Search

In this study utilization of neural networks and their optimization have been investigated in order to monitor the efficiency performance of pressurized water reactor (PWR) nuclear power plants. Various configurations and scaling methods have been examined for accurate monitoring of efficiency, fault detection, fault extent and generation losses under faulty state. Twelve different fault conditions, related to the feed water heaters, along with one steady state operating condition have been analyzed through the use of different architectures of back propagation neural network. Due to highly sensitive nature of efficiency parameter against various operational faults, ten different non-parametric scaling methods, which do not require the recalculation of scaling parameters in case of new data availability, have been studied. A novel scoring method has also been proposed for comparison among the performances from various networks. At the end, the robustness of optimized network has also been tested against the noisy input data. • Neural networks are used to develop online efficiency monitoring system for PWR. • Proposed model is able to monitor efficiency, detect faults, and report fault extent. • 12 configurations and 10 scaling methods have been studied for network optimization. • A novel scoring method has been proposed in order to compare the performances. • Robustness of proposed model has also been tested against the noisy data. [ABSTRACT FROM AUTHOR]

• The model, FIDCAV, can enable real-time fault monitoring and hierarchical fault diagnosis for nuclear power plant systems. • The relationship between the detection results and data was examined through the visualization of SPE contribution of KPCA. • Through comparative analysis, the superior diagnostic performance of the hierarchical diagnostic model was validated. • The diagnostic model balances timeliness and accuracy, while addressing the practical needs of nuclear power plants. To study the fault intelligent detection and diagnosis method of nuclear power plant systems and improve the detection and diagnosis effect of internal fault of nuclear power plant Chemical and Volume control System (CVS), this study presents an intelligent F ault D etection and D iagnosis model for the C hemical and V olume control S ystem (FDD-CVS) in nuclear power plants (NPPs). The model is based on failure mode and effects analysis of the CVS system and is implemented by combining kernel principal component analysis (KPCA) with decision tree and support vector machine (SVM). FDD-CVS can rapidly and visually recognize faults in CVS based on independent time-point system parameters, and it is capable of diagnosing fault types and specific fault locations. The model is characterized by clear principles, hierarchical diagnostics, fast diagnostic speed, and visualized results. The model is trained and tested by using the data of the passive nuclear power simulation analyzer. The fault detection rate of FDD-CVS is 96.38%, the false alarm rate is 4.34%, and the average accuracy rate is 98.40%. Overall, the fault monitoring and diagnostic method proposed in this article is innovative and provides valuable references for fault diagnosis research in nuclear power plants. [ABSTRACT FROM AUTHOR]

• The study reviews NPPs' organizational vulnerability, formation mechanisms, and evaluation methods. • Develops an AHP-Fuzzy-based evaluation model for NPPs' organizational vulnerability, enhancing assessment objectivity. • Applies the model to identify flaws in NPPs' structures, proposing measures to improve safety standards. The operational efficiency and developmental progress of nuclear power entities are significantly challenged by organizational vulnerability, which can lead to severe outcomes if neglected. This paper presents a systematic approach to identifying and quantifying the primary factors influencing organizational vulnerability within nuclear power plants (NPPs). An evaluative index is established, and an innovative hybrid methodology combining the Analytic Hierarchy Process (AHP) and fuzzy sets theory is applied to assess overall organizational vulnerability. A case study validates the approach, demonstrating low vulnerability and strong organizational reliability in the NPPs. The research contributes a valuable tool for enhancing safety and sustainability in the nuclear power sector. [ABSTRACT FROM AUTHOR]

• The cavity-vortex evolution in cavitation flow within control valves are analyzed. • A method for analyzing the unsteady characteristics of flow parameters is proposed. • The flow details of cavity-vortex evolution were captured by the LES model. • Increasing the flow area within the allowable range helps to reduce cavitation. Nuclear control valves operating under high pressure drop often experience damage such as cavitation, vibration, and noise, which pose a significant safety risk to the nuclear plant. However, there is currently a lack of research on the unsteady cavity-vortex evolution, resulting in a limited understanding of cavitation flow in valves. In this study, the evolution of the cavity-vortex structures in control valves with different openings and structural parameters was studied, and the unsteady characteristics of flow were analysed by mechanical energy pulsation. Results indicate that, under the same pressure ratio conditions, reducing valve opening increases the influence range of quasi-periodic cavity-vortex structures and cavitation intensity. Furthermore, the unsteady pulsation of flow parameters becomes more obvious. Additionally, increasing the throat length-to-diameter ratios or decreasing the expansion angle enhances the evolution of cavity-vortex, which consequently has a stronger impact on flow unsteadiness. Therefore, it is recommended to avoid small openings and maximize the effective flow area of the throttling passage within the allowable range. [ABSTRACT FROM AUTHOR]

• Adual-layer correlation analysis framework (PCAF) based on five different models. • Use PCAF framework to conduct parameter analysis on thermohydraulic bench. • Achieve mapping of parameter relationships on bench to digital topological graph. • Design a method to rebuild topological graph to parameter correlation network. • Create a multi-dimensional parameter prediction framework based on LSTM and GCN. In nuclear power plants (NPPs) operations, the prediction of multi-dimensional parameters is found to help operators to grasp the condition of the system. However, majority of existing studies are focused on single-dimensional parameter prediction. In this study, a multi-dimensional parameter prediction framework of NPPs based on Long Short-Term Memory Network and Graph Convolution Network (LSTM-GCN) and a multi-model integrated parameter correlation analysis framework (PCAF) are proposed, in which PCAF is used to build a parameter correlation network for GCN, and LSTM-GCN is used to predict multi-dimensional parameter of NPPs. To verify the feasibility of the LSTM-GCN framework, multi-dimensional parameter prediction researches are conducted using data from a thermohydraulic experimental bench that simulates the operation of NPPs. Results indicate that compared to traditional prediction models, LSTM-GCN framework enhances the prediction accuracy of multi-dimensional parameter, which benefits from the ability of LSTM-GCN to utilize the temporal dependencies and spatial correlations of parameters. [ABSTRACT FROM AUTHOR]

• Development of a thermal power dispatch (TPD) system to transfer thermal energy from boiling water reactors to hydrogen plants. • TPD system heat exchangers designed with Aspen Exchanger Design and Rating software. • System thermal hydraulics analysis of TPD using Simulink Simscape. • TPD system performance assessed for both transient and steady state operations. Nuclear power plants are crucial to meeting net zero emission goals and achieving energy sustainability. Integrating these plants with clean energy technologies such as high-temperature steam electrolysis (HTSE) may improve the efficiency and economic competitiveness of these plants. The current study investigates the design and operation of a thermal power dispatch (TPD) system for coupling boiling water reactors (BWRs) to HTSE plants. The TPD system extracts a portion of the steam from the reactor's main steam line and transfers its thermal energy to an HTSE plant through a power transport loop. A TPD system for 5 % steam extraction has been designed and the system performance during steady and transient operations has been analyzed. The TPD system dispatched a total of 197 MW thermal energy to the HTSE plant under nominal design conditions. Saturated steam at 7.17 MPa from the BWR plant was condensed and subcooled to a temperature of 168 °C, while a mass flow rate of 91.1 kg/s of superheated steam was dispatched to the HTSE plant. Furthermore, the system performance during transient operation showed a continuous transition from the initial hot standby mode to the nominal power dispatch level. The transient simulation results emphasized the importance of investigating component level performance for the TPD system design. The current results will guide future works on the development of integrated energy systems for hydrogen production or process heat applications. [ABSTRACT FROM AUTHOR]

To develop the expected isotopic signatures of online nuclear power plants, a CANDU-6 quarter-core reactor was modeled using Serpent, a Monte Carlo and Burnup code. The reactor model was simulated using nominal operating parameters, steady power levels and standard refueling procedures, to set the baseline for online operations. The model was burned for 500 refuelings totaling 1400 effective full power days. The core was divided into 1140 spatially-discrete fuel bundles with each tracking the density of 237 isotopes. Instantaneous core inventory snapshots were recorded at the time of each refueling to create a continuous inventory database. These snapshots provide the expected isotopic densities and ratios for virtually any fuel bundle position or burnup under nominal operating parameters. These values are useful in the event of accidents, short-cycles, or nuclear proliferation. The time-dependent and spatially dependent results for xenon effluent are used to develop an analytical method for calculating the expected International Monitoring System xenon ratio measurements based on fuel bundle leak rates. A possible false-positive nuclear proliferation scenario for a CANDU-6 operating under nominal parameters is also identified. • Online isotope inventories modeled and analyzed over 500 channel refuelings. • Transient radioxenon analysis of online inventory for IMS monitoring. • Possible false-positive nuclear proliferation scenarios identified. [ABSTRACT FROM AUTHOR]

Multiple uncertainties still exist about the state of the debris in Fukushima Daiichi Nuclear Power Station (1F). In the past, the attention of the nuclear safety community was focused on the heat transfer characteristics in the case of an homogeneous pool, but little attention was given to address the melting and heat transfer in the presence of a debris bed constituted of materials with different melting points. This condition represents a challenge for CFD analyses, because it includes multi-physics conditions, such as a low melting point fluid convecting into a debris bed surrounded by a crust on the vessel wall which has received little attention compared to classical CFD analyses. Even though a comprehensive analysis of a related experiment (i.e. LIVE-J2) has been performed recently by Madokoro et al. (2023) little attention on the results has been paid to the effect of debris bed porosity and the existence of a gap between the vessel wall and the crust. In the paper we have modified the porosity resistance based on the Ergun equation and proposed a simple model for the gap conductance in the lower part of the crust. The results show an improvement in the prediction of the thermal stratification and the vessel temperature in the lower locations. In addition, highlight that such phenomena constitute key parameters to keep into consideration in the simulation of prototypical cases both for CFD and lumped parameter codes (e.g. MELCOR, MAAP). [ABSTRACT FROM AUTHOR]

• Introduces the Fuzzy Analytic Hierarchy Process (AHP)-VIKOR technique for risk ranking in the nuclear power industry. • Emphasizes the need for thorough examination and sequential assessment of each component's risks to minimize failures and risks. • Utilizes expert evaluations based on Severity, Occurrence, and Detection criteria to calculate risk weights and rankings. • Provides a higher-level risk analysis that incorporates multiple performance parameters. The critical systems in industries such as nuclear power plants rely on various preventive methods to minimize failures or risks through efficient strategies and equipment. However, in many businesses, maintenance tasks are carried out infrequently, improperly, and without consideration for the overall state of the plant or its equipment. To choose the appropriate risk preventive approach, a thorough examination of each component's risks in a sequential manner becomes imperative. This paper introduces the Fuzzy Analytic Hierarchy Process (AHP)-VIKOR technique, a Multicriteria Decision Making Approach employed to rank the various risks prevalent in the nuclear power industry. By identifying the proper sequence of risks, this approach aims to reduce the occurrence of unfortunate mishaps, along with minimizing recovery time and costs. Five experienced researchers and experts assessed the impact of risk based on three risk criteria: Severity, Occurrence, and Detection. Utilizing the opinions and judgments of these experts, the Fuzzy AHP-VIKOR approach was employed to calculate the weight of each performance criterion and the ranking of each risk or hazard. The suggested method is designed to assist supervisors in resolving discrete problems characterized by incommensurable and conflicting criteria. This study contributes to the industry by providing a higher risk analysis incorporating various performance parameters. The paper concludes by presenting a result with the priorities of all risks in the industry using the fuzzy AHP-VIKOR method. [ABSTRACT FROM AUTHOR]

This paper introduces methods to develop a surrogate model based on deep learning methods and rolling-window forecast for fast and accurate prediction of severe accidents in a nuclear power plant. The surrogate model was trained using time series data, which represents thermal–hydraulic behavior in the nuclear power plant under multi-component failures while various mitigation strategies are also implemented. The model uses a rolling-window forecast to predict selected thermal–hydraulic variables for each time step using the previous time-step variables. To improve the accuracy, the model was further refined to consider the hysteresis effect of the variables using the previous three-time steps. The value of the performance metrics measured by the mean absolute error was reduced by 64 percent in the three-step model compared to the single-step model. The proposed surrogate model has the potential as a practical severe accident simulator for accident management support tools. [ABSTRACT FROM AUTHOR]

• Designed and constructed a fault data acquisition system for nuclear power plant electric gate valves. • Based on the study of the fault mechanism of electric gate valves, analyzed the occurrence locations of faults in electric gate valves, completed fault settings, and provided theoretical guidance for designing relevant experimental platforms for other scholars. • Analyzed the changes in acceleration signals and AE signals when different types of faults occur in electric gate valves, selected corresponding sensors, and completed the arrangement of sensors and signal acquisition. • To effectively improve the accuracy of fault classification and the precision of fault severity assessment, this paper improved the original algorithms and developed a fault diagnosis system for nuclear power plant electric gate valves based on the SAE algorithm and a fault severity assessment system based on the Bi-LSTM algorithm. Electric gate valve failure is a common type of fault in nuclear power plants. Among all factors leading to reactor shutdown in nuclear power plants, valve failures account for a significant proportion. Due to the difficulty in obtaining valve failure data in nuclear power plants, to effectively obtain such data and provide data support for the development of subsequent fault diagnosis algorithms, this paper adopts an experimental research method. It designs and constructs an electric gate valve failure simulation test bench, obtains experimental data under various states of electric gate valves, and develops a fault diagnosis system for electric gate valves in nuclear power plants based on this. The experimental results show that the data generated in this experiment can well achieve the purpose of classifying valve failure types and evaluating the degree of failure. Moreover, the developed fault diagnosis system exhibits high diagnostic accuracy and low error in evaluating the degree of failure. [ABSTRACT FROM AUTHOR]

• The SIRIUS-3D rod bundle tests were validated using the CUPID code. • The drag coefficient interfacial friction model without the EVVD model provided the best prediction. • At a low velocity of 0.1 m/s, void friction was over-predicted, which could be improved by reducting interfacial friction. In nuclear power plants, the occurrence of a three-dimensional mixing phenomenon in two-phase flow can help mitigate the peak cladding temperature of fuels in a core. Simulating this phenomenon necessitates subchannel-scale analysis. CUPID is a three-dimensional thermal hydraulic analysis code developed by KAERI. This paper presents the validation results of the code against rod bundle tests conducted using SIRIUS-3D, a 5 × 5 rod bundle test facility. One test utilized subchannel void sensors, while the other employed X-ray CT. To validate the SIRIUS-3D test data, subchannel-scale nodes were constructed for CUPID calculation. For improved convergence, the implicit friction calculation was conducted with a small CFL ratio. The drag coefficient interfacial friction model without the EVVD model yielded the best prediction. The void fraction was over-predicted at a low velocity of 0.1 m/s, which could be improved by reducing the interfacial friction. [ABSTRACT FROM AUTHOR]

Natural Circulation Loops (NCLs), where fluid is driven through a closed circuit solely by thermal imbalance, offer potential for use in passive cooling systems within nuclear power plants. Transient CFD simulations of two-dimensional NCLs of different aspect ratios have been performed using the Unsteady Reynolds-averaged Navier–Stokes framework for a range of Rayleigh numbers (8 × 1 0 9 < R a m < 8 × 1 0 13 ), at three Prandtl numbers (P r = 0. 01 , 0. 71 , 7. 1) and for two different loop aspect ratios in order to provide insight into the transient and stability behaviour of such systems. Results predict that NCL systems exhibit complex and rich dynamic behaviour, with strong sensitivity to both the imposed Rayleigh number and Prandtl number. Initial thermal transients led to the establishment of unstable and oscillatory flow behaviour which, in several cases, led to flow reversals. At the high and moderate Prandtl number, cases at the lower R a m = 8 × 1 0 9 reached a statistically steady-state whilst flows at the higher R a m = 8 × 1 0 13 demonstrated more oscillatory, but relatively regular, behaviour, including persistent flow reversals. A significant reduction in Prandtl number to P r = 0. 01 led to much more stable behaviour, with cases at all R a m producing shorter initial transients and settling quickly to a statistically steady-state. This is primarily as a consequence of the increased dominance of heat transfer by conduction and subsequent reduction in overall temperature variations. For cases that produced statistically steady-state behaviour, output Reynolds numbers demonstrated good agreement with existing experimental trends and established correlations. These results suggest that utilizing fluids with low Prandtl numbers may be an effective means of improving the stability of NCLs for a given heat input without compromising the output Reynolds number. • Transient URANS simulations of 2D Natural Circulation Loops. • Assessed the influence of Prandtl number, Rayleigh number and loop aspect ratio. • Range of stable and unstable flow behaviour predicted, including flow reversals. • Results in good agreement with experimental trends and established correlations. • Lower Prandtl number improves flow stability without reducing output Reynolds number. [ABSTRACT FROM AUTHOR]

This paper presents the results of the assessment of the radiological consequences of LOCA at DBA and DEC-A for NPPs with VVER-440 and 1000 reactors using approaches that contribute to a realistic assessment of the radionuclide sources, taking into account the thermomechanical criterion of the gas gap activity release and the process of their transfer from the primary circuit to the environment. The results of the improved modelling are compared with conservative estimates. [ABSTRACT FROM AUTHOR]

• Propose KSCAE combining K-Means and stacked convolutional autoencoder for better similar fault diagnosis. • Apply K-Means for data segmentation, guided by the effective elbow method. • Develop classifiers with stacked convolutional autoencoder, tuned via Bayesian optimization, enhancing fault diagnosis. • KSCAE achieves top accuracy, boosting diagnostic accuracy by 14.67%-21.54%, excelling in distinguishing similar faults. • K-Means enhances robust classification boundaries, improving accuracy in diagnosing similar faults. Nuclear power plants consist of multiple subsystems characterized by nonlinear correlations, numerous monitoring parameters, and faults with similar characteristics. These similar faults often exhibit closely distributed data points, resulting in unclear classification boundaries and reduced diagnostic accuracy for fault detection algorithms. A single deep learning model cannot establish a diagnostic model that is broad in scope, highly accurate, and robust. This paper proposes a robust, multi-system similar fault diagnosis model for nuclear power plants, combining K-Means and Stacked Convolutional Autoencoders (KSCAE). First, K-Means data segmentation model partitions complex multi-system fault data into several similar data clusters. Then, specialized diagnostic models are developed for each cluster using powerful stacked convolutional autoencoders to focus on learning and classifying similar faults. Testing across three nuclear power plant fault diagnosis scenarios based on data from the Fuqing nuclear power plant simulator demonstrates that KSCAE outperforms single deep learning models in diagnostic accuracy, particularly when fault severity differs between training and testing sets. The results show that the KSCAE algorithm achieves a maximum accuracy of 99.31% and a minimum accuracy of 79.11% under severe noise and data distribution differences. Compared to the baseline algorithms, KSCAE achieves an average accuracy improvement of up to approximately 20% across multiple tests, particularly in Scenario 2. This study demonstrates the effectiveness and robustness of the proposed model for diagnosing similar faults, providing a reliable approach for multi-system fault diagnosis in nuclear power plants. [ABSTRACT FROM AUTHOR]

• A HTransformer method is proposed to improve the accuracy of fault diagnosis in nuclear power plants. • The paper designes a HTransformer model interpretability framework based on the SHAP method. • The test results show that the fault diagnosis performance of HTransformer is significantly improved. • The SHAP method is used to analyze the local and global interpretability of the HTransformer model. In order to improve the accuracy of the fault diagnosis, a hybrid Transformer (HTransformer) model for fault diagnosis in nuclear power plants (NPPs) is proposed. By concatenating convolutional neural networks and gated recurrent unit networks in front of the Transformer encoder, the spatiotemporal feature information of the input information is effectively capture. However, the high reliability demand for NPPs requires researchers to be able to explain the decision-making behavior of deep learning models. The interpretability of the HTransformer model has been preliminarily studied based on the Shapley additive explanations (SHAP) method. The result shows that the HTransformer model has a higher fault diagnosis accuracy of 99.5% compared to other deep learning fault diagnosis models. The preliminary research result on the interpretability of fault diagnosis models based on the SHAP method provides a possibility for deep learning models to be applied in the practical engineering of NPPs. [ABSTRACT FROM AUTHOR]

• This paper implements a statistical and theoretical analysis of iodine spiking phenomena during SGTR events. • An amended iodine spiking model was proposed with operation data collected from 97 cases of U.S. PWR, VVER and Chinese PWR. • The theoretical analysis results of a iodine spike initiated by SGTR are roughly equivalent with the statistical method. • Both the statistical and theoretical results indicate that the current treatment of iodine spiking in regulatory is conservative. • This paper recommends an envelope value of 35 for iodine spiking factor and provides a fitting formula for different PWRs. Iodine spiking phenomena has long been the focal point in the design and review of nuclear power plants. This paper implements a statistical and theoretical analysis of iodine spiking phenomena during Steam Generator Tube Rupture (SGTR) events to enhance our understanding of iodine spiking phenomena, ensuring that analyses are aligned with current plant operations and theoretical models. In contrast to previous literatures, this paper proposes an amended iodine spiking model based on the statistical method which is combined with the practice of safety analysis. Additionally, a concurrent iodine spike initiated by the SGTR event is calculated using the theoretical model. Both the statistical analysis and the theoretical calculation results indicate that the current treatment of iodine spiking in regulatory is indeed quite conservative. Combining both methods, this paper recommends an envelope value for the iodine spiking factor and provides a fitting formula considering the applicability across different PWRs. [ABSTRACT FROM AUTHOR]

• A simulation platform for the small pool-type lead–bismuth fast reactor is established. • The operation scheme is determined using a double-constant operation scheme. • Based on the energy balance analysis, a power-temperature coordinated control system is designed and optimized. Small lead–bismuth fast reactors have become one of the most important types of Generation IV reactor types because of their high inherent safety. Small pool-type lead–bismuth reactors are often used in mobile nuclear power plants. Conventional control logic cannot fulfill the load-following requirements. Therefore, the control systems for nuclear power plants were studied in detail. First, a simulation platform for small-pool lead–bismuth fast reactors was developed through modular modeling. Second, the operation scheme was analyzed and designed. Then, the system was analyzed from an energy perspective, and a power-temperature coordinated control system was developed. The load-following characteristics of nuclear power were improved through load-feedforward control with energy balance correction. Finally, the power-temperature coordinated control system was verified on the basis of a simulation platform. The results show that the designed power-temperature coordinated control system has a good control effect. In this study, a reference for the development of nuclear power control systems for other pool type metal cooled fast reactors was created. [ABSTRACT FROM AUTHOR]

Utilizing fault diagnosis methods is crucial for nuclear power professionals to achieve efficient and accurate fault diagnosis for nuclear power plants (NPPs). The performance of traditional methods is limited by their dependence on complex feature extraction and skilled expert knowledge, which can be time-consuming and subjective. This paper proposes a novel intelligent fault diagnosis method for NPPs that combines enhanced temporal convolutional network (ETCN) with sparrow search algorithm (SSA). ETCN utilizes temporal convolutional network (TCN), self-attention (SA) mechanism and residual block for enhancing performance. ETCN excels at extracting local features and capturing time series information, while SSA adaptively optimizes its hyperparameters for superior performance. The proposed method's performance is experimentally verified on a CPR1000 simulation dataset. Compared to other advanced intelligent fault diagnosis methods, the proposed one demonstrates superior performance across all evaluation metrics. This makes it a promising tool for NPP intelligent fault diagnosis, ultimately enhancing operational reliability. [ABSTRACT FROM AUTHOR]

• Investigates reactor coolant pump performance under uniform and non-uniform inflow conditions. • Demonstrates significant vortex intensity and backflow under non-uniform inflow. • Quantifies 1.67% head and 3.51% efficiency losses with non-uniform inflow. The operational performance of the reactor coolant pump is crucial to the safety and reliability of nuclear power plants, However, in practical applications, non-uniform inflow at the inlet can significantly affect the pump's stability and efficiency. This study focuses on the nuclear reactor coolant pump, conducting refined numerical simulations of its internal flow. The internal flow field distributions of the reactor coolant pump's overflow components are analyzed under both uniform and non-uniform inflow conditions, and the Q-criterion is applied to capture vortex structures inside the pump. Results show that the scale and intensity of vortex structures increase under non-uniform inflow, especially in the outlet section of the pressurized water chamber casing. This study aims to gain a deeper understanding of the reactor coolant pump's performance under complex conditions, providing theoretical support for its efficient operation and optimal design. [ABSTRACT FROM AUTHOR]

• A hybrid LSTM-TCN model is developed for rotating machinery fault diagnosis. • The model can extract deep spatio-temporal correlation features in vibration data. • It has excellent diagnostic performances for complex faults in rotating machinery. Vibration signals typically used for health monitoring of rotating machinery has highly integrated spatio-temporal correlations. However existing studies rarely explore the impact of spatial correlation features of rotating machinery internal components on their vibration signals. To identify the health condition of rotating machinery in NPPs in terms of spatial–temporal correlation, we propose a fault diagnosis method with combination of the long short-term memory and temporal convolutional networks. The spatial and temporal features in the vibration signals of rotating machinery are extracted using the two networks and then fused to diagnose its faults. The model was assessed against the Case Western Reserve University bearing dataset, University of Ottawa bearing dataset and Southeast University gearbox dataset. The results show that its diagnostic accuracy reaches up to 99.56 %, 100 %, and 100 % on the three datasets, respectively, and outperforms other five well-designed comparative models, demonstrating its effectiveness and superiority in rotating machinery fault diagnosis. [ABSTRACT FROM AUTHOR]

• We find that shortcomings in the water intake, including an inadequate design basis for water intake open channels and sewage interception networks, as well as deficiencies in marine organisms early warning and monitoring equipment and layout. The sea ice also represents a novel challenge confronting the nuclear power industry in coastal regions situated at high latitudes. • We can set the encircled cofferdam to achieve active guidance and avoidance, joint wave forecasting to achieve risk investigation and three-dimensional monitoring of sea ice at sea, land and sky to improve the performance of the water intake system. • We propose to install water jet pumps in the open channel dredging system to meet requirements for efficient and reliable long-term operation of nuclear power plants. Nuclear energy, classified as a clean energy source, enjoys extensive application in numerous countries worldwide. In recent years, the growth rate of marine organisms has accelerated, resulting in the formation of large aggregations of marine organisms and other large congregations that have the potential to obstruct the cold source water intake system of nuclear power plants. The safety situation about the cold source is of significant concern. This paper presents a summary of water intake safety events at nuclear power plants around the world in recent years. It also analyzes the current status of comprehensive prevention and control technology for disaster-causing floating bodies at nuclear power plants' water intakes. The analysis is based on the latest developments in global nuclear power technology. This study provides a comprehensive overview of the advancements in fine offshore monitoring and early warning systems, passive interception techniques for open channel dredging, the evolution of active defense measures, the development of composite water intake open channel structures, bubble curtain interception methods, cutting and grinding techniques, and the subsequent transportation of marine organisms. Nevertheless, shortcomings remain in the water intake safety of nuclear power plants. These include an inadequate design basis for water intake open channels and sewage interception networks, as well as deficiencies in marine organism early warning and monitoring equipment and layout. Furthermore, the issue of sea ice represents a novel challenge confronting the nuclear power industry in coastal regions situated at high latitudes. Therefore, it is essential to undertake research and implement improvements to the comprehensive technology used to prevent and control water intake safety in nuclear power plants. It is imperative to provide robust backing for the long-term advancement of nuclear power technology. [ABSTRACT FROM AUTHOR]

• An efficient method is proposed to contain redundant nonlinear characteristic constrains. • Stable and dynamic reconciliation procedure are combined for recursive data reconciliation. • The uncertainty of measurement data estimation is alleviated. • A new method to apply data reconciliation technology is provided. Data reconciliation has been extensively studied in the Nuclear Power Industry because of its benefits including reducing the uncertainty of measurement data and economic superiority. Previous reconciliation methods usually neglect necessary characteristic constraints, causing certain deviation under stable or dynamic situations. By making full use of redundant information from equipment thermodynamic state equations and control transformation functions, a recursive data reconciliation method is proposed to narrow estimation deviation in two aspects. First, different reconciled methods including implicit method, explicit method, coupled method and synthesized method were established based on bilinear orthogonal transformation. Second, recursive process was designed for reconciliation between virtual device and real device. Two typical heat exchanger systems in nuclear plant were selected as case studies. Results show that the proposed reconciliation method decreases the system error in both stable and dynamic situations. Moreover, when implementing the new proposed data reconciliation method to a preheating system with two heat exchangers, it can converge to the specified residual error within 10 iterations. Recursive reconciliation method which was proposed in this paper provides systematic guidance for nuclear power plant operating and maintenance involving data reconciliation. [ABSTRACT FROM AUTHOR]

Accurately assessing personal radiation doses in real radiation environments like nuclear power plants requires precise and real-time reconstruction of three-dimensional radiation fields. The Kriging algorithm, known for its accuracy in spatial interpolation, provides a promising approach for this task. However, its computational demands can be significant, especially in real-time scenarios. To address this, we enhance the Kriging algorithm with GPU acceleration and data partitioning strategies, enabling efficient and accurate reconstruction of three-dimensional nuclear radiation fields. Using Fluka software for Monte Carlo simulations, we generated a virtual radiation field of dimensions 5 m × 5 m × 5 m for a single-source, unshielded scenario, and a field of dimensions 20 m × 6 m × 8 m for a multi-source, shielded scenario. Using the simulated data, we compared the prediction accuracy of the improved algorithm with the conventional Kriging algorithm and further explored factors influencing the acceleration ratio of the improved algorithm. The results indicate that the GPU-accelerated and data-partitioned Kriging algorithm achieves nearly identical accuracy compared to the traditional method. In the single-source, unshielded scenario, with more than 343 known (measurement) points and predicting 95 × 95 × 95 = 857,375 points, the prediction accuracy remains above 92.25%. In the multi-source, shielded scenario, with more than 8000 known (measurement) points and predicting 95 × 95 × 95 = 857,375 points, the prediction accuracy remains above 91.17%. The acceleration performance of the improved algorithm is consistent across both scenarios, with the acceleration ratio increasing as the number of known and predicted points grows, reaching approximately 20 for smaller datasets and up to 93 for larger datasets. Additionally, the acceleration effect of the improved algorithm varies with data partition size, initially increasing and then decreasing as the partition size increases. When the number of known points is 512 and the number of predicted points is 884,736, the optimal partition size lies between 80,000 and 90,000, resulting in a prediction time of only 0.24 s. • GPU-accelerated Kriging algorithm introduced for 3D radiation field reconstruction. • Algorithm's accuracy is comparable to traditional methods. • Speedup of up to 93 times demonstrated on large datasets. • Optimal data partition size identified for enhanced performance. • RTX 4090 outperforms RTX 2080Ti in algorithm acceleration. [ABSTRACT FROM AUTHOR]

• A partial defect detector design proposed for spent fuel inspection before transfer from storage pool to difficult-to-access storage facility. • Performance indicator of a detector applied to the spent fuel inspection was quantitatively analyzed from a regulatory standpoint. • Detection capability of new design was assessed with the randomly distributed fuel defect scenarios. The current nuclear safeguards approach to spent nuclear fuel inspection at nuclear power stations is based on item counting and limited partial defect analysis. With the expected surge in spent fuel storage, limited spent fuel storage pool capacity, and the increasing need for transferring fuel to long-term storage facilities, there is a growing demand for more efficient and cost-effective nuclear safeguards approaches for nuclear materials management in civilian nuclear power facilities. This study proposes a scintillator-based indirect gamma detector for spent fuel inventory screening inspection, specifically designed for use in interim storage pools prior to fuel transfer to difficult-to-access storage. This paper presents the design of the proposed detector, its application for spent fuel screening inventory inspection, and analysis using MCNP for partial defect detection. Results of analysis indicated that verifying a ∼ 13.6 % level of randomly distributed fuel defect for the Westinghouse 17x17 fuel assembly is possible using this approach. The performance evaluation also indicates that inspection of spent fuel assemblies of various vendor types against 1 SQ diversion may be possible. [ABSTRACT FROM AUTHOR]

• The extent to which PSFs contribute to the organizational performance of operators in the main control room of a nuclear power plant is an crucial issue in decision-making. A questionnaire was used to study the effect of PSFs on organizational performance, and the survey data were quantified using hierarchical analysis. After excluding invalid questionnaires, the final statistics were expressed in the form of x ¯ ± s. • To address the bias of traditional hierarchical analysis methods that rely on point values to estimate weights, this paper utilizes the interval hierarchical analysis method. This approach replaces point estimates with intervals, mitigating the limitations of the traditional method. • To determine the optimal values within these intervals, this paper constructs a quadratic programming model. The weights of PSFs serve as the decision variables, and the objective function aims to maximize organizational performance. The constraints are linear, making the model a convex optimization problem, which ensures that the global optimal solution can be easily found within the feasible domain. • This paper introduces a high-precision power setting method based on the optimization model for maximizing organizational performance. This method is highly transferable and offers a new model and approach for research in power setting theory. The organizational performance of main control room (MCR) operators is crucial in digital nuclear power plants (NPPs). A comprehensive study determined the weights of performance-shaping factors (PSFs) influencing operators' organizational performance. Firstly, a questionnaire survey was conducted, and an interval judgment matrix was innovatively established based on the expectation and standard deviation of statistical data acquired. Next, interval hierarchical analysis (IHA) was employed to derive reasonable intervals for each PSF weight. Finally, a hierarchical interval weighting optimization mathematical model was established for high-precision weighting calculation. This method offers a significant reference for optimizing the PSF weighting process. The approach presented in this paper is highly adaptable and provides new technical and theoretical support for the study of weighting theory. [ABSTRACT FROM AUTHOR]

• Optimization of condenser and pump in the circulating water system is proposed. • The optimization schemes are calculated by the minimum annual cost method. • A sensitivity analysis was carried out on the factors affecting the profits. The optimization of the circulating water system of pressurized water reactor nuclear power plants can effectively improve the profit without affecting the primary circuit operation, which is of great significance in enhancing the market competitiveness of nuclear power plants. The circulating water system of the nuclear power plant is modeled, and the circulating water system optimization scheme is formed after the nuclear power unit is reformed from three aspects: condenser heat transfer area, condenser connection mode, and circulating water pump operation mode. The circulating water system optimization calculation is carried out for each scheme by using the minimum annual cost method. Based on the sensitivity analysis of the factors that affect the profits of nuclear power plants, the best circulating water system optimization scheme is finally given for new nuclear power plants and nuclear power plants in operation, which provides a reference for the construction of related nuclear power projects. [ABSTRACT FROM AUTHOR]

• Thermal performance of AMTEC, TEG, and AMTEC/TEG system is analyzed for nuclear power space stations in space exploration. • The AMTEC/TEG integrated system's thermoelectric conversion performance is determined by the AMTEC subsystem. • The output power and efficiency of the AMTEC/TEG are 4.35% and 15.39% higher as TEG using waste heat from AMTEC condenser. • Provide a new route for constructing long-distance nuclear space power stations and lunar bases in extreme environments. For deep space exploration, the performance of an alkali metal thermoelectric converter (AMTEC) can be improved by utilizing the thermoelectric generator (TEG) subsystem. In this work, a coupling system comprised of an AMTEC and a TEG is proposed, in which the waste heat produced by the AMTEC top cycle is used by the TEG bottom cycle cooled by a heat pipe radiator. A general mathematical model is built to analyze the output power and efficiency of each subsystem and the whole system. Finally, the effect of critical parameters on the output power and efficiency are evaluated, including the current density of AMTEC, dimensionless current i of TEG, geometrical structure, and operating environment. The results show that the combined AMTEC/TEG system is 4.35% and 15.39% higher than the efficiency and output power obtained when using the TEG subsystem to recycle the waste heat from the AMTEC condenser. The thermoelectric conversion performance of the AMTEC/TEG system is related to that of the AMTEC subsystem while less affected by the TEG subsystem. The present work could provide a new route for constructing long-distance nuclear space power stations, rovers, and lunar bases in extreme environments. [ABSTRACT FROM AUTHOR]

The fault diagnosis model is used to diagnose the current state of nuclear power plants (NPPs), thus effectively improving the safety of NPPs. An important challenge is to diagnose faults accurately in the early stage of fault, which requires parameters prediction using a long-term prediction model. In this work, an improved Informer model, i.e., Weights Embedding-Informer (WE-Informer) is proposed for long-term prediction of NPP parameters. Weights Embedding (WE) is designed, so that the key parameters of data can be read by this method, thus reducing the error of long-term prediction. The results show that the minimum mean square error of the prediction model is 0.074. Besides, a fault diagnosis framework based on Stacking is proposed, which is used to diagnose faults based on the prediction results. This framework can effectively diagnose faults in the early stage, which makes it of great significance in the fault diagnosis task of NPPs. [ABSTRACT FROM AUTHOR]

• This study proposes a novel fault detection, reconstruction, and fault-tolerant control architecture specifically designed for sensors in nuclear power plants (NPPs) under both Steady Running Conditions (SRC) and Variable Running Conditions (VRC). Compared to traditional sensor fault detection or reconstruction models, this architecture offers a broader range of application scenarios and practical value. • A Long Short-Term Memory (LSTM) deep neural network sensor fault diagnosis model based on Bayesian Optimization Algorithm (BOA) has been developed. This model is capable of effectively detecting faults under two conditions: load step changes and load slope changes, significantly enhancing the accuracy of signal reconstruction. • Through simulation experiments using a 900 MW nuclear power station as an example, the designed fault-tolerant control architecture has been verified to diagnose and restore the control system to its pre-fault state in the event of sensor failure, demonstrating its effectiveness and practicality. • This research provides solutions for offshore nuclear power stations operating under limited space conditions, expanding the application scope of sensor fault management. The Instrument & control system is the center of the Nuclear Power Plant (NPP), which is equipped with a large number of sensors to ensure the safe and stable running of the NPP. In order to improve the safety of the system, the traditional redundancy method uses the sensor redundancy design. However, due to the limited space of Marine NPP, not all sensors can adopt the traditional redundancy method. In the existing research on sensor fault diagnosis, the sensor fault signal during Variable Running Condition (VRC) cannot be diagnosed and reconstructed. In this study, we focused on solving two problems: (1) training a sensor fault diagnosis model for VRC using Bayesian Optimization (BOA) Long Short Term Memory (LSTM) deep neural network, (2) by using the sensor reconstruction signal in Stable Running Condition (SRC) and the sensor normal signal in VRC, a complete sensor fault-tolerant control system under normal operation of NPP is designed. Finally, taking the 900 MW NPP as an example, experiments were conducted using the NPP and its control system simulation model. The simulation results show that even if a sensor fails during normal operation of NPP, the sensor fault can be isolated, so as to realize fault tolerance control. [ABSTRACT FROM AUTHOR]

To predict the thermal hydraulic behavior in nuclear power plants, traditional system codes such as RELAP, MARS, and CATHARE, designed for one-dimensional two-phase flows, have been employed. However, these codes face limitations in simulating unknown phenomena or new geometries, highlighting the need for advanced three-dimensional simulations for the safety analysis of reactors, especially small modular reactors with passive safety systems. Recent efforts in multidimensional thermal hydraulic analysis have employed various codes for specific scenarios, such as pressurized thermal shock and hydrogen distribution in containment vessels. Consequently, the Korea Atomic Energy Research Institute (KAERI) has developed the CUPID code since 2007, which utilizes a two-fluid, three-field model on unstructured grids, enabling efficient large-scale simulations through MPI-based parallelization. Despite its successes, the computational efficiency of pressure-based solvers, which consume a significant portion of simulation time, is identified as a crucial area for improvement. This study introduces and verifies the meshless Geometric Multi Grid (GMG) technique, a novel approach to enhance the efficiency of solving the pressure correction equation on unstructured grids. The GMG technique, aimed at reducing computational time and increasing scalability, is compared to the conventional Bi-Conjugate Gradient method through various numerical simulations, demonstrating its effectiveness and potential as a superior alternative for thermal hydraulic simulations in nuclear power plant analysis. • Introduces the CUPID code, a three-dimensional simulation tool developed by KAERI, utilizing a two-fluid, three-field model on unstructured grids. • Proposes a novel meshless Geometric Multi Grid (GMG) method, designed to significantly improve the computational efficiency. • Demonstrates through comparative studies that the GMG method substantially reduces computational time compared to Bi-Conjugate Gradient methods. • Confirms through extensive numerical simulations that GMG offers stable convergence on structured and unstructured grids. [ABSTRACT FROM AUTHOR]

• A TD3-NNMPC method based on neural network and TD3 algorithm was designed. • Based on the ε-greedy strategy, the exploration mechanism of TD3 is improved. • TD3-NNMPC solves the problem of dynamic adjustment of control parameters of NNMPC. • Based on TD3, the multi-objective optimization of the control effect was completed. • TD3-NNMPC based on TD3 solves the problem of insufficient robustness of MPC. The power change of Pressurized water reactors (PWRs) is a highly complex nonlinear coupling process, which requires the power controller to have a high nonlinear control capability. Since the traditional model predictive control (MPC) algorithm has the problems of insufficient robustness and adaptivity, this study proposes a neural network-based neural network model predictive control method based on model predictive control and optimizes its control parameters in real time by using the Twin Delayed Deep Deterministic Policy Gradient (TD3) algorithm. The designed control method is validated on the simulation model of the Qinshan nuclear power plant (NPP). Simulation results show that the proposed control method can realize effective control of power under a variety of power changing conditions and dynamic adjustment of control parameters. [ABSTRACT FROM AUTHOR]

• Introducing the Gaussian-Cauchy hybrid mutation into SSA to enhance optimization, shown effective in comparisons with SSA. • Using MDI indicators and K-L divergence in feature extraction for better fault signal representation from sensor data. • Improving SSA for optimizing random forest model parameters, achieving 96.375% accuracy in electric gate valve fault diagnosis. Electric gate valve (EGV) is an essential equipment within nuclear power plant (NPP). This paper presents an advanced fault diagnosis (FD) approach, leveraging Variational Modal Decomposition (VMD), Mutual Dimensionless Indicator (MDI) and the Random Forest (RF) optimized through Improved Sparrow Search Algorithm (ISSA), aimed at improving the accuracy of fault diagnosis and optimizing the FD model during EGV failure events. To commence, we employ the VMD algorithm for modal decomposition of raw electric gate valve signals. This process yields several Intrinsic Mode Function (IMF) components with diverse frequencies, enabling the capture of the underlying dynamics of the signals and facilitating a more comprehensive analysis of the fault conditions. We subsequently apply the K-L divergence to identify key IMF components that closely resemble the original signals. These selected key IMF components serve as the foundation for extracting dimensional indicators (DI) and mutual dimensionless indicators (MDI) as signal features. Furthermore, the Improved Sparrow Search Algorithm (ISSA) is enlisted to optimize the maximum feature count and the number of decision trees in the Random Forest (RF) algorithm. Ultimately, the optimized RF algorithm is deployed for fault diagnosis. Our paper offers a comparative analysis, pitting the VMD method against Empirical Mode Decomposition (EMD) and Local Mean Decomposition (LMD). Additionally, we compare our proposed fault diagnosis model with traditional RF algorithm and the SSA-RF algorithm. Through rigorous experimentation, our results achieved an average fault diagnosis accuracy of up to 96.375%. [ABSTRACT FROM AUTHOR]

• A variable universe fuzzy control method is proposed for OTSG feedwater control. • OTSG pressure and flowrate fuzzy controllers were designed using this method. • The fuzzy controllers provide better steam pressure control that PID controllers. The feedwater control of once-through steam generators (OTSGs) is critical to the safety and stability of nuclear power plants. As traditional PID control methods are difficult to ensure good control of OTSGs under flexible operating conditions, a variable universe fuzzy control (VUFC) method is proposed for OTSG feedwater. The widely-used cascade control structure was first adopted for the OTSG feedwater control system. Then the outer-loop pressure and inner-loop feedwater controllers were designed as fuzzy controllers with their fuzzy universes adjusted adaptively during plant operations. To assess the performance of the VUFC method, dynamic responses of an OTSG employing the fuzzy controllers were compared with those adopting the well-designed PID controllers under typical load change transients. Comparison results illustrate that the VUFC method can provide much better OTSG pressure control with smaller overshoots and shorter settling times, demonstrating the applicability of the VUFC method for OTSGs to provide excellent feedwater control. [ABSTRACT FROM AUTHOR]

• Proposing and developing a Virtual Domain Technology (VDT) scheme that inherits the computational advantages of porous media models while achieving the spatial resolution of CFD for thermal-hydraulics analysis. • The proposed VDT enables flexible creation and adjustment of virtual domains within porous media through localized resistance without changing the mesh. • VDT shows great computational resolution and accuracy. This study verifies the thermal-hydraulics calculation of two-dimensional Karman vortex street and CARR reactor standard fuel assembly. • Thermal-hydraulics calculation for the standard fuel assembly of CARR reactor, compared with the refined CFD, the grid amount of VDT is reduced by 94%, and the calculation time is reduced by 94.303%. Thermal-hydraulics analysis of nuclear reactor core is crucial for the safe operation of nuclear power plants. Porous media models in computational fluid dynamics (CFD) code are commonly used to simulate flow and heat transfer in nuclear reactor cores because they simplify geometry and reduce computational cost. However, their low spatial resolution limits detailed analysis of internal flow and temperature fields. This research develops a Virtual Domain Technology (VDT) scheme to improve the resolution of porous media models without increasing mesh density. VDT sets localized resistance coefficients to create virtual solid and fluid domains within the porous core representation. The flow and thermal effects of these virtual domains mimic real structural effects. VDT was verified on a cylinder flow case and showed excellent agreement with CFD for velocity and pressure fields. Application to a plate-type research reactor core demonstrated VDT's ability to capture detailed temperature profiles and flow traces consistent with CFD, using only 6% as many mesh cells. By improving spatial resolution, VDT enhances porous media-based modeling of reactor cores to provide more accurate thermal-hydraulics analysis, while maintaining computational efficiency. This will aid optimization and safety analysis for advanced reactor designs. [ABSTRACT FROM AUTHOR]

• Modeling and analysis of the entire small pressurized water reactor nuclear power plant were conducted using the MELCOR code. • This paper analyzed the total release quantity and distribution patterns of radioactive aerosols and inert gases. • This paper analyzed the impact of the activation of the containment spray system on the release and distribution of radioactive aerosols and inert gases. The small-sized pressurized water reactor (PWR) characterized by its high inherent safety and wide applicability, has become a promising type of reactor with broad prospects for development. Currently, there is a limited amount of research dedicated to the analysis of accident source terms specific to small-sized pressurized water reactors, this study adopts the ACP100 as a reference to establish a model of a small-sized pressurized water reactor power plant using the severe accident analysis code MELCOR. This study investigated the release and distribution characteristics of radioactive materials under the scenario of a large break loss of coolant accident in the heat pipe section. Specifically, it analyzed the impact of opening and shutting down the containment spray system on the release and distribution of radioactive materials. The results indicate that in the absence of the containment spray system, the majority of radioactive aerosols are primarily distributed in liquid form within the containment, with a small portion depositing on other components and dispersing in the gas phase; inert gases are primarily distributed in the gaseous phase within the containment. With the containment spray system engaged, there is an increase in the total quantity of radioactive aerosols and the concentration of radioactive aerosols in the gaseous phase significantly decreases, while the concentration of liquid-phase radioactive aerosols dramatically increases. The containment spray system exhibits strong capabilities for the removal and retention of radioactive aerosols. While the total release quantity remains largely unchanged, the time to reach the peak release of inert gases is significantly shortened The containment spray system significantly affects the release rate of inert gases. [ABSTRACT FROM AUTHOR]

• Heating experiments were conducted using a small pool test apparatus to investigate thermal stratification in a water pool with a vertical heat source. • Spatial distributions of temperature and velocity in thermal stratification in the small pool were measured using thermography and PIV and analyzed to characterize their spatial distributions. • The mesh size of CFD simulations must be sufficiently small to resolve the flow structure of internal convection, including the thermal stratification boundary. Thermal stratification is a common phenomenon observed in various systems, including bathtubs, lakes, and pools in nuclear facilities. It is considered system-dependent; thus, its occurrence in different systems must be investigated. In this study, heating experiments were performed in a two-dimensional water pool to investigate the thermal stratification mechanism where a source is immersed in a pool, as in spent fuel pools in nuclear power plants. In addition to temperature measurements using thermocouples, the spatial structure of thermal stratification was obtained by visualizing the temperature and velocity fields using thermography and particle image velocimetry (PIV). Computational fluid dynamics (CFD) simulations were also performed as a benchmark analysis using the experimental data, and the simulated and experimental results are in good agreement. [ABSTRACT FROM AUTHOR]

• NPPs must adjust output for load-following to meet power system needs. • The PWR model is built using point kinetics and thermal–hydraulic equations. • Effectiveness of coolant flow rate vs. control rod positioning is investigated. • PID controller is designed and optimized using SQP. • Variable coolant flow rate outperforms control rod positioning. To ensure safety and stability, it is optimal for nuclear power plants to operate at their nominal power level. However, to meet the needs of power systems, NPPs must also be able to adjust their output for load-following purposes. This necessitates finding an effective control algorithm to maintain stable/safe operating conditions. The current study investigates the performance of PID controllers optimized using Sequential Quadratic Programming algorithm and applied using two control techniques: control rod positioning and variable coolant flow rate. Simulation is done using model of four-loop Westinghouse PWR-1200 MW nuclear reactor. The proposed controller's performance is evaluated in three scenarios: sudden control rod withdrawal, sudden change in coolant inlet temperature, and linear load tracking. The influences of model parametric uncertainty are also investigated. Simulation results revealed that the performance of the control algorithm applied using variable coolant flow rate was superior to the one using control rod positioning. [ABSTRACT FROM AUTHOR]

• The CL-4.1-03 SB-LOCA experiment of the PSB-VVER has been analysed with system codes. • APROS and TRACE were able to predict all the relevant physical phenomena. • Modelling approaches and lessons learned from sensitivity analyses are summarised. • Quantitative assessment has been performed using the FFTBM-SM and SARBM methods. • Our results were compared to measurements, to each other, to a reference evaluation. Best-estimate thermal–hydraulic system codes are widely used for the analysis of various postulated events of NPPs, therefore, an extensive verification and validation of these codes is essential. In the current study, an SB-LOCA scenario (CL-4.1-03) of the PSB-VVER integral test facility has been investigated using NRC's novel system code, TRACE, and the Finnish APROS, which had already been in long use at our Institution. PSB-VVER is currently the most comprehensive physical model of the VVER-type reactors, particularly VVER-1000, in most of the aspects. The experiment under study was carried out in 2004 in the framework of an OECD NEA benchmark. The obtained results have been compared to the benchmark and to each other both from a qualitative and quantitative point of view. As a figure of merit, the improved FFTBM-SM and the SARBM quantitative methods were used. Furthermore, the main lessons learnt during the model development are also presented. [ABSTRACT FROM AUTHOR]

• During the accident, the total hydrogen release was 390 kg, with a maximum release rate of 0.9 kg/s. • The diffusion path of hydrogen in the containment is "source compartment → top space of the containment → bottom compartment". • The absence of hydrogen risk in most containment compartments demonstrates the effectiveness of the CHC system. • Post-inerting can reduce hydrogen risk in containment at severe accident. Limited HPR1000 operational data under accident conditions emphasize the importance of analyzing its Containment Hydrogen Combination System (CHC) for improving nuclear safety. GASFLOW is employed to simulate a large break loss of coolant accident (LB-LOCA) in the steam generator's hot leg section, assuming the Safety Injection System's active injection fails. Results show that the movement of hydrogen within the containment follows the pathway 'source compartment → top space of the containment → bottom compartment.' Initially, a distinct stratification of hydrogen occurs throughout the containment. Subsequently, the stratification phenomenon begins to manifest in the source compartment and the lower areas of the containment. The CHC system effectively mitigates the risk of hydrogen, prevents flame acceleration and the likelihood of deflagration-to-detonation transitions in most compartments. However, the source compartment still poses a risk. Additionally, injecting inert gas into the source compartment is an effective method to control the hydrogen risk within the HPR1000. [ABSTRACT FROM AUTHOR]

• A mathematical model is built for the prediction of operation temperature and RH of exhaust moist gas for tritiated water transportation. • The relationship between the curve of saturated moist air and the line linked atmospheric point and operation point illustrate the occurrence of condensation. • The critical temperature of the exhaust gas emit under saturation is obtained by the calculation according to the tangent line from the atmospheric point and is tangent to the fitting curve of saturated moist air. A mathematical model is built for the prediction of operation parameters for tritiated water transferred by air from nuclear reactors. Exhaust moist gas containing tritiated water vapor is emitted and diffused into the atmosphere. It should be ensuring that no condensate water formation during the mixture of exhaust moist gas and atmosphere to avoid local radiation. The occurrence of water condensation during the mixture of exhaust moist gas and atmospheric air is analyzed according to a fitting equation of saturated moist gas in a psychrometric chart. The mathematical model which predict operation temperature and RH of exhaust moist gas for certain atmospheric temperature and RH is built. The mathematical model is verified in CFD for five typical operating conditions. The effect of atmosphere temperature/RH on operation temperature and RH is analyzed. The operation strategy for tritiated water transferred by air under different atmospheric conditions is discussed and presented. [ABSTRACT FROM AUTHOR]

• The part of the failure criteria and uncertainty design parameter has been rewritten • The part of the results has been redone and the computational time has been supplemented. • The part of the conclusion has been rewritten. • The test of the improved kriging model of the passive residual system has been redone. Accurately quantify the failure probability of passive safety systems can promote the application of passive safety technology in nuclear power plants. However, the complex and variable marine conditions introduce more uncertainty to the operation of passive safety systems, posing additional challenges to the reliability analysis of passive safety systems under marine conditions. Furthermore, huge calculation cost is one of the problems in the reliability assessment of passive safety system. In response to these limitations of the reliability assessment, a reliability analysis method for passive safety systems is proposed based on the improved Kriging model and SORM method in this paper. It is applied to the reliability analysis of the passive residual heat removal systems of the marine IPWR200. The calculation results show that the failure probability of the passive residual heat removal systems of the IPWR200 under marine conditions is approximately 2.9308 × 10-4. Compared to the Monte Carlo method, the method which coupling with improved Kriging and SORM can reduce the computational cost of reliability analysis while ensuring the accuracy of the results. The sensitivity analysis result show that the marine movements are the major uncertainty parameters which is significant for system performance. The marine movements significantly impact the driving force of natural circulation and generate a periodic additional force which lead to the periodic oscillation of natural circulation. [ABSTRACT FROM AUTHOR]

Severe accident scenarios for nuclear power plants are determined through probabilistic safety analysis (PSA). In this process, it is possible to identify the failure sequence of specific components, but assessing the impact of component failure time on the severity of accident remains a challenge. In this study, a novel approach is presented that utilizes machine learning methodologies such as reinforcement learning (RL) to complement traditional PSA. The proposed process was validated by comparing whether the most severe accident scenarios obtained through critical accident simulations can be reproduced by a RL, since this is a novel use of machine learning techniques. The comparison shows feasibility of exploring critical accident scenarios using RL. To implement the reinforcement learning methodology based on the existing system code, supervised learning model that can predict the remaining time of reactor vessel failure was implemented in this study. Based on this prediction model and data from existing catastrophic accident simulation, RL was implemented. The results obtained from RL were subsequently validated with the results of severe accident code simulation. In summary, new methodology for applying machine learning techniques to the nuclear accident analysis process was presented, and the feasibility and potential of the proposed methodology were discussed. [ABSTRACT FROM AUTHOR]

• Simulations during normal operations showed higher concentrations and ground depositions in Veracruz, Mexico. • The probability of exceeding screening doses was acceptably low (<5%), indicating minimal cancer risk from controlled releases. • During nuclear accidents, dispersed plumes were observed with directions varying depending on wind speed. • Veracruz recorded highest air concentrations, depositions, and radiation doses among all sites during the incident. • Study insights: Minimal controlled release impacts, potential accident risks; valuable for stakeholders. Monitoring and regulating radioactive emissions from nuclear power plants is important for evaluating and managing the associated radiological risks to humans and the environment. Radioactive emissions can occur during different stages of nuclear power generation, posing concerns about exposure to ionizing radiation. Modeling, particularly using advanced simulation techniques, is a crucial tool for assessing both routine and accidental emissions. This work assesses the potential environmental and radiological impacts of radioactive emissions from the Laguna Verde nuclear power plant (Veracruz, México) using the CALPUFF model for dispersion and the ERICA tool for ecological risk assessment. Normal operations simulations showed that the higher concentrations and ground depositions were observed in Veracruz. The probability of exceeding screening doses was acceptably low (<5%), indicating minimal cancer risk from controlled releases. During nuclear accidents, dispersed plumes were observed with varying directions depending on wind speed. The study identified potential risks to five designated cities as receptor sites, with some plume sweeps diverting pollutants. The most affected receptor site (Veracruz) recorded the highest air concentrations, depositions, and radiation doses. These findings offer valuable insights for regulators, decision-makers, and the public, emphasizing low impacts from controlled releases but highlighting potential risks during nuclear accidents. [ABSTRACT FROM AUTHOR]

• A large-scale 6-DOF visual liquid level measurement system was established. • A theoretical model for pressure tube-type measuring instrument under ocean motions was proposed. • A new differential pressure level measurement scheme was proposed and verified by experiment. • An additional pressure drop model was used to further reduce the measurement deviation. • A Kalman filter for liquid level measurement under ocean motions was proposed and verified by experiment. The high-power density of the floating nuclear power plant (FNPP) makes it suitable for marine resource development and coastal island construction. However, the complex motion of the floating plant under actual ocean motions poses a challenge for the use of pressure tube-type measuring instruments, which can lead to significant measurement deviation and inaccurate predictions of the nuclear reactor system's running state. To address this issue, this study employed the optimization of differential pressure level measurement technology for free level equipment under ocean motions as an illustrative case, utilized the theoretical analysis of gravity pressure drop under ocean motions to optimize the pressure tube layout scheme, and established a motion additional pressure drop model to reveal the variation of differential pressure and further correct the measurement deviation. In addition, combined with the pressure drop theory model, the applicability of Kalman filter under ocean motions is developed. The results showed that the optimized differential pressure level measurement scheme based on gravity pressure drop can reduce the measurement deviation to 3% under most ocean motions. Moreover, the motion additional pressure drop model further reduced the deviation of the measurement to 0.5%. Furthermore, the Kalman filter modified liquid level measurement scheme significantly reduced the measurement deviation. This complete correction scheme opens a window for the optimized design of pressure pipe-type measuring instrument under ocean motions. [ABSTRACT FROM AUTHOR]

• The development of a fault diagnosis and localization system for equipment in nuclear power plants was achieved under high-dimensional feature parameter conditions. • To address the issue of high-dimensional data in nuclear power plants, which affects the effectiveness of subsequent fault diagnosis and localization, the utilization of PCA for dimensionality reduction of the collected data was proposed. • To overcome the problem of gradient vanishing in CNN, which leads to network degradation and impacts fault diagnosis performance, the adoption of ResNet for fault diagnosis was proposed. • To address the limitations of AE in capturing temporal relationships, such as gradient vanishing, information blurring, and loss, the utilization of LSTM-AE for fault localization was proposed. Nuclear energy, as a clean energy source, has advantages such as environmental friendliness and low consumption. However, it also carries the drawback of potential radioactive leaks. The impact of radioactive leaks from nuclear power plants on the environment and humans is far greater than that of other types of power plants. Therefore, there are higher requirements for safety in nuclear power plants. Prompt and accurate identification of equipment failures, as well as precise fault localization, can assist operators and maintenance personnel in taking appropriate actions to prevent further deterioration and improve the economic and safety aspects of nuclear power plants. To address these issues, this study has developed a nuclear power plant equipment fault diagnosis and fault localization system. Firstly, considering the large number of sensors in nuclear power plants and the high dimensionality of the collected data, which can affect the effectiveness of subsequent fault diagnosis and localization models, the use of Principal Component Analysis (PCA) for dimensionality reduction of high-dimensional feature parameters is proposed. Then, a Residual Network (ResNet) is employed for fault classification. For fault localization, a Long-Short Term Memory Auto-Encoder (LSTM-AE) is used. The experimental results demonstrate that the developed system achieves high accuracy in nuclear power plant equipment fault diagnosis and fault localization. [ABSTRACT FROM AUTHOR]

• Decay Heat Removal is an important aspect of nuclear Reactor Design. • Need to keep fuel and structure temperatures within limits. • Designs to ensure proper decay heat removal in sodium cooled fast reactors. • Needed under loss of grid power and station power. • Studies and experiments in support of decay heat removal. Shutdown systems and decay heat removal systems form the backbone of the nuclear plant protection system. While the former ensures safe shutdown of the fission reaction, the latter is essential to remove the heat from the decay of the fission products during earlier fissions. Thus, heat continues to be generated even after shutdown. Residual or decay heat removal (DHR) systems are needed to ensure that fuel clad temperatures do not rise beyond limits after the reactor shutdown. In fast reactors which have higher power density in the core, decay heat removal becomes very important to keep the core temperatures and structures within acceptable limits as otherwise fuel failure can result with its attendant consequences of adding radioactivity to the primary coolant. Also, in situations of loss of off-site and on-site power, this heat needs to be removed. Toward this, one has to ensure that the design is amenable to natural circulation cooling in different situations. The primary coolant system in Sodium cooled fast Reactors (SFR) can easily be configured to provide natural circulation shutdown heat removal. Different reactors designed/built/operated have resorted to different ways of decay heat removal. This paper traces the evolution of different decay heat removal options used by different fast reactor designs in different countries. Also presented is a brief review of some of the studies carried out on the efficacy of the different DHR systems besides parametric investigations carried out in different countries, backed up by experiments. The paper makes important observations on the type of DHR system to be adopted based on the above studies. [ABSTRACT FROM AUTHOR]

Intelligent process monitoring has shown significant promise in ensuring the operational safety of nuclear power plants. However, once an anomaly is detected, the localization of fault-relevant sensors can be quite challenging due to the cross-dimensional smearing effect caused by anomalous data. In this work, an autoencoder model is introduced for anomaly detection of nuclear power operational data. To avoid interference from anomalous data on identifying fault variables, a modified loss function utilizing the maximum correntropy criterion is employed in a two-phase model fine-tuning approach to enhance the model's robustness and achieve accurate anomaly isolation. Experimental results on the synthetic dataset demonstrate that the model fine-tuning approach can obviously alleviate the interference from anomalous data, the mean F1 score is improved by up to 0.3 compared to the original AE model. Further anomaly contribution analysis has verified the model's capability to identify the crucial variables responsible for detected anomalies. • A two-phase model fine-tuning approach is proposed to achieve anomaly isolation. • Smearing effect is mitigated utilizing model fine-tuning based on MCC loss function. • Anomaly detection and isolation is achieved by a unified two-stage fine-tuning model. [ABSTRACT FROM AUTHOR]

The Essential Service Water System (SEC) is a nuclear safety level system in nuclear power plants. It's responsible for transporting the heat load collected by the equipment cooling water system (RRI) to the final heat sink – seawater. When the SEC system equipment is damaged, it can lead to ineffective heat transfer in RRI system, ultimately resulting in a major accident. Therefore, this article focuses on the SEC system of nuclear power plants and a technical solution combining model similarity experiments and theoretical analysis was adopted to restore the pressure fluctuation phenomenon that occurred after long-term pump shutdown of the SEC system, and the mechanism of pressure fluctuation phenomenon after long-term pump shutdown of the SEC system was analyzed. The results indicate that: the mechanism of pressure fluctuations in header of the nuclear power plant after long-term pump shutdown is the combined effect of an increase in the volume of the air chamber in the system and leakage in the check valve. In the design process of the unit, it is possible to consider reducing the high point position reasonably to improve the negative pressure situation of the SEC system, and it is recommended to replace the butterfly check valve with a two-stage slow closing check valve. This will provide certain experience and technical support for the further development of the SEC system. [ABSTRACT FROM AUTHOR]