broader considerations related to governance, fairness, and confidence in institutions. The findings suggest that societal considerations should be treated as a deployment parameter for advanced reactors, alongside safety, performance, and economics.

Conference participants will gain early insights into how citizens in two European countries perceive advanced lead-cooled reactor technologies and how subtle differences in technology descriptions can influence public attitudes and support. These findings can help reactor developers, researchers, and policymakers better anticipate societal expectations and design communication and engagement strategies that contribute to the successful deployment of next-generation nuclear systems.

Acknowledgment: This project has received funding from the European Union’s Horizon EURATOM 2021 Research and Training Programme under grant agreement No 101061185.

Slovenia is entering a new phase of energy development marked by discussions surrounding the construction of a second nuclear power plant unit (JEK2), long-term decarbonization goals, and the strengthening of national energy security. These developments have exposed a critical shortage of nuclear professionals and highlighted the need for a systematic renewal of nuclear education in Slovenia. In response, the University of Maribor is developing a new undergraduate study program in Nuclear Engineering that aims to provide interdisciplinary, practice-oriented, and internationally connected education aligned with the future needs of the Slovenian and European nuclear sectors.

The proposed Bachelor’s program is being developed at the Faculty of Energy Technology in Krško in cooperation with Slovenian industry stakeholders, including GEN Energija and the Krško Nuclear Power Plant (NEK), as well as with the French International Institute for Nuclear Energy (I2EN). The program is designed as a strategic contribution to the development of national competencies required for future nuclear projects, operation and maintenance of nuclear facilities, radiation protection, regulation, and related technological sectors.

The curriculum combines strong scientific and engineering foundations with specialized nuclear content. The educational philosophy of the program follows a three-step progression: (1) building strong scientific and engineering foundations, (2) developing nuclear-specific knowledge and operational understanding, and (3) integrating multidisciplinary competences through practical applications, project work, and professional training. The program is intended not only as preparation for Master-level nuclear studies but also as a pathway toward immediate professional integration into the nuclear industry.

A major characteristic of the program is its interdisciplinary structure. Experts from multiple faculties of the University of Maribor participate in curriculum development and future implementation, including faculties specializing in mechanical engineering, electrical engineering, computer science, chemistry, mathematics, civil engineering, economics, law, logistics, and organizational sciences. This structure reflects the inherently interdisciplinary nature of nuclear technology and modern energy systems. Cooperation with I2EN and potential French academic partners aims to strengthen educational quality, student mobility, hands-on training opportunities, and alignment with European nuclear education standards.

An important milestone in the development process was the preliminary assessment performed by I2EN experts in 2025. The evaluation recognized the program as a credible and strategically important initiative for Slovenia’s nuclear future. According to the assessment, the program demonstrates strong institutional commitment, clear alignment with national workforce needs, and a solid academic structure. I2EN experts particularly highlighted the strong foundations in nuclear science and engineering, the involvement of Slovenian industry stakeholders, and the relevance of the program in the context of the planned JEK2 project and the broader European nuclear renaissance. The I2EN review also provided several recommendations that are currently being integrated into further program development. Based on the evaluation, the University of Maribor’s Bachelor program in Nuclear Engineering was awarded the status of “I2EN Seal Candidate.”

The presented initiative demonstrates how smaller European countries can strategically rebuild nuclear knowledge ecosystems through interdisciplinary collaboration, strong industry involvement, and international partnerships. The program represents not only an educational project but also a broader contribution to technological sovereignty, sustainable energy transition, and long-term societal resilience.

Passive safety systems are increasingly considered for Small Modular Reactors (SMRs), where buoyancy-driven natural convection may be encountered under various operating and accident scenarios. Such configurations can involve Rayleigh numbers approaching 10¹⁵, which represent a significant challenge for turbulence modelling. Validating the thermal-hydraulic codes used to simulate such systems under accident scenarios requires high-quality experimental data, particularly at the elevated Rayleigh numbers (approaching 10¹⁵) characteristic of SMR operating conditions. To address this need, a new experimental campaign is being conducted at the PANDA facility (Paul Scherrer Institut, Switzerland) under the OECD/NEA PANDA project, with the objective of producing benchmark-grade measurements of natural convection heat transfer at high Rayleigh numbers.

The present work is part of an ongoing computational effort using EDF's open-source code_saturne software to support the design and interpretation of this experimental campaign. Previous studies [1,2] established a 2D axisymmetric modelling framework and demonstrated, through preliminary simulations, that Rayleigh numbers of 10¹⁵ are achievable within the facility's operational constraints. That work also identified key sensitivities affecting the predicted Nusselt–Rayleigh correlations, namely the influence of the initial bulk temperature of the water pool and the choice of turbulence model. Two RANS turbulence models were compared: a k-ω Shear Stress Transport (SST) eddy viscosity model and the Elliptic Blending Reynolds Stress Model (EBRSM). Results showed that turbulence model selection has a significant impact on predictions of the laminar-to-turbulent transition that occurs in the lower part, peak wall temperatures, and boundary layer development, and underscored the importance of a detailed near-wall temperature measurement strategy for future code validation.

Based on these findings, the present contribution reports the next stage of the computational program: the transition from the simplified 2D axisymmetric geometry to a full 3D model based on the actual PANDA facility CAD geometry. This step is essential, as the real experimental configuration features geometric asymmetries, complex boundary conditions, and three-dimensional flow structures that cannot be captured in an axisymmetric framework. The 3D model is expected to provide more realistic predictions of the flow field and heat transfer distribution within the pool, including the development of large-scale circulation patterns and their interaction with the heated walls.

The 3D simulations are used to anticipate the experimental results expected from the PANDA campaign, the data from which are scheduled to be released in mid-2026. By providing pre-test predictions, the present work aims at supporting the interpretation of the forthcoming measurements. In particular, the predicted Nusselt–Rayleigh correlations obtained with the 3D model are compared against those derived from the earlier 2D calculations and against established empirical correlations, to assess the impact of geometry on the predicted heat transfer behavior.

References

[1] A. Morente, J. Uribe, S. Benhamadouche, P. Rotach, J.-L. Vacher, M. S. Chae, D. Paladino, "Scoping Analyses for the Definition of New Experiments for Natural Convection at High Rayleigh Numbers," in Proceedings of Advances in Thermal Hydraulics (ATH 2024), Orlando, FL, USA, November 17–21, 2024, pp. 202–211. DOI: 10.13182/T131-45695.

[2] A. Morente, J. Uribe, S. Benhamadouche, P. Rotach, M. Montout, J.-L. Vacher, D. Paladino, M. S. Chae, "New Experiments for Natural Convection at High Rayleigh Numbers: Definition, Sizing and Analysis using CFD," in Proceedings of the 21st International Topical Meeting on Nuclear Reactor Thermal Hydraulics (NURETH-21), Busan, Republic of Korea, August 31–September 5, 2025.

Cavitation in marine propellers can lead to significant performance degradation, damage, and unwanted noise. The suppression of cavitation is challenging, however, it should be considered to enhance the efficiency and performance of the propeller.
This study investigates the effect of the water injection technique on the suppression of cavitation in a propeller tunnel setup. CFD simulations were performed with and without water injection. A set of injection velocities (1, 1.911, 2.5, 3, 5, 10, and 50 m/s) was used. The results show that the cavitation can be effectively suppressed up to 15% if the injection velocity is in the range of 52.91% to 264.06% of the tunnel flow velocity which was 1.89 m/s.
A significant reduction in pressure & thrust fluctuations was recorded along with the shrinkage of the separation zone. Typical bow thruster vortex structures have also been identified. It was found that the injection velocity close to the tunnel flow velocity is not an optimal choice as it leads to an increase in thrust fluctuation. Moreover, a high velocity of 10 or 50 m/s is also not suggestive.

During reactor operation, nuclear fuel assemblies are subjected to a combination of irradiation-induced and thermal-mechanical phenomena, including swelling, creep, and thermal expansion. These effects progressively lead to geometrical deformations of the fuel bundle, which may significantly alter the local flow distribution and heat transfer characteristics. In lead-cooled fast reactors (LFR), where thermal margins can be strongly affected by local flow perturbations, such deformation-induced changes may result in temperature non-uniformities and, in some cases, the formation of local hot spots. These aspects are particularly relevant for reactor safety and for the reliable design and licensing of advanced systems such as the ALFRED reactor.

Despite their importance, the impact of fuel bundle deformation on thermal-hydraulic performance is still not fully understood, and the predictive capability of current computational fluid dynamics (CFD) tools under such conditions remains to be thoroughly assessed. In this context, the present work aims to provide a comprehensive experimental and numerical investigation of heat transfer modifications in a deformed fuel bundle representative of a lead-cooled reactor configuration. The study combines a dedicated experimental campaign with a structured code validation and benchmarking activity involving multiple partners.

The experimental campaign will be carried out in the NACIE-UP loop at the ENEA Brasimone Research Centre in Italy, a facility designed for thermal-hydraulic investigations in flowing liquid lead under controlled chemistry conditions. The test section consists of a 19-pin rod bundle equipped with grid spacers, designed to simulate the geometry of the ALFRED fuel assembly. The bundle features a pin diameter of 10 mm, a pitch-to-diameter ratio (p/d) of 1.3, and an active heated length of 600 mm. The nominal operating conditions include an average heat flux of approximately 700 kW/m² and flow velocities representative of the ALFRED reactor.

A key feature of the experimental set-up is the ability to introduce controlled geometric deformations within the active region of the pins. The fuel pin simulator accommodates three deformed rods, one for each radial rank, each characterized by a maximum deformation at the mid-plane of the active length. The relative position and orientation of the deformed rods was defined during the design phase through CFD simulations and the deformation of the pins was measured by profilometer before assembling the final rod bundle.

The instrumentation of the test section has been specifically designed to provide high-resolution data for code validation purposes. The rods are equipped with multiple wall thermocouples to capture detailed temperature distributions along both axial and circumferential directions. Additional thermocouples are installed outside the bundle to quantify heat losses to the surroundings, ensuring an accurate thermal balance. Pressure losses along the test section are continuously monitored, and the lead mass flow rate is measured using a thermal mass flow meter. Different flow conditions may be explored for selected configurations to assess the sensitivity of the system response to varying Reynolds numbers, while maintaining relevance to nominal reactor conditions.

The experimental program is structured within the EU project ANSELMUS that wants to pave the way for the LFR technology. In this context, several participants joined the effort to reach this goal. CFD calculations will serve both as an initial assessment of model performance and as a support for optimizing the experimental design, including sensor placement and test matrix definition.

Following the blind phase, a benchmark exercise will be conducted, allowing a direct comparison between numerical predictions and experimental measurements. The outcomes of this phase enable the identification of modelling deficiencies and guide the improvement of numerical approaches. Subsequently, enhanced models will be applied to simulate additional configurations experimentally investigated in the NACIE-UP loop.

Overall, the present work aims to generate a high-quality experimental database and to advance the validation of CFD tools for lead-cooled reactor applications under geometrically perturbed conditions. The results are expected to contribute to a better understanding of deformation-induced thermal-hydraulic phenomena and to support the development of more reliable predictive models for the design and safety assessment of advanced nuclear systems.

Nuclear research reactors (RR) are essential facilities for advancing nuclear science and technology, educating students and nuclear professionals, producing medical isotopes, and testing future energy technologies such as fusion, small modular reactors (SMRs), and Generation IV reactors. The fleet of European research reactors is ageing (average age of 60 years) and its numbers are decreasing, while the demand for RR services is growing in line with Europe’s renewed interest in nuclear energy and the transition to a zero-carbon society. The VERONICA (Versatile European Reactor for Neutron Irradiation and Nuclear Research) project, a collaboration between the Jožef Stefan Institute (JSI) and CEA, addresses these challenges by proposing a new dual-core research reactor facility in Slovenia, comprising a Multi-Purpose Research Reactor (MPRR) and a Zero-Power Reactor (ZPR). The MPRR will mainly focus on irradiation experiments to support radiation hardness studies, material science, instrumentation testing and qualification, neutron beam applications, isotope production, and novel multiphysics experimental capabilities, while the ZPR will serve education and training, reactor physics measurements, neutron code validation, and benchmark experiments. So far, in line with IAEA’s strategic planning for RR [1], the technical requirements for both reactors were formalized by covering more than 35 identified applications for the MPRR and 15 for the ZPR. The main conclusions [2] are that the MPRR should achieve an in-core thermal neutron flux on the order of 10¹⁴ n∙cm^-2∙s^-1 in an open pool-type, water-cooled configuration with a high level of versatility. This feature includes, in particular, the ability to operate the reactor while modifying the geometric configurations of the core and its reflector, as well as the operating modes, enabling a high utilization level across multiple fields. The ZPR should enable the flexibility in performing various benchmark experiments, while being used for high-level education & training. This paper will report on the progress in technical design of both reactors and the initial computational characterization of the MPRR.

The first part will present and analyze the main design features of both the MPRR and ZPR. For the MPRR, this includes the design of several novel experimental facilities. The reactor will include various locations that will enable a wide range of flux levels (up to several 10¹⁴ n∙cm^-2∙s^-1 ) as well as different spectral properties, ranging from hard-spectrum locations in the core to thermal-spectrum locations in the reflector, and even cold-spectrum locations in the reflector, which in the MPRR design is constituted of a heavy water tank. A spectrum-shifting device is considered in the reflector, based on dedicated neutron filters or absorbers, which will enable access to some degree of neutron spectrum shifting. A dedicated experimental loop for multiphysics investigations will allow controlled variation of coolant flow rate, inlet temperature, and pressure to generate coupled neutronics–thermal-hydraulics benchmark datasets under both steady-state and transient conditions, in line with OECD/NEA EGMUP objectives. High and low temperature irradiation ports will enable the irradiation of samples in different environments, crucial for the development of fusion and Gen IV technologies. Additionally, more general MPRR components will be presented, which include the irradiation positions for medical isotope production, silicon doping and neutron beam experimental ports. For the ZPR, the main design features will be discussed, such as the flexible core configuration, the fuel loading approach, and the experimental capabilities for reactor physics measurements, neutron code validation, and benchmark experiments.

In the second part, preliminary Monte Carlo neutron transport calculations for the MPRR will be presented. These simulations will provide initial estimates of the neutron flux levels and spectral characteristics achievable in the reactor core, reflector, and experimental positions, and assess the effect of spectrum-shifting inserts on the available neutron energy distributions. The results will serve as an early computational characterization to verify that the main design targets identified in the technical requirements phase – in particular the in-core flux level and spectral versatility – are attainable within the chosen reactor concept, and to guide further design decisions ahead of the conceptual design phase.

By presenting the progress in technical design of the VERONICA facility and preliminary neutron transport results for the MPRR, this paper aims to contribute to the ongoing European nuclear dialogue and invite collaborations and feedback from the wider nuclear science and technology community as the project advances towards its conceptual design.

[1] INTERNATIONAL ATOMIC ENERGY AGENCY, Strategic Planning for Research Reactors, IAEA Nuclear Energy Series No. NG-T-3.16, IAEA, Vienna (2017)

[2] Pungerčič, Anže, et al. "Versatile European Reactor for Neutron Irradiation and Nuclear Research (VERONICA): Technical Requirements and Multiphysics Relevance," PHYSOR 2026 – The International Conference on Physics of Reactors, Torino, Italy, April 19–23, 2026.

Good characterisation and realistic modelling of radiation source is essential first step in the benchmark experiment analysis. Examples of neutron and gamma source modelling for Monte Carlo and deterministic codes for a list of shielding benchmarks will be provided, including:

- Frascati Neutron Generator (FNG) D-T 14 MeV neutron source for W benchmark, replacing the neutron source subroutines (source.F90 and srcdx.F90) in the MCNP source code,

- IFMIF-DONES heavy concrete shielding mock-up experiment performed in 2024 at Rez facility cyclotron using 35 MeV protons impinging on an 8mm thick Be target and providing neutrons of energies up to 33 MeV,

- KFK neutron/gamma Iron Sphere Benchmark using a bare ²⁵²Cf(s.f.) source

- TIARA Iron/Concrete benchmarks using neutrons at 0° produced by 43- and 68-MeV proton on a lithium

Source description for use with the deterministic solvers is somewhat different compared to the description in MCNP. Furthermore, different sources in MCNP are often described using specialised source subroutine, e.g. ENEA-JSI DT source subroutine or McDeLicious. To deal with this a simple way to convert essentially any MCNP source to description used in deterministic solvers was devised. The method used is also valuable for Monte Carlo code users allowing to avoid tedious use and adaptation of source subroutines from one (version of) code to another.

Increasing renewable resource electric generation capacity has further exacerbated the need for flexible nuclear power plants. Load following operation would be a great asset for nuclear power plants, especially for complementing the variable output of renewable energy sources. In this work, we present the workflow for a simple load follow simulation of a small modular reactor. The simulation is performed using the OpenMC code, which allows for detailed modeling of the reactor core and its behavior under different operating conditions.

In this work, we enhanced the Critical Density Iteration algorithm to allow for in-memory adjustment of control rod positions during a single simulation step, without a need for an outerloop iteration search. A simple thermal-hydraulic feedback model is implemented to account for the changes in coolant and fuel temperatures during load following, under channel-flow assumption. Thermal-hydraulic feedback is also implemented in-memory to increase the simulation efficiency. At the beginning-of-life steady-state full-power conditions, 100–50–100% load following maneuver is simulated by adjusting the control rod positions to achieve the desired power output curve. In the inactive part of the Monte-Carlo simulation, the fuel temperature and
coolant density are updated, which in turn affects the neutron flux distribution and the reactivity of the core. Fuel composition is assumed constant, with only xenon and samarium decay chain concentrations being updated during the simulation using a simple point-kinetics model in each fuel segment.

This work is a feasibility study for future development on nuclear complementarity to renewable energy sources, taking into account the transient behavior of the thermal feedback effects on the reactor core during load following operation. Such workflow can be used as a basis for more detailed multi-physics coupling, which would allow for more accurate simulations of load following operation and its impact on the reactor core.

High-porosity MOX fuels are of interest for advanced fast-reactor concepts, but their reduced thermal conductivity can strongly affect fuel temperature, thermal gradients, and safety margins. Reliable thermal-conductivity data are therefore needed to support fuel-performance modelling and fuel qualification.

This work presents an integrated experimental and modelling approach for assessing the thermal conductivity of high-porosity MOX samples produced in the framework of ESFR-SIMPLE. Thermal-diffusivity measurements are combined with density and heat-capacity information to derive thermal conductivity as a function of temperature and porosity. In parallel, a heat-transfer model of the experimental setup is developed in COMSOL Multiphysics and used as a digital twin of the measurement configuration. The model helps evaluate the influence of sample geometry, boundary conditions, heat losses, contact effects, and measurement uncertainty on the inferred thermal-conductivity values.

A Bayesian design-of-experiments strategy is also introduced to support the selection of the next most informative sample or measurement condition. Based on the available data and uncertainty in the conductivity-porosity relationship, the method ranks candidate measurements according to their expected contribution to reducing model uncertainty. This can help prioritise experimental work when sample availability, fabrication effort, or measurement time are limited.

The proposed workflow links thermophysical characterization, multiphysics modelling, and data-informed experimental planning. It provides a practical route to improve the reliability, traceability, and usefulness of thermal-conductivity data for high-porosity MOX fuels relevant to advanced reactor applications. In addition, the resulting dataset and uncertainty analysis are intended to support the assessment and future refinement of thermal-conductivity correlations for fuel-performance modelling, including applications in TRANSURANUS.

Boron carbide (B₄C) is one of the most important ceramic materials for nuclear technology due to its high neutron absorption capability, excellent thermal stability, chemical resistance, and superior mechanical hardness. In nanoscale form, B₄C particles can exhibit enhanced sensitivity to irradiation-induced structural and electronic modifications because of their high surface-to-volume ratio and defect-active interfaces. In this study, neutron irradiation-induced paramagnetic centers and vacancy-type defects in nanocrystalline B₄C particles were investigated using Electron Paramagnetic Resonance (EPR) spectroscopy. Rhombohedral B₄C nanoparticles with an average particle size of approximately 30 nm and a specific surface area of 150 m²/g were irradiated in the F19 channel of the TRIGA Mark II research reactor at the Jozef Stefan Institute, Slovenia. The irradiation was performed under full reactor power conditions with neutron fluences of 1.6 × 10¹⁵, 8 × 10¹⁵, 4 × 10¹⁶, and 2 × 10¹⁷ n/cm². EPR measurements were carried out at room temperature using an X-band Bruker EMX II plus spectrometer. Spectra were recorded over a broad magnetic field range of 500–5500 G and were further analyzed in the narrower region around 3300–3700 G, where the most intense irradiation-sensitive signals were observed.

The EPR spectra revealed that neutron irradiation significantly modifies the paramagnetic structure of nanocrystalline B₄C particles. A strong resonance signal near 3500 G became more pronounced with increasing neutron fluence, especially at the highest irradiation dose. The signal with g = 2.003252 was attributed to free-electron-like paramagnetic centers and stable irradiation-induced defect states. The observed spectral evolution is mainly associated with neutron transmutation reactions involving boron isotopes. In particular, the interaction of neutrons with ¹⁰B and ¹¹B isotopes leads to the formation of secondary isotopes such as ⁷Li and ¹²C, which contribute to vacancy formation and defect restructuring in the B₄C lattice. The increase in carbon isotope concentration and the formation of boron vacancies were identified as key mechanisms responsible for the strengthening of EPR signals.

At higher neutron fluences, the signal observed around 2400 G gradually weakened, indicating the transformation or annihilation of specific pre-existing defect centers. Simultaneously, the intensification of the main EPR signal confirmed the generation of new stable paramagnetic centers. The results suggest that neutron irradiation promotes the formation of boron vacancies, carbon vacancies, and complex defect centers, including vacancy-related boron configurations. These changes can strongly influence the electronic and radiation-response properties of nanocrystalline B₄C. Overall, this work demonstrates that EPR spectroscopy is a highly sensitive and effective technique for detecting neutron-induced paramagnetic defects in B₄C nanoparticles. The findings provide important insight into the defect formation mechanisms of boron carbide under neutron irradiation and support its relevance as a neutron absorber and radiation-resistant material for advanced nuclear applications.

Neutron transport calculations are essential for the analysis, control, and safe operation of nuclear reactors. In practice, they are usually performed using one of two main approaches: deterministic solution of the neutron transport equation or stochastic Monte Carlo neutron transport. Deterministic methods can provide detailed three-dimensional neutronic quantities throughout the modeled system. However, for realistic reactor geometries such calculations may become computationally demanding. Monte Carlo (MC) methods are the reference approach for criticality and detailed neutronics analyses, owing to their use of exact geometry and continuous energy treatment. There are several MC codes available for the nuclear community (e.g. MCNP, Serpent-2, TRIPOLI, SuperMC, etc.), however none of them are open-source. OpenMC is a community-developed open-source Monte Carlo neutron transport code capable of standard reactor calculations, including fixed-source, subcritical multiplication, and k-eigenvalue calculations. Its open-source and highly programmable structure make it particularly suitable for transparent model development, systematic validation, and research-oriented reactor-physics studies [1].

This work presents validating the model and verifying the OpenMC implementation against MCNP. Two well-documented JSI TRIGA benchmark experiments were used: the ICSBEP criticality benchmark and the IRPhEP axial fission rate benchmark. For both experiments the OpenMC calculated reactor parameters were validated by comparison with the experimental results and verified by comparison with results obtained with MCNP Monte Carlo code. All calculations presented in the paper were performed using OpenMC version 0.15.2 together with the ENDF/B-VIII.1 nuclear data library. Both, criticality calculations and axial fission rate distributions, were performed using 1000 batches and 5·10⁶ neutrons.

The paper presents verification and validation of two experiments. The first is the criticality benchmark experiment that was performed in 1991 and is described in the ICSBEP Handbook [2]. In this experiment, two critical core configurations (core no. 132 and 133) were operated. Both configurations had the same number of fuel elements but different loading pattern. For both benchmark core configurations, the OpenMC calculated multiplication factors are within 1σ statistical uncertainty of the experimental results. OpenMC overestimates the multiplication factor for around 100 – 200 pcm. The OpenMC results show slightly better agreement with the experimental values than MCNP, with deviations smaller by approximately 150 pcm and 240 pcm for core configurations 132 and 133, respectively.

The second experiment is the axial fission rate benchmark, published in the IRPhEP handbook [3] for which the OpenMC calculated axial fission rate distributions were compared with the corresponding experimental benchmark data and with the reference MCNP calculations. In the benchmark experiment, axial fission rate profiles were measured in the reactor core using absolutely calibrated miniature fission chambers inserted into aluminum guide tubes at several radial measuring positions. For all measuring positions, the OpenMC and MCNP results are in very close agreement over the entire axial range. Compared with the experimental benchmark data, OpenMC reproduces the MCNP reference profile within statistical uncertainty, while both codes underestimate the measured fission rates by 5 % - 12 % in the central region. The OpenMC implementation is therefore considered verified against MCNP for this geometry, with the model-experiment discrepancy attributed to the underlying nuclear data, geometry, or modeling assumptions shared by both codes. The two calculated profiles are nearly superimposed in both the absolute and normalized representations, which indicates that the OpenMC model reproduces the same axial response as the reference MCNP model to a very high degree. This is an important result, since it shows that the OpenMC implementation of the reactor geometry, materials, and tally treatment is consistent with the MCNP code and its geometrical model. Main conclusion of the presented research is that OpenMC calculated results are almost completely consistent with the MCNP results and in very good agreement with the experimental results. This indicates that the OpenMC model is consistent with the MCNP reference and is suitable for further research applications.

[1] Paul K. Romano, Nicholas E. Horelik and K. Smith. Openmc: A state-of-the-art monte carlo code for research and development. Annals of Nuclear Energy, Vol. 82, pp. 90-97, 2015.

[2] Jeraj, R., Ravnik, M., 1999. TRIGA Mark II Reactor: U(20)-Zirconium Hydride Fuel Rods in Water with Graphite Reflector, IEU-COMP-THERM-003. In: International Handbook of Evaluated Criticality Safety Benchmark Experiments. NEA 7328, OECD Nuclear Energy Agency, Paris.

[3] Štancar, Ž., Snoj, L., Barbot, L., Destouches, C., Lell, R., 2017. Reaction Rate Distribution Experiments at the Slovenian JSI TRIGA Mark II Research Reactor, TRIGA-FUND-RESR-002. In: International Handbook of Evaluated Reactor Physics Benchmark Experiments. NEA 7329, OECD Nuclear Energy Agency, Paris.

This article presents a framework for evaluating the boron-10 (B-10) isotopic fraction within the total boron concentration of the primary coolant system over a typical 18-month fuel cycle in pressurized water reactors (PWRs). Reactivity control is achieved through the addition of boric acid to the coolant, with B-10 serving as a key thermal neutron absorber. During extended fuel cycles, the limited addition of fresh boron to the reactor coolant system (RCS) leads to gradual B-10 depletion, resulting in changes to the isotopic composition of boron. The study provides practical examples illustrating how the B-10 fraction can vary significantly under transient operating conditions and examines the resulting impact on chemically measured boron concentrations, which represent total boron content but do not directly reflect the effective B-10 fraction. These findings highlight the importance of accounting for isotopic variation when assessing boron behavior and reactivity control in PWR systems.

Water activation remains one of the key unresolved challenges in fusion reactor technology. In water-cooled systems such as ITER and DEMO, neutron-induced reactions in oxygen isotopes generate short-lived radionuclides, primarily N-16, producing intense gamma radiation fields extending up to 7 MeV, and lower intensity neutrons from N-17 decay chain. These distributed radiation sources strongly influence shielding design, detector performance, maintenance planning, and operational safety. Despite their importance, benchmark-quality experimental data for validation of water activation models remain extremely limited, particularly under controlled and accessible irradiation conditions.

The KATANA water activation facility at the Jožef Stefan Institute TRIGA reactor was commissioned in 2024 in Slovenia as a dedicated experimental platform for neutron-induced water activation studies relevant to fusion and advanced nuclear systems. The facility represents, to the best of current knowledge, the only operational and accessible closed-loop water activation platform capable of producing controlled high-energy gamma and neutron radiation fields under fusion-relevant conditions. Initial experimental campaigns have successfully demonstrated stable operation, reproducible activation conditions, and clear detection of characteristic radiation signatures associated with activated cooling water.

This contribution presents the strategic roadmap for the further development of the KATANA facility toward benchmark-quality experimental capability and the establishment of KATANA as a reference facility for high-energy gamma-ray detector calibration in support of ITER and future fusion facilities. The presented roadmap is structured around several key milestones required to transform the existing operational platform into a benchmark-standard experimental infrastructure.

The first milestone focuses on the establishment of benchmark-quality experimental conditions through the absolute calibration of gamma and neutron detector systems, neutron-field characterisation, reduction of systematic uncertainties, optimisation of shielding and collimation arrangements, and definition of reproducible reference configurations. The second milestone addresses validation of advanced fluid-activation and radiation transport methodologies through dedicated benchmark-oriented experimental campaigns relevant for ITER water activation studies.

A major milestone is establishing a dedicated reference environment for calibration and qualification of high-energy gamma-ray detectors up to 7 MeV energy range (greatly extending calibration curve beyond typically <2 MeV range from well-known standards) using activated water as the radiation source. Such capabilities are directly relevant for ITER diagnostics and fusion-related radiation measurements, where conventional calibration sources are insufficient. Additional roadmap activities include shielding studies using ITER-relevant materials, analyses of long-term JSI TRIGA reactor fuel burnup effects on KATANA irradiation conditions, and optimisation of reactor core reshuffling strategies to preserve stable fast-neutron conditions for future benchmark campaigns. Further development directions include redesign of the inner irradiation part toward improved ITER-relevance and as well preparation of upgrade paths toward Activated Corrosion Product studies.

The presented roadmap positions KATANA as a promising long-term experimental asset for fusion neutronics, benchmark activation studies, detector development, and validation of advanced computational tools in support of ITER operation and future fusion energy systems.

As the development of fusion reactors advances, the safety of personnel during maintenance periods and decommissioning phases remains a primary concern. Radiation safety, maintenance planning, and licensing of future fusion facilities such as ITER and DEMO, can be ensured through accurate assessment of material activation and shutdown dose rate (SDDR). To achieve precise SDDR calculation, codes are being developed based on the Rigorous Two-Step (R2S) and Direct One-Step (D1S) methods. Given the limited number of fusion facilities and relevant benchmarks, new experiments are required for the assessment of fusion-relevant materials, as well as for the validation of SDDR codes, particularly the JSIR2S code, developed at the Jožef Stefan Institute, which is based on the R2S method.

This work presents the design of an experimental campaign at the TRIGA Mark II research reactor to generate benchmark data for future fusion-relevant activation and SDDR analyses. The primary objective is to expand the established irradiation and measurement methodology to support subsequent validation of the JSIR2S SDDR computational code. This focus is on the experimental design and preparatory neutronics analyses. Experimental results and code-to-experiment comparisons are beyond the scope of this paper.

The TRIGA reactor has a well-characterised neutron spectrum, stable operation, and standardised irradiation processes. The experimental setup includes selected material samples positioned in the reactor core via the pneumatic mail system. Irradiation parameters, including reactor power, irradiation time, and sample cooling time before measurements, are determined based on preliminary neutronics calculations to ensure relevance to fusion applications. The experiment will be conducted in two parts. In the first part, smaller material samples, such as foils, will be used to measure activation and material composition with gamma spectroscopy, using an HPGe detector. In the second part, larger samples, such as pellets, will be used to measure the shutdown dose rate with an ionisation chamber. Both measurements will be performed at predefined cooling intervals. These results will provide direct input for validation of the JSIR2S code. Materials are selected based on high activation of in a fast neutron spectrum, varying half-lives of daughter isotopes, and current or future use in fusion facilities such as ITER.

Preliminary simulations using the JSIR2S, FISPACT, and MCNP codes will support the experimental design, including geometric modelling of the irradiation configuration and selection of appropriate nuclear data libraries. Based on these calculations, a validation methodology is defined, with comparisons between calculated and measured dose rates using calculation-to-experiment (C/E) metrics planned for future work.

The outcome of this work is a well-defined experimental and computational framework generating benchmark data suitable for SDDR code validation. The planned experiment will enable systematic evaluation of JSIR2S performance in future studies. This approach contributes to ongoing efforts in the verification and validation of SDDR methodologies and highlights the role of research reactors such as TRIGA in supporting fusion technology development.

The Kingdom of Saudi Arabia (KSA) is now among the elite group of countries that will enjoy the benefits of nuclear reactor technology. The ideal approach to building a nuclear reactor for a desert region like KSA is to develop micronuclear power plants that can be installed instantly in remote regions. The performance of nuclear reactors, including their novel design, their safe operation, and their economic viability are the most important issues. Therefore, the proposed research is based on the novel design of reactor. This invention presents the conceptual design and thermal-hydraulic analysis of a 10 MW modular nuclear reactor employing light water as the primary coolant in a pressurized configuration. The design emphasizes modularity, inherent safety, and compactness, making it suitable for decentralized power generation, industrial heat applications, and deployment in regions with limited grid infrastructure. The reactor operates on natural circulation principles, supported by forced circulation on the secondary side. Cold water enters through the lower plenum, flows upward across the reactor core, and absorbs fission-generated heat. Reduced coolant density drives buoyancy flow toward the riser and helical coil steam generator, where heat is transferred to the secondary loop. Feed-water entering the steam generator undergoes phase transition, producing superheated steam at 579.8 K for turbine operation. This modular reactor offers technical viability for clean electricity, hydrogen production, desalination, and remote energy supply. By combining compact design, scalability, and validated thermal-hydraulic performance, it represents a competitive and sustainable nuclear energy solution. This innovation directly supports Saudi Arabia’s commitment to sustainable development, energy security, and carbon reduction goals. By introducing a viable small modular reactor concept, the Kingdom can advance its leadership in clean energy technologies while aligning with global trends toward carbon-neutral energy systems.

AC²-ATHLET is a thermal-hydraulic system code for simulating coolant circuits in nuclear reactor systems. In this work, the modeling of pressure losses is extended to non-circular flow geometries, such as rectangular ducts as well as concentric and eccentric annular ducts, and to hydrodynamically developing flows. This is relevant for compact reactor components, short flow paths, and small control volumes, where the assumption of fully developed flow may not be sufficient, as is often the case in modern reactor designs and small modular reactors.

For laminar flow, the implemented model is based on the scaling approach by Muzychka and Yovanovich (2009) for developing and fully developed flow in non-circular ducts. Instead of using only the hydraulic diameter, the square root of the cross-sectional area is used as the characteristic length. This allows for a more consistent treatment of different cross-sectional shapes. The local development of the flow is described by a dimensionless axial length, so that the pressure-loss correction varies along the duct and approaches the fully developed limit downstream.

The geometric description is based on the aspect ratio for rectangular ducts, on the radius ratio for annular ducts, and, in the case of eccentric annular ducts, on the eccentricity. In addition, geometry-dependent critical Reynolds numbers are introduced. They serve to distinguish the laminar regime from the transitional regime and thus form the basis for a consistent regime selection. For rectangular ducts, the approach by Tosun (1988) is used for this purpose, while concentric annular ducts are classified according to Hanks (1980) using radius-ratio-dependent thresholds. In addition, the approach proposed by Gnielinski (2007) is taken into account, in which geometric effects are incorporated into the pressure-loss calculation via a modified Reynolds number, with the Darcy friction factor calculated from the Muzychka and Yovanovich model serving to modify the Reynolds number.

The turbulent extension is integrated into the existing Colebrook-based part of the pressure-loss model. The extension therefore builds upon this computational path and supplements it with geometry-dependent corrections for fully developed non-circular flows, as well as an additional treatment of turbulent entrance flow.

RESULTS
For the laminar regime, a Darcy friction factor is determined for the geometry under consideration and the local state of development using the approach by Muzychka and Yovanovich (2009). The basis for this is the cross-sectional area, the respective geometric parameterization, and the dimensionless axial length. This allows both the entrance region and the fully developed limit case to be described. For rectangular ducts, the correction factor is determined using the aspect ratio. For annular ducts, the radius ratio and eccentricity are included in the calculation. The resulting pressure-loss factor is then used, in accordance with the Gnielinski approach (2007), to modify the Reynolds-number-based calculation.

The implemented laminar model was verified using literature correlations and reference data from the VDI Heat Atlas. For rectangular ducts, the calculated pressure-loss factors show very good agreement across the range of aspect ratios examined. The largest deviations occur near the parallel-plate limit, where the geometry becomes highly asymmetric. For annular ducts, the agreement is also good for concentric and moderately eccentric configurations. Larger deviations are to be expected for highly eccentric annuli, where the flow field becomes increasingly geometry-dependent.

For fully developed turbulent flows, the approach by Duan, Yovanovich, and Muzychka (2012) is used as the basis for further development. Analogous to the laminar model, the square root of the cross-sectional area is also used there as the characteristic length. This reduces the geometry dependence of the turbulent friction-factor formulation and enables a consistent transfer from circular to non-circular cross sections.

The turbulent entrance length is determined according to Anselmet et al. (2009). The approach describes the axial development of turbulent flow using a Reynolds-number-dependent development length based on the hydraulic diameter. In the implementation, this length is used to distinguish between the turbulent entrance region and the fully developed region.

For turbulent developing flows, there is no complete general correlation available for all non-circular geometries under consideration. Therefore, the pressure-loss correction in the entrance region must be approximated based on the axial development of the velocity profile. The intended approach is to use the developing velocity profile in the momentum balance and derive an effective wall shear stress from it. This wall shear stress can then be used to calculate an additional pressure-loss correction for the turbulent entrance region.

For annular turbulent flows, the work of Singh, Nigam, and Mishra (1980) provides a suitable reference. Their model describes fully developed and developing turbulent flows through annular ducts using separate velocity-profile descriptions for the inner and outer regions. Friction factors and pressure losses in the entrance region are derived from these profiles.

CONCLUSION
The extension of AC²-ATHLET enables pressure-loss calculations for rectangular as well as concentric and eccentric annular ducts, taking into account hydrodynamically developing flow. Previous simulations of the laminar model show good agreement with literature data, particularly with the VDI Heat Atlas and the underlying reference models. For the turbulent regime, suitable approaches for fully developed flow, entrance length, and annular ducts have been identified, which serve as the basis for further implementation.

With the rapid expansion of renewable energy sources such as solar and wind power, grid operators are increasingly required to manage greater variability in electricity demand and supply. To accommodate this shift, nuclear power plants (NPPs) are increasingly recommended to participate in load-following operation — a mode in which reactor power is deliberately varied in response to grid demand. However, load-following operation introduces complex neutron-physics challenges, particularly those associated with xenon transient behavior, which can significantly affect the core power distribution.

Xenon-135 (135Xe) is one of the most significant fission products in nuclear reactor operation due to its high neutron absorption cross-section. During load-following operation, reactor power is intentionally reduced and subsequently increased, causing the xenon concentration in the core to undergo substantial dynamic changes. Changes in reactor power during load-following operation result in significant xenon transients, which continue to influence the core power distribution even after the power maneuver is completed. This xenon-induced transient gives rise to significant spatial redistribution of the neutron flux and, consequently, the core power distribution. The resulting power distribution asymmetry can persist for several hours after the load-following maneuver is completed, during which the core power distribution remains in a transient condition. A key objective of this study is to determine the stabilization time required for the core power distribution to return to a steady-state condition following load-following operation. Understanding this stabilization time is important for planning load-following maneuvers and evaluating compliance with power-distribution-related operational limits.

This study analyzes core power distribution behavior during and after load-following operation for the APR1400, a pressurized water reactor (PWR) developed in South Korea.

To evaluate the impact of xenon transients on core power distribution, this study simulates and compares three sets of load-following scenarios. The first set examines the effect of different power ramp-up durations on subsequent xenon-induced power distribution distortion. The second set varies the reduced-power hold time, i.e., the duration for which the reactor is maintained at a lowered power level before power recovery. The third set considers different levels of power reduction during load-following operation. By varying these parameters, the study evaluates their effects on xenon-induced power distribution transients and the stabilization time required for the core power distribution to return to a steady-state condition.

The results of this study provide insight into the stabilization behavior of core power distribution following load-following operation in the APR1400 and improve understanding of xenon-induced power distribution transients under various load-following conditions.

The Work Package 1 of the EVEREST project revolves around the validation of modelling tools for vessel fluence prediction against experimental data gathered from the Siemens KWU/Vor-Konvoi. The modeling activities involve both in-core calculations for the determination of an effective neutron source and ex-core calculations tallying the fast flux at the reactor pressure vessel (RPV). The in-core low-fidelity (LF) modelling methodology adopted at EPFL consists of lattice calculations with SCALE Polaris and full core calculations with the nodal diffusion code PARCS. In agreement with results from literature, a tilt appears in the radial power distribution at hot zero power (HZP) and beginning of life (BOL) when the LF full core solution is compared to a high-fidelity (HF) one. More specifically, for the analyzed high-leakage core loading pattern, the LF methodology underestimates the power of the outermost assemblies and overestimates the power at the center of the core. For vessel fluence prediction, the correctness of the power distribution in the outermost assemblies is crucial to avoid systematic biases; therefore, efforts are made to improve the agreement between LF and HF solutions. Past studies have attributed this power tilt to the modelling of the radial reflector, which is hence the focus of this paper. This work explores different methodologies to model the radial reflectors and their potential impact on the neutron source used for vessel fluence calculations. The comparison encompasses conventional 1D reflector models developed using Polaris, together with Serpent 2 models constructed in both 1D and 2D geometries. Here, “1D” refers to a simplified reflector modelled as a series of slabs with a single adjacent fuel assembly, whereas “2D” refers to a quarter-core model representing the real core radial geometry. The impact of the application of the Serpent 2 Monte Carlo code is hence assessed both in terms of neutron transport methods and in terms of geometric fidelity. For the former, the possibility to use transport correction ratios (TCR) for hydrogen in the calculation of the diffusion coefficient in Serpent yields a non-negligible impact, when compared to the out-scatter approximation employed by Polaris. For the latter, Serpent allows the use of 2D models to generate the cross-sections of flat and corner reflectors. The generation of group constants comprises also the calculation of assembly discontinuity factors (ADFs). In order to calculate ADFs at the interfaces with the neighboring fuel assembly, a diffusion solver is applied to obtain the homogeneous flux. The need for the diffusion solver stems from the use of the null net current assumption by Serpent to tally the homogeneous surface flux, which is not applicable at the interface between assembly and reflector. In accordance with previous findings reported in the literature, specific attention is devoted to the development and application of a diffusion solver that is coherent with the one used by the nodal code simulator PARCS. This solver addresses a 2-group fixed source diffusion problem on the fuel assembly and reflector nodes. The homogeneous condensed cross sections and fission source are taken from Serpent. The net current across assembly-reflector interface is also calculated by Serpent and imposed as a boundary condition in the diffusion solver. Finally, the developed models are compared in terms of radial reflector cross-sections, diffusion coefficients and ADFs. Their impact on vessel fluence calculations is assessed through a comparison of the radial power distribution in PARCS with a reference full-core Monte Carlo solution. Assembly and pin power distributions directly influence the neutron source definition adopted in the subsequent shielding calculations.

This paper presents an overview of the temperature monitoring systems installed on spent fuel dry storage casks at the Krško Nuclear Power Plant. A comprehensive description of the hardware and software employed for temperature measurement and data acquisition is provided for each cask, with the primary objective of monitoring cask performance and status of dry storage building. Historical temperature trends were systematically evaluated and analyzed to assess the reliability and effectiveness of the passive cooling design. Furthermore, the paper describes preparatory activities associated with the second dry storage campaign and discusses considerations relevant to long-term spent fuel storage.

This paper explores solutions for modernizing Instrumentation and Control (I&C) in nuclear and military settings, with a focus on digitization. The goal is to replace outdated analog instruments and systems with digital technologies in a way that will minimize operational interruptions. The challenges associated with digitizing these complex systems can be addressed through two key approaches: the use of universal adapter plates for gradual meter replacement and the implementation of a modular compact I&C cabinet for comprehensive overhaul and both options compliant to Class 1E or commercial grade.

The adapter plate solution facilitates the transition from analog to digital meters by converting existing panel cutouts to accommodate a range of digital displays. This approach allows facilities to replace outdated meters without the need for a complete new control panel. These adapter plates offer the flexibility to perform upgrades incrementally, either as analog meters fail or during planned maintenance outages. The digital meters provide improved accuracy, reliability, longevity, and compatibility with modern Human-Machine Interface (HMI) and Machine-Machine Interface (MMI) systems. Additionally, converting all meter input signals to a standardized 4-20mA current loop simplifies maintenance by reducing the variety of costly spare parts required, enabling the use of a single meter type for various signals across the facility.

A more comprehensive solution is the implementation of a modular compact I&C cabinet, which centralizes the control of multiple I&C functions into a single, unified system. Such a cabinet allows for full or partial digitization of control systems without interfering with ongoing plant operations. This modular system can be installed in any available space with access to power and signal connections, and can function alongside or as a replacement for existing Supervisory Control and Data Acquisition (SCADA), Data Acquisition Systems (DAS), or Programmable Logic Controllers (PLC). A digital I&C cabinet supports both automated and manual control of process, enhancing flexibility in operation. A digital I&C cabinet would be required to be designed to house a large number of digital meters, controllers, and signal transmitters. Desired features include programmable logic transmitters and controllers that can manage signal outputs via various relay options including an OCT or MOSFETS or dry contact reel relay up to 4 amps for Hi, Hi/Hi, Low, and Low/Low alarms, and fail-safe mechanisms. These systems should be fully programmable and customizable to meet specific operation needs. The development and deployment of modular I&C cabinets have been informed by lessons learned from various industries, including nuclear power, aerospace, and military applications. These robust steel cabinets construction provides housing for up 400 digital meters, while redundant UPS and individual isolated power supplies and advanced monitoring systems ensure operation reliability. Sophisicated fail-safe capabilities are crucial for critical operations and must include mechanisms for controlled shutdowns during emergencies. This approach to digitizing I&C systems can be customized to meet the specific needs of both civilian and military applications. The system’s ability to consolidate control functions into a centralized digital cabinet eliminates the need for multiple scattered control panels, streamlining operations and improving overall system performance. Furthermore, the use of standard current loop signals simplifies maintenance and reduces long-term operational costs by extending the lifespan of the system and its components.

Over the past two decades, advanced multi-physics solvers have been developed primarily to reduce the conservative assumptions traditionally employed in the analysis of pressurized water reactors. Despite this progress, a major limitation persists: the scarcity of high-quality experimental data suitable for the validation of coupled neutronic–thermal-hydraulic simulations. The Training Reactor of the Budapest University of Technology and Economics (BME TR), a 100 kW pool-type research reactor operating under natural circulation, represents a unique validation platform within the EVEREST project. Its power level is sufficiently high to produce measurable thermal feedback effects, while its open-pool geometry enables flexible experimental access and instrumentation.

This work presents a high-fidelity coupled simulation framework developed for the BME TR, combining the OpenMC Monte Carlo neutron transport code with an in-house developed 2.5D finite element thermal-hydraulic solver. The work directly supports EVEREST’s objective of quantifying the impact of advanced multi-physics models on the prediction of Long Term Operation (LTO) relevant parameters in PWR and VVER reactors. The neutronics model is based on a detailed Constructive Solid Geometry representation converted from a validated MCNP reference model and incorporates 24 EK-10 fuel assemblies with assembly-specific burnup compositions, graphite reflector, control rods, and experimental irradiation channels. The thermal-hydraulic solver resolves two-dimensional cross-sectional heat conduction within fuel assemblies together with one-dimensional axial bulk energy transport, employing a dual-channel formulation separating the inner cassette flow from the outer bypass pool. Natural circulation velocities are obtained from an iterative buoyancy–friction balance, with geometry-specific friction factors derived from a two-dimensional viscous Poisson solve on the finite element mesh. The two codes are coupled through a Picard iteration scheme with axially resolved temperature and coolant density feedback applied per assembly and per axial layer.

Validation is performed against experimental datasets obtained within the EVEREST project. Thermocouple measurements at fuel assembly outlets across multiple reactor power levels are compared with calculated outlet temperatures for cold insertion and long irradiation transients. Additionally, axial reaction rate profiles from gold foil activation measurements at four radial core positions (E3, B5, F7, and E6) are compared with calculated Au-197 capture rates at different reactor power levels. A feedback decomposition analysis is additionally performed to quantify the separate contributions of Doppler broadening and coolant density reduction to the overall reactivity feedback, demonstrating that the two mechanisms dominate at different axial regions of the assembly.

Preliminary results demonstrate stable convergence behaviour of the coupled framework and good agreement between calculated and measured outlet temperatures across the investigated power range, while analysis of axial reaction rate distributions is ongoing. The work contributes to the EVEREST benchmarking effort and demonstrates the applicability of open-source high-fidelity tools for coupled reactor analysis in research reactor environments.

Climate change, environmental pollution and energy security are among the primary concerns in the twenty-first century. Rapid advancements in carbon emission mitigation are essential, as well as the mitigation of widespread plastic and microplastic pollution. Nuclear power represents a safe, reliable and carbon-neutral base-load energy source. A byproduct of the exploitation of nuclear energy is the ionizing nuclear radiation emitted during power plant operation or by high-level radioactive waste. Recently, the Jožef Stefan Institute (JSI) and the National Institute of Chemistry in Slovenia, have initiated collaboration efforts aimed at the investigation of nuclear radiation as a catalyst for efficient carbon capture and the synthesis of high-value chemicals from abundant, low-value feedstocks such as CO₂, plastics and biomass derivatives. This initiative was motivated by past irradiation experiments carried out at the JSI TRIGA reactor which have shown gamma-driven conversion of waste glycerol into acetol (a solvent) and solketal (a fuel additive). New radiolytic pathways could offer promising routes to activate stable molecules like CO₂, convert plastic and biomass wastes, and co-produce chemicals and fuels alongside electricity.

The JSI in collaboration with the French Alternative Energies and Atomic Energy Commission (CEA) has developed STARSHIP, a high-temperature irradiation device for testing and qualification of sensors (e.g. fission chambers, Rayleigh optical-fibre detectors) and materials for Generation IV fission and fusion systems (including molten-salt compatibility). The STARSHIP device consists of a tube furnace installed in the Tangential Channel of the JSI TRIGA reactor, enabling an internal target temperature range of 500 °C – 900 °C. The irradiation volume is cylindrical, approximately 40 mm in diameter and 600 mm in length; the maximal neutron flux and gamma air kerma rate are, respectively 2.7×10¹² n cm^-2 s^-1 and 2.1×10⁴ Gy h^-1 and typical irradiation times are of the order of hours. The device employs low-activation materials, including SiC heating elements and advanced insulating layers (nanoporous silica and Ca₂SiO₄). The device design, based on thermal analyses and material irradiation testing, as well as the device assembly, laboratory qualification, and reactor commissioning have been completed. Initial tests have demonstrated temperatures up to 870 °C.

This paper presents the development and initial testing of a coupled system, including the STARSHIP device located in the Tangential Channel of the JSI TRIGA reactor and a custom mass-spectrometry setup currently under construction, enabling online, real-time monitoring of gaseous and volatile radiolysis products directly during irradiation in controlled temperature conditions. Sample irradiations will be performed within a sealed stainless steel capsule located within the STARSHIP device, approximately 24 mm in diameter and 50 mm in length; a carrier gas (He) flowing through the capsule will be used to convey reaction products for analysis. The coupled setup will allow detailed kinetic studies, product-yield quantification, and mechanistic insights into radiation-driven transformations of CO₂, waste plastics, biomass, and other feedstocks under conditions relevant for industrial applications.

The Dispatchable Adaptive Reactor With Interchangeable compoNents (DARWIN) offers a flexible solution designed for a wide range of urgent and emergency applications, including flood water pumping, desalination, and district heating and cooling. Rather than optimising a reactor core for a single purpose, the DARWIN concept relies on interchangeable modules that can be configured to meet specific operational requirements, making versatility a core design principle.

The current work focuses on reactor core design using light water as coolant and fuels relevant to light water reactor technology, namely UO₂, MOX, UZrH_X, and UYH_X, arranged in a hexagonal lattice. Neutronics calculations are performed using a two-dimensional infinite geometry model in the Serpent Monte Carlo code. The infinite geometry approximation is adopted as a preliminary screening approach, enabling rapid evaluation of a large candidate design space.

To efficiently explore the large candidate design space, a genetic algorithm is used to optimise the reactor core design for a specific application. Core design parameters including fuel pin pitch, fuel composition, enrichment level, and assembly geometry are encoded as genes. These genes define an initial population of candidate designs. Each fuel type is treated as a discrete variable within the gene encoding, allowing the algorithm to explore different fuel types within a single optimisation run. Through successive generations of crossover and mutation, the population evolves toward an optimal solution, where the fitness function is defined as the maximisation of fuel cycle length. Strictly negative temperature reactivity coefficients for both fuel and moderator are enforced as a hard safety constraint throughout the optimisation process. Results will include comparisons of fuel cycle performance across candidate fuel types and identification of optimal assembly configurations.

The analysis of severe accident scenarios in nuclear reactors relies heavily on numerical simulations to understand complex thermodynamic phenomena. In the late phase of such accidents, debris beds can form from corium either within the reactor vessel or in ex-vessel configurations. The effective cooling of these debris beds is a key safety objective, as insufficient heat removal may lead to further degradation, melting or in the worst case, failure of the containment. Therefore, accurate simulation tools are essential for assessing cooling behaviour under a wide range of conditions. However, these simulations are computationally expensive and often impractical for large parametric studies requiring large amounts of data. This limits studies into uncertainty quantification, sensitivity analyses, or real-time applications.

The COCOMO (COrium COolability MOdel) code, developed by the Institute of Nuclear Technology and Energy Systems (IKE) at the University of Stuttgart, is used to simulate the thermal-hydraulic behaviour of debris beds, capturing key physical processes such as heat transfer, phase change, and flow through porous media. While COCOMO provides detailed and accurate calculations of debris bed cooling, the computational cost associated with large numbers of simulations limits its applicability in broader safety studies.

Recent advances in artificial intelligence (AI), particularly in deep learning, offer a promising alternative for accelerating such analyses. AI methods are well-suited for identifying complex, non-linear relationships in large datasets and have demonstrated strong performance in spatiotemporal prediction tasks. This work follows up on contributions presented at NENE 2024, which established AI-based approaches for predicting the quench front of simplified cylindrical geometries and NENE 2025, where the coolability and end state of more complex truncated conical debris beds were predicted. The present work expands on the data from NENE 2025, by focusing on not only the end state but the spatial and temporal evolution of the cooling process itself.

Central to this work is the prediction of spatially resolved quench and melt front maps. Understanding where and in what order a debris bed quenches or where melt forms is of direct safety relevance. The spatial progression of the quench front determines whether sufficient cooling is established throughout the bed, or whether dry, uncooled regions persist and risk further melting or structural degradation. A bed that quenches rapidly and uniformly indicates effective coolability, while delayed or localised quenching may signal insufficient heat removal, leading to hot spots and re-melting in certain regions. By predicting the time at which each spatial location reaches saturation temperature (indicating quenching), or a melting temperature, the model produces an interpretable map of cooling progression across the debris bed. This formulation compresses the full transient evolution into a single 2D field, enabling rapid assessment of cooling behaviour and potential risk locations without the need for computationally expensive full simulations.

This work uses a comprehensive dataset of approximately 9,000 ex-vessel, reactor-scale COCOMO simulations. Each simulation is defined by a set of uniformly distributed input parameters, including corium composition, particle size, porosity, system pressure, initial bed temperature, and geometric characteristics of the debris bed; of which parameters are used as inputs to the model. Further information is given to the model, such as the porosity map of the bed, which is also used as a mask to prevent predictions outside the bed region. To enable efficient learning across varying geometries, all simulations are interpolated onto a common spatial grid, allowing the application of convolutional neural network (CNN) based architectures.

To Summarise, the paper will present the results of these models, showing the potential of AI-based surrogate models to significantly reduce computational effort while maintaining acceptable predictive accuracy. Such models could support faster parametric studies, uncertainty quantification, and, in the long term, real-time decision support in severe accident management.

Heat pipe-cooled micro-reactors are receiving renewed interest driven by the passive cooling capabilities of heat pipes. These systems offer highly safe operations due to their self-regulating, passive nature and the elimination of a traditional bulk coolant, therefore, removing any possibility of a loss-of-coolant accident (LOCA) or loss-of-flow accident (LOFA) [1]. However, achieving a safe and reliable integration of heat pipes presents distinct engineering challenges. One important challenge among these is establishing an optimized internal design capable of matching the high thermal load (4.1-7.5 kW per heat pipe) and high temperature demands (627–727 °C) of proposed reactor concepts [2]. Consequently, this necessitates the utilization of high-temperature heat pipes operating with liquid metals such as potassium, sodium, and lithium.

To develop a successful optimization strategy, additive manufacturing (AM) serves as a highly viable fabrication option, enabling complex heat pipe geometries that would be exceedingly challenging to produce conventionally. This capability bypasses standard geometric and dimensional constraints while ensuring reliable replicability, facilitating systematic design optimization through iterative experimental refinement. However, the feasibility of additive manufacturing for high-temperature heat pipes, as well as the performance and structural characteristics of such components require detailed experimental validation.

This work investigates an AM-fabricated heat pipe featuring an innovative, strut-based capillary wick, based on a 0.5 mm body-centered cubic (BCC) unit cell. The samples were fabricated in 150 mm sections in length with a 20 mm outer diameter, using selective laser melting process of Inconel 718 at the Material Testing Institute (MPA). Subsequent manufacturing steps and experimental testing were conducted at the Institute of Nuclear Technology and Energy Systems (IKE). The sections were joined by electron beam welding process into a 90 cm heat pipe. The wick characteristics were evaluated using a mass-based rate-of-rise (mRoR) method with low-contact angle working fluids to ensure complete wetting that decouples the contact angle from the capillary pressure. Based on the solution proposed by Elkholy et al. [3], the effective pore radius and permeability were evaluated to be 160 µm and 374 µm2. The resulting empirical data were utilized to analytically calculate the operational limits of the AM-HPs. Performance and feasibility were further assessed through leakage tests, activation tests and steady state operation at various power levels using potassium as the working fluid. This work demonstrates the viability of additively manufactured heat pipes in high-temperature environments and establishes a foundational framework for systematic optimization to support the development of the advanced nuclear micro-reactors.

Reference

[1] I. Yilgor, Z. D. Sellers, J. L. Hartvigsen, P. Sabharwall, K. M. Sweetland, Heat pipe cooled microreactors, Tech. rep., Idaho National Laboratory (INL), Idaho Falls, ID (United States) (12 2023). doi:10.2172/2337608. URL https://www.osti.gov/biblio/2337608.

[2] J. W. Sterbentz, J. E. Werner, A. J. Hummel, J. C. Kennedy, R. C. O’Brien, A. M. Dion, R. N. Wright, K. P. Ananth, Preliminary assessment of two alternative core design concepts for the special purpose reactor. doi:10.2172/1413987.

[3] A. Elkholy, M. Bardoel, J. Durfee, and R. Kempers, "A mass rate-of-rise model for additively manufactured wick structures," International Communications in Heat and Mass Transfer, vol. 146, p. 106934, 2023, doi: 10.1016/j.icheatmasstransfer.2023.106934.

Nuclear experiments in research reactors require more than measurement functionality. They require experimental devices that can prove, continuously and independently, that operation remains safe. In practical nuclear environments, the critical risk is not only that a parameter exceeds a limit, but that the experimental system continues operating when safety information is missing, unreliable, or provided by only one source.

In TRIGA research reactors, argon extraction systems are used to continuously remove air from irradiation channels and reactor cavities, since atmospheric argon can become neutron-activated during reactor operation. In the presented work, this existing argon extraction infrastructure was additionally used as an independent safety supervision layer by integrating smoke detection, gas monitoring, and thermal monitoring directly into the extraction system itself.

This work presents an independent safety monitoring system integrated into the argon extraction infrastructure of the STARSHIP high-temperature irradiation device connected to TRIGA reactor irradiation-channel operation.

The central design principle is simple: the experiment is allowed to operate only when safety is actively confirmed. If valid safety data from the argon subsystem are not received by the main control system, operation is inhibited. In this architecture, missing safety data are treated as an unsafe state rather than as a minor communication fault.

The argon safety subsystem includes optical smoke detection, MQ-2 and MQ-135 gas sensors for monitoring changes in gas composition, and six thermocouples for distributed temperature measurement across the device. These measurements are processed by the argon system and exposed remotely through an application programming interface, allowing the main experimental control system to continuously verify safety status.

This creates an independent safety permission layer. The main experiment does not rely only on its own controller, operator observation, or on a single measurement channel. Instead, it must continuously receive confirmation from a separate subsystem responsible for atmospheric, smoke, and thermal supervision. If this confirmation is lost, abnormal, or unavailable, the experiment cannot proceed as if conditions were normal.

The practical contribution of this work is the transformation of automation from a convenience feature into an enforceable safety mechanism. Remote monitoring, API-based supervision, redundant sensing, and communication-dependent permission logic allow the system to reduce uncertainty during operation. The result is not merely a device that can be controlled remotely, but a device that refuses to operate when safety cannot be demonstrated.

This approach is relevant for experimental nuclear facilities, irradiation laboratories, and developers of reactor instrumentation where complex devices must operate reliably in restricted-access environments. The presented system demonstrates a practical method for designing nuclear experimental hardware around a fail-safe principle: no single sensor, controller, communication link, or operator action should be the only barrier between normal operation and an unsafe condition.

During the reprocessing of spent nuclear fuel (PUREX process, operated at the Orano Recyclage La Hague plant), nitric acid solution containing fission products (FP) are stored in tanks prior to their conditioning for final disposal (by means of vitrification process). The significant heat release generated by the radioactive decay of the fission products is offset by a cooling system. In the event of a total loss of cooling, the solution could heat up to its boiling point, potentially leading to the release of fission products into the tank ventilation system, including gaseous ruthenium in the form of RuO₄ (a scenario of interest following the post-Fukushima stress tests). In addition to its ability to form highly volatile oxides, ruthenium exhibits significant radiotoxicity due to its radioactive isotopes ¹⁰⁶Ru and ¹⁰³Ru, with respective half-lives of almost one year and 40 days. It should be noted that in the event of a loss-of-cooling accident affecting fission product storage tanks, without restoration of the cooling, the peak release of ruthenium would be observed after approximately three days (Philippe et al., Behavior of ruthenium in the case of shutdown of the cooling system of HLLW storage tanks », CEA Centre d’Etudes Nucléaires de Fontenay-aux-Roses, 1990).

This phenomenology is still uncertain and needs further insight (Nerisson et al., Behaviour of ruthenium in nitric media (HLLW) in reprocessing plants: a review and some perspectives, Journal of Radioanalytical and Nuclear Chemistry, 2022). This need is addressed within the framework of a Joint Interest Project Orano–ASNR (formerly IRSN) launched in 2021. Among different topics, one issue is to better understand the physicochemical phenomena governing the interactions between ruthenium and nitrogen oxides (NO_x) released from the boiling nitric acid solution, taking surface effects into account. To this end, gas-phase reactivity is analytically investigated in considering possible catalytic effects induced by steel (304L steel surfaces simulating ventilation ducts, with or without prior ruthenium deposits).

To study the reactivity of gaseous RuO₄ batch under different atmospheres, a device consisting of a glass reactor and a robust sampling system was designed. The reactor consists in a 3 L double-walled chamber in which RuO₄(g), generated by ozonation of RuO₂·xH₂O powder, is injected and brought into contact with steel coupons and/or NO_x.

The first tests on the decomposition of gaseous RuO₄ were conducted at 90 °C, both with and without 304L stainless steel coupons. The temperature was chosen to make comparison with previous results and allows us to get appropriate kinetics of reaction to our scale time study.

At 90 °C and in the presence of stainless steel (surface area of 2500 mm²), it has been shown, from the literature, that the decomposition kinetics of RuO₄ exhibit two distinct stages: an induction period followed by an autocatalytic process, with respective rate constants of 7.0 × 10⁻⁶ s⁻¹ and 2.3 × 10⁻⁵ s⁻¹. Our first results do not distinguish these two stages, with a first-order decomposition rate of 5.0× 10⁻⁶, 7.8 × 10⁻⁶, and 3.2 × 10⁻⁵ s⁻¹ for 0, 424, and 1272 mm² of available stainless steel area, respectively, within a gas volume of approximately 3 L. The steel surface promotes the decomposition leading to a shorter half-life of gaseous RuO₄. This is likely due to the increase of adsorption/condensation sites, since the decomposition of gaseous RuO₄ into solid RuO₂ is not congruent with catalytic process. The presence of 1272 mm² of available steel surface leads to a metallic-like deposit on the reactor glass surface, which cannot be removed using the usual cleaning solution (1 M NaOH with 10–20 g·L⁻¹ K₂S₂O₈). However, the formation of a metallic ruthenium deposit at this temperature is highly unlikely, and XRD analyses indicate a RuO₂ deposit (X-ray Photoelectron Spectroscopy analyses pending).

Regarding the impact of the presence of NO/NO₂ on the decomposition kinetics of RuO₄, the results indicate a first-order decomposition rate of 1.9 × 10⁻⁶ s⁻¹. The presence of nitrogen oxides leads to a significant increase in the half-life of gaseous RuO₄, which is consistent with the literature reporting that NO_x species help stabilize the tetroxide in the gas phase by inhibiting its decomposition (Yoshida et al., Effect of nitrogen oxides on decomposition behavior of gaseous ruthenium tetroxide, Japan Atomic Energy Agency, 2020) (Hessou et al., Ab initio investigation of the competitive adsorption of RuO₄ and NO_x on a RuO₂ surface: Assessment of RuO₄ release in a severe nuclear accident, The Journal of Physical Chemistry C, 2025).

Future experiments will aim to evaluate the combined effects of the presence of nitrogen oxides and stainless steel at 90 °C, as well as to investigate the influence of temperature on the decomposition kinetics of RuO₄(g) through a test conducted at a lower temperature (50 °C).

Chemical reagent injection and radwaste decomposition (CRI_RWDecom) was developed by Korea Hydro & Nuclear Power (KHNP) as a full system decontamination process for pressurized water reactors (PWRs). This process sequentially employs oxidation, reduction, decomposition, waste treatment, and purification. It uses permanganic acid as an oxidizing agent and oxalic acid as a reducing agent to dissolve metal oxides during system decontamination. The decontaminated liquid waste contains radionuclides (Co²⁺, Cs⁺, and Sr²⁺), dissolved metal ions (Cr³⁺, Fe²⁺, Ni²⁺), and residual decontamination agents (Mn²⁺, NO₃^-, HC₂O₄^-).

For the treatment of this waste, a UV-C based Photo Fenton process is employed to decompose approximately 90% of the oxalic acid. Subsequently, the remaining ions are polished using an ion exchange process. Although mixed-bed ion exchange process is usually used due to high effluent quality, resin regeneration requires complex separation steps, which often leads to direct disposal and increases disposal cost.

This study investigated the removal mechanisms of CRI_RWDecom liquid waste using strong acid cation (IRN 77) and strong base anion (IRN 78) exchange resins. These resins were evaluated under single-bed, sequential separated-bed, and mixed-bed configurations. The effects of pH variation on ionic speciation were analyzed to determine underlying mechanisms, removal efficiencies, and optimal bed order. High concentrations of sulfate in the waste formed neutral and charged complexes, which compromised oxalic acid treatment efficiency. Furthermore, the speciation shift from oxalic acid to hydrogen oxalate or oxalate, induced the formation of colloidal oxalate complexes.

Based on a comparative performance analysis and resin regeneration feasibility, a sequential cation-to-anion configuration with a 1:1 theoretical requirement of cation and anion exchange resins capacity removed 92.4% of oxalic acid and more than 99% of the remaining metal ions within 10 minutes. This optimized sequence achieved a superior acid-base balance and minimized competitive binding from sulfate and nitrate ions. Therefore, this configuration is proposed to enhance polishing efficiency, simplify resin regeneration, and reduce disposal costs in CRI_RWDecom wastewater treatment systems.

The geopolymer is a new environment-friendly material that represents an alternative to cementitious waste form for the immobilization of radioactive liquid organic waste (RLOW) by combining better chemical durability with lower environmental impact.

The aim of this paper is to investigate the physical-chemical, microstructural and mechanical properties of three alkali-activated geopolymer formulations, constituted of metakaolin, blast furnace slag, and fly ash, which are the key parameters influencing and affecting the durability and the stability of the conditioned waste in the long-term period (disposal conditions).

Thermal degradation test, performed at 100° for 48 hours without interruption, XRD, thermogravimetric analysis (TGA), compression tests, and scanning electron microscopy with energy-dispersive spectroscopy (SEM-EDS) were carried out to assess geopolymer durability and stability.

SEM observations reveal a heterogeneous microstructure characterized by spherical particles, porous regions, and a fine amorphous matrix, while EDS analyses confirm a Si–Al–O dominated composition with minor Ca, K and Fe contributions. XRPD indicates the coexistence of crystalline phases and a significant amorphous fraction.

This study is performed in the framework of EURAD-2 project, geopolymer characterization is a primary focus for Work Package 7 (WP7) "L'OPERA", which evaluates advanced matrices for immobilizing low- and intermediate-level radioactive waste (LILW).

Flow boiling is an important mechanism of heat removal in nuclear reactors, effectively removing large amounts of heat from a heated component and preventing surface damage due to overheating. The study and prediction of this multiphase flow phenomenon is therefore important in determining the safety margins and operating conditions of light water reactors. The present work uses the Morphology-adaptive Multifield Euler-Euler (MultiMorph) model within the OpenFOAM Foundation software to model the THELMA (Thermal Hydraulics Experimental Laboratory for Multiphase Applications) flow boiling experiment, depicting flow boiling around a reactor fuel rod.

The MultiMorph model combines a modelled approach for the interface of flow structures too small to be resolved on the grid and a resolved approach for the interface of large structures resolvable with the grid at hand. Therefore, it distinguishes a dispersed and continuous morphology for each phase, using multiple numerical phases with appropriate closure laws to describe the interaction and the transition between different representations.

The THELMA flow boiling experiment is designed as a double pipe heat exchanger, where a copper pipe with an outer diameter of 12 mm is incased in a larger borosilicate glass tube with an inner and outer diameter of 16 mm and 24 mm, respectively. This creates a 2 mm wide annular gap between the two, through which a refrigerant (R245fa) flows. The copper pipe is heated from the inside by flowing hot water, causing the refrigerant in the gap to boil, which is then recorded with high-speed cameras. Alongside this, thermo couples on the inside of the copper pipe enable the measurements of the axial temperature of the wall and flowing hot water.

Present work considers a three-dimensional slice with a small angle of the annular gap and the accompanying copper wall for simulation. Due to the vertical orientation axial symmetry is assumed. A multiphase conjugate heat transfer simulation of the boiling refrigerant in the annular gap is carried out. The heat losses to the environment through the glass tube and the heat gains from flowing hot water are modelled accordingly. Operating conditions at a system pressure of 2.6 bar and refrigerant mass flux of 600 kg/m²s with an inlet subcooling of 14 °C are considered. Two hot-water inlet temperatures are selected: 50 °C, representing a lower-temperature case with only modest bubble growth and coalescence, and 60 °C, where small nucleating bubbles rapidly grow and merge into larger vapor structures. This transition in bubble morphology is then represented using the MultiMorph model. Simulations with differing mesh resolutions are performed to investigate the effects on morphology transfer between the dispersed and continuous gas phases. Results are validated with experimental data and compared with previous two-fluid model (dispersed gas only) simulations.

As a rule, regulatory inspections of nuclear and radiation facilities as well as other activities involving radiation sources require highly specialised inspectors. Risks associated with such practices span from very high risks, e.g. related to nuclear reactors, industrial radiography and radiotherapy, to activities with very low risk e.g. a use of Ni-63 in ECD in quality assurance laboratories. Therefore, the technical measures to be inspected, along with the associated organizational measures, may be highly complex at one site while remaining basic at another. In addition, the facilities with highly specialized designs and novel practices that are the first of their kind, either within a State or worldwide, make inspectors’ tasks even more challenging. Such facilities may include research accelerators, sterilization units, radioactive waste storage facilities, or facilities producing new radioisotopes. Furthermore, certain activities may possess unique characteristics, for example the first decommissioning of a university research laboratory involving open radioactive sources. Moreover, each inspector is confronted at one point of the time to inspect a practice with he or she never inspected before and therefore he or she has limited on site experiences with such practice.

Inspection is as stated by the IAEA, one of the core regulatory processes. Assuring competences of inspectors when inspecting such first of that kind facilities or practices, e.g. using new types of spectrometers, requires systematic approach. This is a particular challenge in countries with very limited number of regulatory staff. Training should address newcomers, i.e. persons qualified to become a member of the regulatory staff who go through elaborate training period to gain expertise in line with graded approach including on-job training programmes. Systematic approach to training programmes is particularly important when first of that kind facilities or practices are involved.

The Slovenian Nuclear Safety Administration (SNSA) gained experiences with building needed competences when addressing first of that kind practices or activities. The process of building competences has five steps, namely:

• Preparatory phase with preparation of the Action plan for gaining competences and identification of team to be involved in inspection process;
• Identification of other competent authorities, technical support organizations in a state and literature screening;
• Cooperation with investor, supplier, future operator, other national regulatory bodies in particular those from the state of origin and others;
• Management of regulatory processes with due vigilance including documenting all technical questions and answers;
• Continue learning and upgrading the Action plan.

As inspection of first of that kind practice is a challenge such systematic approach is needed. All five steps should reflect graded approach to safety and security. The steps are described in details and practical hints how to gain competences are given. The method is applicable also for other regulatory processes. TSOs as well as others involved in nuclear and radiation area might adopt it as needed. In addition, such approach also result in building a strong national network of highly competent professionals. Such network is the best guaranty that safety and security challenges in nuclear and radiation safety are going to be addressed also in the future.

In large tokamaks such as ITER, powerful neutron sources and substantial positioning
devices are required for in-vessel calibration. Changes in geometry such as the presence
of a remote handling system as well as differences in neutron emission properties between
the calibration source and the plasma, can be compensated for with particle transport
calculations. Furthermore, computational methods enable us to calculate the detector
response in regions where experiments have not been performed, due to restrictions on
device commissioning time. Because of this, sensitivity studies are required to estimate
the impact of the uncertainties in nuclear data, geometry, and materials on the uncertainty
in particle transport calculations. Simple, representative models are required for such
studies both for computational efficiency and to illuminate underlying physical
phenomena.

In this work, we present the creation of a simplified neutronics model of the ITER tokamak
intended to support the development of the neutron calibration strategy, referred to as the
Neutron Calibration Simplified Model (NCSM). This Constructive Solid Geometry (CSG)
model is axially symmetric and has been developed using the Polygon Torus tool [1] where
planar, conical, and cylindrical surfaces are used to approximate toroidal bodies. Its
geometry and materials are based on state-of-the-art detailed ITER neutronics models. The
poloidal section is based on published images of the C-Model, while the material
composition was derived from the E-lite model of the ITER tokamak.

We compare the NCSM and the C-Model in terms of cell and surface counts,
computational performance in MCNP and OpenMC, and detector signals along a toroidal
source scan, observing a 900x reduction in initialization times and a 540x increase in
particle per second computational efficiency with the agreement within 5.2 % for the
integral reaction rate in the Neutron Activation System (NAS) during a toroidal scan with a
neutron source. This combination of geometric representativeness and computational
speed enables uncertainty and sensitivity analyses (e.g. Total-Monte-Carlo with O(10³)
samples) that are inaccessible with the reference detailed models, providing a practical
tool for the iterative development of the ITER neutron calibration strategy.

[1] Fortuna, M. (2025). PolygonTorus: A tool for creating axially symmetric CSG geometry. (Version 0.0.1) [Computer software]. https://github.com/mark-fortuna/polygontorus

Most fission reactors use water as a coolant, and it will also play an important role as a coolant in future fusion reactors. During the fusion of deuterium and tritium nuclei, high-energy neutrons with an energy of 14.1 MeV are produced, activating oxygen isotopes in water. Through the reaction ¹⁷O(n,p)¹⁷N, the isotope ¹⁷N is formed. Via radioactive decay, it emits gamma rays and secondary neutrons with discrete energies of 0.38 MeV, 1.17 MeV, and 1.70 MeV. In future fusion reactors such as ITER, the fusion neutrons will activate the water in the cooling system. Since the water activation is expected to be high, the cooling system requires careful shielding design and appropriate radiation protection considerations. A thorough understanding of the water activation process is therefore essential for the development, licensing, and safe operation of fusion devices. However, experiments on water activation under fusion-relevant conditions remain limited. For studying water activation, improving simulation tools, and directly supporting the ITER project, the KATANA irradiation facility with an activated water loop was established in 2024 at the Jožef Stefan Institute TRIGA reactor in Slovenia.

The characterisation of the neutron field intensity of the KATANA water activation facility was carried out using the neutron activation method with various activation materials, including indium samples. During the irradiation of samples, the TRIGA reactor operated at a constant thermal power of 250 kW, with a water flow rate of 0.681 L/s through the KATANA loop. The main focus was on the inelastic neutron scattering reaction ¹¹⁵In(n,n')^115mIn as its cross section is highest at the energies of the emitted secondary neutrons. In addition, this is a threshold reaction (339 keV), which eliminate the contribution of thermal and most epithermal neutrons. The metastable isotope ^115mIn decays with a half-life of 4.48 h, emitting gamma rays with an energy of 336 keV. After the irradiation on the loop, the activated samples were measured using an HPGe detector, and the reaction rate for the production of the isotope ^115mIn via inelastic scattering was experimentally determined from the number of detected gamma rays at the corresponding energy.

The geometric setup of the irradiation experiment was simulated using Monte Carlo N-Particle (MCNP) code. In the MCNP model of the KATANA outer irradiation part, the neutron flux from the activated water is represented as an isotropically distributed neutron source with discrete neutron energies. Through simulation, the neutron flux and the activation reaction rate in the samples were calculated. By comparing the reaction rates in the indium samples determined experimentally with those obtained from simulations, the neutron source strength of the KATANA was calculated. In addition, the neutron flux and reaction rate maps were also calculated using a mesh tally. The mesh tally results enabled optimisation of the geometric properties of the samples, particularly their thickness, for future measurements.

Heavy metal industry is demanding industry requiring highly-skilled jobs. It demonstrates growth globally. This trend is going to continue. A use of radioactive sources and in particular nuclear material in heavy metal industry is taking place for decades, e.g. for measuring thickness of the stainless stell in metalworking process using Cs-137 sources with activity of several TBq. In addition, heavy metal industry is also using several other types of radiation sources, e.g. X-ray generators with maximum voltage of several hundred kV resulting in dose rates up to Sv/h used in its production process, radiation sources used in quality assurance laboratory and X-ray spectrometers used for analysing input materials.

A use of sources associated with high risk requires due attention of regulators. Safety measures including safeguard control when using nuclear material should be in place.

The Slovenian Nuclear Safety Administration (SNSA) performs control of such practices in Slovenia in line with standards set in European Union (EU) in particular implementation of 2013/59/Euratom which provides basic safety standards to be applied in all EU Member States. The IAEA standards and Code of Conduct on the Safety and Security of Radioactive Sources and its supplementary guidance are followed. Inspections includes safety as well as security measures.

Regulatory inspections of high activity sources in Slovenia is taking place annually as a part of the SNSA Annual Inspection Plan. Analysis of inspector findings from such regular inspections is given. The analysis addresses normal operation of heavy metal facilities as well as emergency preparedness as incidents, accidents and near misses with radiation sources in heavy metal industry cannot be ruled out. Special attention of SNSA is given to such events.

Three events involving high activity sources with Cs-137 used in heavy metal industry which happened in the past are described in detail. It should be stressed that such events are relatively poorly described in literature. Operator activities, competent authority actions and a control provided by qualified experts are described providing detailed descriptions of challenges related to nuclear safety in heavy metal industry. Lessons learned to be used in the future are given.

The main objective of the study was to perform an uncertainty and sensitivity analysis of a doubled-ended large-break loss-of-coolant accident (LBLOCA) in a two-loop pressurized water reactor (PWR) of the Westinghouse type with a thermal power of 2000 MW, using the latest RELAP5/MOD3.3 Patch 6 thermal-hydraulic computer code. A detailed and validated input model of two-loop PWR was used for the RELAP5 computer code. The second objective of the study was to apply the Morris screening method for the selection of influential uncertain input parameters before uncertainty analysis and to use the obtained Morris sensitivity indices for comparison with the results of the sensitivity analysis performed after the uncertainty analysis using the Pearson and Spearman methods. For the uncertainty and sensitivity analysis, the latest version of the SNAP (Symbolic Nuclear Analysis Package) software and the uncertainty plugin using DAKOTA (Design Analysis Kit for Optimization and Terascale Applications) were used.

The scenario selected was double-ended LBLOCA in the cold leg. In accordance with the 10CFR50, Appendix K, discharge coefficients between 0.6 and 1.0 had to be considered for the Moody critical flow model. For the selected two-loop PWR, previous analyses showed that the highest cladding temperatures occurred at a discharge coefficient value of 0.4 (or even lower, but this is not in accordance with 10CFR50, Appendix K). Therefore, the spectrum of breaks with discharge coefficients between 0.2 and 1.0 was studied. Based on the maximum cladding temperatures break size of 0.4 was selected as reference case for uncertainty and sensitivity analysis. A total of 168 uncertainty runs were performed, during which 16 uncertain input parameters were randomly sampled. The input uncertain parameters with their distributions and ranges were determined based on the literature. For the calculation of input parameter influence the Morris screening method was used. It required an additional 85 simulations for each Morris sampling formula (simplified and optimal), where 16 uncertain input parameters were sampled according to the Morris method. The main result of the uncertainty analysis using SNAP uncertainty plugin was the peak cladding temperature of the hot fuel rod in the blowdown phase. Using an Excel spreadsheet, based on 168 uncertainty runs, the peak cladding temperature was determined both in the blowdown and reflood phase, for both the hot and the average fuel rod. Finally, Python script was developed and used to calculate statistical data in each time steps. In these way, trends of calculated statistical data were generated, similar to those that can be produced using Software for Uncertainty and Sensitivity Analyses (SUSA) tool, developed by Gesellschaft für Anlagen- und Reaktorsicherheit (GRS).

Introduction The global nuclear sector is undergoing a major paradigm shift, where science, engineering, and industry must align to meet the dual challenges of decarbonization and energy sovereignty. In France, this ecosystem is structured and driven by GIFEN (French Nuclear Industry Association). This abstract outlines the synergy between the French industrial fabric, the ambitious national nuclear program, and the strategic partnerships required to bridge scientific excellence and industrial performance. 1. GIFEN: The National Unified Voice of the French Nuclear Industry Founded to federate all stakeholders of the nuclear sector in France, GIFEN is the unique professional association gathering the entire value chain, from major utilities to start-ups, SMEs, mid-caps, and research organizations.

• Representativeness: Over 670 member companies spanning engineering, equipment manufacturing, construction, operations, maintenance, and the fuel cycle.
• Mission: To structure the industry, monitor and manage capacity building (notably through the MATCH and COACH programmes to forecast skills and workload requirements), and represent industrial interests to public authorities and international bodies.

2. The French Nuclear Industry and Program: A Historic Trajectory France has embarked on one of the largest industrial programs in its recent history, driven by the dual necessity of extending the lifespan of its existing fleet and building tomorrow's nuclear infrastructure.

• The "Grand Carénage": A massive maintenance and modernization program aimed at safely extending the operation of the current fleet (56 reactors) beyond 40 years under the highest safety standards.
• The New Nuclear Program: The rollout of the construction program for 6 to 14 EPR2 reactors, coupled with the development of a disruptive innovation ecosystem, specifically SMRs and AMRs.
• Industrial Excellence: A high-tech industry generating highly skilled jobs, backed by world-class R&D to ensure technological sovereignty.

3. GIFEN’s Partnership Strategy: Bridging Science, Innovation, and Industry To successfully deliver these unprecedented industrial challenges, GIFEN deploys an open, business-driven partnership strategy designed to strengthen the supply chain, harmonize standards, and secure international alliances.

• Industrial Synergies & Supply Chain Readiness: Fostering deep collaboration between major European developers and the wider supply chain to optimize industrial capacity, accelerate technology transfer, and implement advanced manufacturing processes (such as digitalization and modular construction).
• Strategic Visibility at NENE26: GIFEN is leveraging the NENE26 conference to establish a strong, high-visibility presence. This event serves as a key industrial platform to showcase the readiness of the French supply chain, share operational feedback, and engage with Slovenian partners on upcoming market opportunities.
• The GIFEN-SAF LoI: Capitalizing on the momentum of NENE26, GIFEN will officially sign a strategic partnership agreement with its counterpart, SAF. This milestone agreement aims to build concrete industrial bridges, promote mutual technical cooperation, and align supply chain capabilities to support the global nuclear resurgence.

Conclusion The success of the French nuclear program relies on an indivisible trilogy: the rigor of science, the capability of industry, and the strength of partnerships. GIFEN is fully committed to this dynamic, inviting the scientific community to co-design the reliable, safe, and innovative industrial solutions that will shape the future of energy."

The validation of advanced multi-physics simulation tools is a central objective of the EURATOM-funded EVEREST project (GA №101163288), requiring high-resolution experimental datasets. While previous research established the fundamental methodology for high-precision thermal mapping in the TRIGA Mark II reactor pool, this study details the subsequent advancements in the deployment process, technical qualification, and the regulatory framework required for installing experimental facilities within a nuclear environment.

A primary focus of these advancements is the technical refinement of the measurement hardware to enhance both spatial and temporal resolution. The deployment features a newly developed in-core probe and a pool-area support structure, both manufactured from aluminum to minimize neutron activation. Significant improvements were made to the sensor network by transitioning to thinner GG-K-36 Type K thermocouple wires (0.13 mm diameter), which reduces thermal mass and increases sensitivity to rapid fluctuations compared to previous designs that utilized thermocouples with a diameter of 0.5 mm. These sensors are now spot-welded rather than soldered to ensure mechanical stability and lower thermal inertia. Furthermore, the data acquisition system has been upgraded to support sampling rates exceeding 50 Hz, enabling measuring fast tremperature fluctuations as well as analysis of turbulent flow structures through temperature correlation techniques.

A critical component of the deployment process is the adherence to the RIC-QA-940 regulatory protocol, which governs all interventions and changes within the Reactor Infrastructure Center (RIC). The deployment can require multi-service approvals, including the Radiation Protection Service (SVPIS) for dose rate estimations, the Occupational Health and Safety Service (SVZD), and Quality Assurance (QA) to ensure compliance with the reactor’s Safety Report. The intervention will be classified based on whether it represented a minor, low-importance, or significant change (Categories 1-3), necessitating corresponding levels of notification to the Slovenian Nuclear Safety Administration (URSJV).

The deployment phase also included a pre-operational validation campaign. Extensive calibration tests were conducted in collaboration with the THELMA lab at JSI Reactor Engineering Division to verify the accuracy of the 32-channel thermocouple system across various sampling rates (1 Hz to 90 Hz) and temperature ranges (at room temperature and 80°C). These tests provided essential data on standard deviations and measurement reliability, ensuring that the instrumentation meets the benchmark quality standards required for the OECD NEA repositories.

Following the Fukushima Daiichi accident in 2011, Nuclear Power Plant Krško, a Westinghouse-designed PWR, implemented enhanced safety measures focused on FLEX (diverse and flexible coping strategies) to address beyond-design-basis events. FLEX strategies are designed to restore critical safety functions—core cooling, containment integrity, and spent fuel pool cooling—under extreme conditions involving prolonged loss of AC power and loss of ultimate heat sink.

This article presents an overview of the FLEX equipment deployed at NPP Krško in accordance with the plant specific FLEX Support Guidelines (FSG) (AESP). The displayed equipment includes portable diesel generators, mobile pumps, fuel supply units, and standardized connection interfaces that enable rapid and flexible response in severe scenarios. Emphasis is placed on equipment diversity, mobility, and protected storage, ensuring availability even under challenging external hazards such as seismic events or flooding.

Additionally, the article illustrates practical aspects of equipment deployment, connection strategies to existing NPP Krško systems, and integration into plant procedures. Operational readiness is supported through regular testing, maintenance, and targeted personnel training aligned with FSG implementation.

By visually presenting key FLEX equipment and its role within accident management strategies, this article highlights the strengthened resilience and defense-in-depth at NPP Krško while supporting knowledge exchange in the post-Fukushima nuclear safety environment.

Keywords: Nuclear power plant, critical safety functions, FLEX equipment, AESP.

INTRODUCTION AND MOTIVATION Monte Carlo neutron transport is the most physically rigorous method for reactor analysis, yet its adoption is constrained by computational cost. Existing production codes—OpenMC [2], MCNP, and Serpent 2 [3]—were designed for CPU clusters and cannot be practically restructured for GPU execution. MCNP traces its origins to 1963 and retains Fortran legacy code; OpenMC and Serpent 2, while more modern, use history-based transport loops where each CPU thread follows a single neutron from birth to death—a model that yields negligible GPU benefit due to thread divergence. More critically, restructuring the transport loop of any validated production code would invalidate its entire benchmark record, a risk no national laboratory will accept. The only viable path to a GPU-native transport code is to build one from scratch. Monte Carlo transport is an ideal GPU workload: neutrons are fully independent and the dominant operation— continuous-energy cross-section lookup—is memory-bandwidth-limited, not compute-limited. Modern GPUs deliver 3–10× higher memory bandwidth than server CPUs. Three recent developments now make a complete GPU implementation practical: consumer GPU memory has reached 8–24 GB (sufficient for full ENDF libraries on device); hardware BVH ray traversal is available via the Vulkan API; and Hamilton and Evans [1] demonstrated that sorting neutrons by material identifier before each kernel launch recovers cross-section cache coherence, yielding 3–10× physics kernel speedups at realistic material counts. GOAL AND TECHNICAL APPROACH ReactorSimulator is a Monte Carlo neutron transport simulator under active development toward production-grade capability, GPU-native by design and built on modern C++17 libraries. The intended capability mirrors OpenMC: continuous-energy transport, criticality, depletion, and variance reduction, running entirely on GPU hardware including AMD Instinct accelerators in Frontier-class supercomputers. Implemented physics includes elastic scattering with centre-of-mass frame kinematics, radiative capture (MT=102), and fission with ν¯ multiplicity sampling (MT=18), using ENDF-6 data parsed via njoy::ENDFtk into log-log interpolated tables of up to 65 536 bins per reaction. The pipeline is event-based: all neutrons advance through geometry intersection, cross-section lookup, and reaction sampling simultaneously, maintaining full GPU wavefront utilization. Geometry uses AMD RadeonRays (Vulkan BVH). Neutron buffers are shared between GPU and Vulkan via Windows KMT zero-copy handles, eliminating PCIe transfers during transport. COMPARISON WITH EXISTING CODES Table 1 summarises the architectural position of ReactorSimulator. GPU-native design cannot be added to a legacy code—it requires a ground-up architecture free of CPU-era constraints and validation debt. Table 1: Architectural comparison of Monte Carlo transport codes. ReactorSimulator OpenMC MCNP 6 Serpent 2 GPU-native architecture Yes No No No Event-based transport Yes No No No Modern C++ stack Yes Partial No No keff criticality Roadmap Yes Yes Yes Depletion / burnup Roadmap Yes Ltd Yes ROADMAP Current priorities are GPU power iteration for keff criticality, isotopic depletion via the CRAM solver, weightwindow variance reduction, and expansion of the isotope library beyond H-1 and Fe-56. Enabling the RDNA2 hardware ray-tracing units is expected to reduce the geometry step by 2–4×, and the material-sort optimisation a further 3–10× at realistic material counts. PRELIMINARY RESULTS A 1D diffusion benchmark was performed: 106 source neutrons transported through six alternating slabs of H-1 and Fe-56 with absorbing boundary conditions. The simulation completed in 669 ms on an AMD Radeon RX 6700S (2 304 stream processors, 224 GB/s memory bandwidth, 100 W TDP—a consumer mobile GPU with no hardware ray-tracing units active), achieving 101.5 million neutron-bounces per second. Geometry intersection and physics computation consumed 287 ms and 295 ms respectively; host transfer overhead was 11.5 % of total runtime. The measured absorption fraction was 15.44 % against the 1D diffusion theory prediction of 17.14 % (Marshak boundary conditions). Expressing the deviation in combined uncertainty units (σMC=0.04 %, σtheory=4.0 %, σcombined=4.0 %) gives 0.43 σcombined—well within the 2 σ criterion. The dominant uncertainty is the diffusiontheory approximation, not the simulation. Throughput of 101.5 Mbnc/s on a 100 W consumer GPU is comparable to published results from the Shift GPU transport code [1] on professional HPC hardware costing an order of magnitude more, and exceeds a 32-core CPU node running OpenMC by an order of magnitude on a per-watt basis. These figures are obtained without hardware ray-tracing or material-sort optimisations active, leaving substantial performance headroom. Table 2: Benchmark summary – 1D slab, 106 neutrons, RX 6700S. Metric Value Total simulation time 669.6 ms Throughput 101.5 Mbnc/s (million neutron-bounces/s) Geometry step (RadeonRays BVH) 286.8 ms Physics step (HIP kernel) 294.8 ms Host transfer overhead 77.0 ms 11.5 % of total Absorption – measured 15.44 % Absorption – theory (1D diff) 17.14 % Deviation 0.43 σcombined PASS

A rapidly expanding bubble in nucleate boiling can form a thin layer of liquid between the solid wall and the liquid-vapor interface, known as the microlayer. As the microlayer works as a heat transfer bridge of low thermal resistance between the heated wall and the bubble, its evaporation strongly contributes to the overall boiling growth.

Recently, Tecchio et al. experimentally showed that the formation of microlayer is analogous to the deposition of a thin liquid film on a flat plate that one pulls out vertically from a pool of liquid as described by Landau & Levich. According to their theory, valid in the asymptotic limit of low capillary number Ca << 1, the initial microlayer thickness depends on the speed of receding of the bubble foot edge U, and its radius of curvature r_c. Experimental results on microlayer thickness can be reproduced by the theory assuming that, during the microlayer formation process, r_c is proportional to the bubble radius r_b, so the bubble growth is self-similar. The problem of such an approach is that it is unclear at which point of the bubble foot r_c and bubble foot velocity U should be defined. Both parameters vary along the bubble interface in the bubble foot region. Unfortunately, it is not even possible to observe it by the conventional optical sidewise observation (i.e., by shadowgraphy) because strong thermal gradients that exist near the wall cause a mirage effect: the bubble interface image is uncontrollably distorted. The only way to measure r_c without any distortion is to use the X-ray imaging.

In this work, we provide an analysis of recordings performed by Yu et al. to check if the aforementioned assumption of self-similar growth is valid. For this purpose, we processed the X-ray recordings using the edge-detection algorithms and analysed the radius of curvature in different points of the bubble interface. Our findings show that the ratio r_b/r_c indeed remains close to a constant over time and a minimum radius of curvature r_c^min exists along the bubble edge. Such results support the validity of the microlayer formation model based on the Landau-Levich theory.

A lot has changed since the first attempts more than 25 years ago to establish a common file format for NDE data: With the advent of advanced ultrasound techniques such as full matrix capture, raw data storage has become a highly desirable feature for many end users; with HDF5, the hierarchical data format HDF5 has become the binary layer of choice for vendors and end users alike, and most importantly, vendors have embraced open data formats as a marketable feature.

The OpenNDE data format (ONDE for short) is a joint effort initially led by COFREND, gathering end users, stake holders, vendors and academics to define a pragmatic yet functional data format with the capability to store raw and processed data and the context information necessary to exploit and re-interpret the data. It builds upon efforts done on the MFMC 2.0 format specification, borrowing other elements from prior work done at EPRI and Iowa State University. The recently published specification targets ultrasound, with a first extension to other modalities underway to include eddy current