Conference Agenda

Abstract

To enhance severe accident mitigation capabilities in the APR1000 reactor design currently being exported to the Czech Republic, a passive core catcher system has been incorporated into the containment configuration. This system is intended to capture and cool molten corium released from the reactor pressure vessel (RPV) following a hypothetical vessel breach. The core catcher utilizes a dual-mode cooling approach involving both bottom cooling and top flooding, thereby enabling redundant and robust heat removal paths. In particular, the bottom cooling strategy depends on a buoyancy-driven natural circulation loop that plays a vital role in the long-term cooling of the corium and the overall structural integrity of the system.

To assess the thermal-hydraulic behavior and stability of this natural circulation flow, an integrated experimental and computational study was conducted in collaboration with the Korea Atomic Energy Research Institute (KAERI). The experimental campaign was performed using the Variable PECS EXperimental facility (VPEX), a scaled-down test apparatus designed to replicate the geometric and thermal characteristics of the bottom cooling channel in the APR1000 core catcher. Within the VPEX setup, heater blocks were installed to simulate the decay heat generated from molten corium, and the heat input was adjusted to reflect postulated accident scenarios. The experiments measured key parameters such as coolant temperatures, flow velocities, and mass flow rates under steady-state conditions.

Simultaneously, numerical simulations were carried out using the MAAP5.06 severe accident analysis code. A detailed VPEX model was developed within MAAP to replicate the test geometry and experimental boundary conditions. This included careful alignment of input variables such as the inlet coolant temperature, channel entrance pressure, heater surface heat flux, and elevation head, all of which were matched to the experimental setup. In addition, MAAP-specific user-defined inputs—including drift velocity, interfacial slip conditions in two-phase flow regions, and empirical loss coefficients for the upper plenum and channel junctions—were appropriately determined based on sensitivity analysis and physical reasoning, and further tuned through comparison with the experimental data.

The simulation results showed favorable agreement with experimental measurements in terms of flow patterns, temperature distribution, and heat removal capability. The natural circulation loop was shown to remain thermally stable under the given thermal loading, with the buoyancy forces generated by density differences between heated and cooled fluid regions being sufficient to sustain circulation. The analyses also confirmed that most operating conditions remained within the single-phase or low void fraction two-phase regime, which is desirable for system predictability and control. Parametric sensitivity revealed that the drift flux model and associated two-phase parameters had a significant influence on flow prediction accuracy, underscoring the importance of careful model calibration in MAAP.

This study demonstrates that MAAP5.06, when validated against experimental data, can effectively simulate the thermal-hydraulic performance of passive core catcher systems subjected to severe accident conditions. The integrated use of the VPEX facility and MAAP5 modeling provided a comprehensive framework to assess natural circulation loop behavior and offers key technical justification for the APR1000 core catcher’s thermal reliability in deployment scenarios such as the Czech Republic.

Furthermore, the outcomes of this study contribute to the broader field of severe accident mitigation and passive system analysis by providing benchmark data for model validation and guidance for design optimization. The results may also support licensing applications, reliability evaluations, and the formulation of post-core melt accident management strategies in next-generation pressurized water reactors equipped with in-containment core retention devices.

As the demand for high burnup fuel operation continues to grow, driven by extended fuel cycles and increased enrichment levels, regulatory evaluation of Emergency Core Cooling System (ECCS) performance under such conditions becomes increasingly important. Traditional Large Break Loss-of-Coolant Accident (LBLOCA) analyses often simplify core modeling with limited representation of burnup effects. In response, this study performs a regulatory audit assessment of ECCS performance in OPR1000 LBLOCA analysis by modifying the MARS-KS core modeling to reflect burnup-dependent thermal and mechanical characteristics. Specifically, the reactor core is segmented into fuel groups based on burnup cycles—fresh, once-burned, and twice-burned—and thermal-hydraulic properties such as gap gas composition and pellet thermal conductivity are applied accordingly. The modified model consists of one average core channel and three hot channels, each linked to a specific burnup group, and incorporates nine heat structures categorized by burnup.

Base case LBLOCA calculations were conducted using MARS-KS with this refined core model, and the impact of burnup on peak cladding temperature (PCT) was analyzed. Results indicated that in certain cases, PCT was higher in once-burned fuel than in fresh fuel, primarily due to degradation of heat transfer caused by burnup-induced changes in gap gas composition. Moreover, although pellet thermal conductivity degradation (TCD) also affects PCT, the same TCD was applied to all heat structures except for fresh fuel, limiting the ability to quantitatively evaluate its impact. This highlighted the necessity of incorporating realistic, burnup-dependent material properties in regulatory assessments.

To account for modeling uncertainties, 124 LBLOCA simulations were performed using the KINS Realistic Evaluation Methodology (KINS-REM), applying 18 uncertainty variables based on prior validation studies. The third-order Wilks formula was used to derive a statistically conservative PCT₉₅/₉₅ value, which was determined to be 1,332.5 K—well below the regulatory acceptance criterion of 1,477 K. Conservative modeling choices, such as a narrowed and higher power distribution and increased cladding roughness, contributed to the higher PCT compared to vendor results, yet the system remained within safety limits. This study demonstrates that the ECCS performance for the OPR1000 LBLOCA scenario remains within the regulatory acceptance criteria, even when burnup effects are reflected through modified core modeling.

NuScale is a Light-Water Small Modular Reactor (LW-SMR) with an integral reactor pressure vessel design that relies on natural circulation to establish the primary side mass flow rate. For this purpose, a detailed 3D thermal-hydraulic model of NuScale has been developed in the TRACE code, which includes the Decay Heat Removal System (DHRS), Emergency Core Cooling System (ECCS), the containment vessel, and the reactor pool.

In the current study a Generic Transient is simulated, assuming SCRAM, the normal actuation of the DHRS while the Reactor Safety Valves (RSVs) remain closed. Additional analysis were performed to evaluate system behavior under different safety system failures. These include cases where the DHRS does not initially actuate , but one RSV opens to depressurize the Reactor Pressure Vessel, followed by a recovery of the DHRS. Further cases are made to evaluate the reactor behavior if the RSV successfully closes or fails to reclose. If the RSV closes, the system returns to stable conditions with DHRS cooling. If the RSV fails to close, the core cooling must rely on the ECCS, leading to different system responses depending on whether the ECCS actuates properly or not.

This study provides for an in-depth understanding of the thermal-hydraulic behavior and the interaction of the passive safety systems in NuScale reactors under accident conditions, contributing to the overall safety evaluation of the SMR concept.

In the event of a loss of coolant accident involving a breach in the primary circuit of a nuclear power plant, boron water must be injected by safety system to cool the reactor core. The containment spray system is also used to reduce the pressure and the temperature in the reactor building. These two safety systems are initially fed from a dedicated water reservoir. At a low threshold, a switchover is made to recirculation mode, in which the water collected in the sump at the bottom of the containment (water from the breach and water from the safety systems) is reinjected into the core after first being filtered by filters located in the sump. In such a situation, debris is generated and may be partially transported to the sump filters, contributing to the 'physical' and possibly 'chemical' clogging of these filters. The chemical contribution corresponds to the formation of gels or precipitates within the fiber bed. These gels or precipitates result from the partial dissolution of contaminants (e.g. insulation, paint, concrete) and corrosion of metal surfaces, releasing reactive ions into the solution. This clogging can lead to failure of the safety recirculation systems and potentially to a severe accident due to inadequate cooling of the reactor core.

This research project aims to identify the reaction mechanisms, the chemical species formed, the key parameters influencing their formation, and their impact on filter head loss under accident conditions.

The experimental methodology is based on two complementary scales. First, small-scale
(~1 L) analytical tests were carried out to investigate the chemical mechanisms involved (dissolution and precipitation). Dissolution experiments carried out in H₃BO₃/NaOH and H₃BO₃/NaOH/Borax (Na₂B₄O₇.10H₂O) solutions (pH~7.5) at 80 °C for 48 to 72 hours revealed a significant release of silicon from debris, mainly mineral fibers, exceeding that of calcium and zinc. Studies of the reactivity of silicon, calcium, and zinc in the H₃BO₃/NaOH/Borax solution showed that this medium enhanced reactivity compared to the H₃BO₃/NaOH solution and promotes precipitation phenomena. In addition, the presence of zinc was found to promote the formation of precipitates, particularly silica (SiO₂), and various silicate and borate species. A thermodynamic analysis using a geochemical speciation software (CHESS) confirmed the formation of compounds such as amorphous silica, colemanite (Ca[B₆O₁₁(OH)₂]3H₂O) and willemite (Zn₂SiO₄).

Finally, medium-scale experiments (140 L) were carried out using the COPIN (Clogging of Sumps in the Nuclear Industry) facility to investigate the precipitate formation under recirculation conditions and to assess its impact on filter head loss. Tests were performed in H₃BO₃/NaOH/Borax and H₃BO₃/NaOH solutions, simulating accidental conditions including galvanized steel corrosion and temperature transitions. A significantly higher pressure drop was observed in the H₃BO₃/NaOH/Borax solution, attributed to combined physical and chemical phenomena, an effect not observed in the H₃BO₃/NaOH solution. The results indicate that borax exacerbates filter pressure losses, probably due to additional chemical interactions and precipitate formation. Experimental investigations carried out on COPIN reveal the presence of a chemically induced pressure drop, which is often overlooked due to its modeling complexity, yet plays a significant role in the overall system behavior.

Passive safety systems are widely recognized as fundamental features of advanced reactor designs, including Small Modular Reactors (SMRs), due to their role in enhancing intrinsic plant safety. These systems operate without the need for external power sources (except possibly during activation), relying instead on natural driving forces such as gravity or natural circulation. Their simplicity and the absence of moving parts post-activation reduce the likelihood of hardware failures; nevertheless, their reliability must be rigorously demonstrated.

In particular, specific scenario conditions that affect gravity-driven phenomena play a crucial role in determining the operation of passive systems. These conditions directly impact the fulfillment of the safety function and, therefore, need to be deterministically characterized. In this context, the so-called functional failure—linked to the failure of gravity-driven (in general “natural”) phenomena, even in the absence of mechanical or electrical failure—can affect the desired passive system operation and its reliability. Ensuring the robustness and predictability of passive systems is essential for the safety demonstration.

To this end, alongside dedicated experimental campaigns designed to characterize passive system behavior, the use of validated thermal-hydraulic system codes is crucial. These tools enable simulation across the full spectrum of plant conditions, supporting both design and safety assessments.

This paper presents a preliminary study conducted by ENEA within Work Package 4 (WP4) of the Horizon Euratom EASI-SMR project, coordinated by EDF (France). The EASI-SMR initiative aims to ensure the highest level of safety of LW-SMRs based on passive systems integrating them into EU regulatory framework. WP4, coordinated by UJV (Czech Republic), aims to consolidate the current and well proven methodologies for assessing the reliability of passive safety systems in LW-SMRs.

The passive system investigated in this study is a Decay Heat Removal System (DHRS), designed to passively extract residual heat from the reactor core under accident conditions via natural circulation. The system transfers heat from the primary circuit to a pool through two heat exchangers arranged in series: (i) a plate-type Compact Steam Generator (CSG) coupling the primary and secondary loops, and (ii) a tube-type heat exchanger immersed in the pool and connected to the secondary side. Experimental data were derived from tests performed at the ELSMOR facility, constructed in 2022 at the SIET laboratories in Piacenza, Italy.

This study performs an uncertainty analysis using the thermal power transferred by the passive system to the pool as the Figure of Merit (FOM). Key parameters influencing this performance include pressure losses, materials thermal conductivity, and heat transfer coefficients. Parameter variability ranges and Probability Density Functions (PDFs) were defined based on literature review and expert engineering judgment. The best thermal hydraulic system code used is the USNRC RELAP5/Mode 3.3. code. The objective of the study is to identify and characterize the key parameters governing DHRS performance and characterize their effect on the FOM. These will serve as the foundation for the upcoming ENEA application of the REPAS methodology in EASI-SMR WP4.

ATHLET-CD (Analysis of THermal-hydraulics of LEaks and Transients with Core Degradation), which is part of the system code package AC² (ATHLET, ATHLET-CD, COCOSYS), is designed to simulate the behaviour of a nuclear power plant during accident scenarios, including core damage progression, fission product release and behaviour as well as lower plenum phenomena. Previously, models and input decks assumed that the cladding material is made from standard, zirconium-based materials. These cladding materials have been used in water-cooled reactors for decades all over the world, despite some of their drawbacks at high temperatures, like the exothermic reaction with hot steam that can aggravate accident progression and produce easily combustible hydrogen. To eliminate or mitigate these drawbacks of the zirconium-based claddings new materials were developed. These so-called accident tolerant fuel (ATF) materials have different properties that can significantly alter the accident evolution compared to the standard cladding materials.

For predicting and analysing the impact of ATF on accident scenarios, GRS is currently implementing models for such ATF materials in AC²/ATHLET-CD. In this paper we briefly introduce the newly developed and adjusted models that aim to model the phenomena occurring with iron-chromium-aluminium (FeCrAl) cladding material, one of the most promising ATF materials. Using the new models, a hypothetical accident scenario of how the accident of TMI-2 would have evolved, if the fuel rods would have had FeCrAl cladding instead of the zirconium-based cladding, has been analysed. It is found that FeCrAl cladding would have altered accident progression strongly and might have prevented a core melt scenario. A short outlook on future model development tasks and analyses is also provided.