• Journal of the Korean Society for Nondestructive Testing
• /
• v.18 no.4
• /
• pp.321-329
• /
• 1998

Structural integrity of nuclear components is important for a safe operation of nuclear power plants. Therefore, nuclear power plants require to perform reliable, periodic inservice inspections. Korea Electric Power Company(KEPCO) operates the entire Korean nuclear power plants. Since nuclear power plant safety and the associated inservice inspection(ISI) are under the plant owner's responsibility, Korea Electric Power Research Institute(KEPRI), the R&D division of KEPCO, has established the ISI NDE Total Support system(TSS) for an efficient performance of ISI tasks, and initiated both key ISI NDE technology development program and traing & qualification system development program for an independent ISI operation. This paper describes details of these programs.

• Nuclear Engineering and Technology
• /
• v.46 no.6
• /
• pp.737-752
• /
• 2014

As condition-based maintenance (CBM) has risen as a new trend, there has been an active movement to apply information technology for effective implementation of CBM in power plants. This motivation is widespread in operations and maintenance, including monitoring, diagnosis, prognosis, and decision-making on asset management. Thermal efficiency analysis in nuclear power plants (NPPs) is a longstanding concern being updated with new methodologies in an advanced IT environment. It is also a prominent way to differentiate competitiveness in terms of operations and maintenance costs. Although thermal performance tests implemented using industrial codes and standards can provide officially trustworthy results, they are essentially resource-consuming and maybe even a hind-sighted technique rather than a foresighted one, considering their periodicity. Therefore, if more accurate performance monitoring can be achieved using advanced data analysis techniques, we can expect more optimized operations and maintenance. This paper proposes a framework and describes associated methodologies for in-situ thermal performance analysis, which differs from conventional performance monitoring. The methodologies are effective for monitoring, diagnosis, and prognosis in pursuit of CBM. Our enabling techniques cover the intelligent removal of random and systematic errors, deviation detection between a best condition and a currently measured condition, degradation diagnosis using a structured knowledge base, and prognosis for decision-making about maintenance tasks. We also discuss how our new methods can be incorporated with existing performance tests. We provide guidance and directions for developers and end-users interested in in-situ thermal performance management, particularly in NPPs with large steam turbines.

• Nuclear Engineering and Technology
• /
• v.56 no.2
• /
• pp.558-567
• /
• 2024

Diagnostic support systems are being researched to assist operators in identifying and responding to abnormal events in a nuclear power plant. Most studies to date have considered single abnormal events only, for which it is relatively straightforward to obtain data to train the deep learning model of the diagnostic support system. However, cases in which multiple abnormal events occur must also be considered, for which obtaining training data becomes difficult due to the large number of combinations of possible abnormal events. This study proposes an approach to maintain diagnostic performance for multiple abnormal events by training a deep learning model with data on single abnormal events only. The proposed approach is applied to an existing algorithm that can perform feature selection and multi-label classification. We choose an extremely randomized trees classifier to select dedicated monitoring parameters for target abnormal events. In diagnosing each event occurrence independently, two-channel convolutional neural networks are employed as sub-models. The algorithm was tested in a case study with various scenarios, including single and multiple abnormal events. Results demonstrated that the proposed approach maintained diagnostic performance for 15 single abnormal events and significantly improved performance for 105 multiple abnormal events compared to the base model.

• Proceedings of the Korean Institute of Building Construction Conference
• /
• 2013.05a
• /
• pp.271-272
• /
• 2013

The Korean nuclear industry acquired technology for each construction stage including engineering, procurement, construction and commissioning from advanced nuclear countries. While the levels of technology and quality have greatly improved, the same cannot be said for the level of project management. In particular, the level of project performance measurement and forecasting project risk still remain at the single project management level. Thus, this paper reviewed the concept of the EVMS method and requirement for the system. The adoption of the EVMS, an advanced project management method, can enable efficient management of project risks and promote an adequate environment for project implementation.

• Nuclear Engineering and Technology
• /
• v.28 no.1
• /
• pp.36-43
• /
• 1996

In this work, a PC-based thermal performance monitoring system is developed for the nuclear power plants. The system performs real-time thermal performance monitoring and diagnosis during plant operation. Specifically, a prototype for the KORI-2 nuclear power unit is developed and examined in this work. The analysis and the fault identification of the thermal cycle of a nuclear power plant is very difficult because the system structure is highly complex and the components are very much inter-related. In this study, some major diagnostic performance parameters are selected in order to represent the thermal cycle effectively and to reduce the computing time. The Fuzzy ARTMAP, a self-organizing neural network, is used to recognize the characteristic pattern change of the performance parameters in abnormal situation. By examination, this algorithm is shown to be able to detect abnormality and to identify the fault component or the change of system operation condition successfully. For the convenience of operators, a graphical user interface is also constructed in this work.

• Nuclear Engineering and Technology
• /
• v.37 no.5
• /
• pp.491-502
• /
• 2005

Reliable human performance is a prerequisite in securing the safety of complicated process systems such as nuclear power plants. However, the amount of available knowledge that can explain why operators deviate from an expected performance level is so small because of the infrequency of real accidents. Therefore, in this study, a database that contains a set of useful information extracted from simulated emergencies was developed in order to provide important clues for understanding the change of operators' performance under stressful conditions (i.e., real accidents). The database was developed under Microsoft Windows TM environment using Microsoft Access $97^{TM}$ and Microsoft Visual Basic $6.0^{TM}$. In the database, operators' performance data obtained from the analysis of over 100 audio-visual records for simulated emergencies were stored using twenty kinds of distinctive data fields. A total of ten kinds of operators' performance data are available from the developed database. Although it is still difficult to predict operators' performance under stressful conditions based on the results of simulated emergencies, simulation studies remain the most feasible way to scrutinize performance. Accordingly, it is expected that the performance data of this study will provide a concrete foundation for understanding the change of operators' performance in emergency situations.

• Nuclear Engineering and Technology
• /
• v.45 no.5
• /
• pp.597-604
• /
• 2013

Containment Filtered Vent Systems (CFVSs) have been mainly equipped in nuclear power plants in Europe and Canada for the controlled depressurization of the containment atmosphere under severe accident conditions. This is to keep the containment integrity against overpressure during the course of a severe accident, in which the radioactive gas-steam mixture from the containment is discharged into a system designed to remove the radionuclides. In Korea, a CFVS was first introduced in the Wolsong unit-1 nuclear power plant as a mitigation measure to deal with the threat of over pressurization, following post-Fukushima action items. In this paper, the overall features of a CFVS installation such as risk assessments, an evaluation of the performance requirements, and a determination of the optimal operating strategies are analyzed for the Wolsong unit 1 nuclear power plant using a severe accident analysis computer code, ISAAC.

• Nuclear Engineering and Technology
• /
• v.51 no.2
• /
• pp.453-462
• /
• 2019

Mental workload (MWL) is a major consideration for the user interface design in nuclear power plants (NPPs). However, each MWL evaluation method has its advantages and limitations, thus the evaluation and control methods based on multi-index methods are needed. In this study, fuzzy comprehensive evaluation (FCE) theory was adopted for assessment of interface designs in NPP based on operators' MWL. An evaluation index system and membership functions were established, and the weights were given using the combination of the variation coefficient and the entropy method. The results showed that multi-index methods such as performance measures (speed of task and error rate), subjective rating (NASA-TLX) and physiological measure (eye response) can be successfully integrated in FCE for user interface design assessment. The FCE method has a correlation coefficient compared with most of the original evaluation indices. Thus, this method might be applied for developing the tool to quickly and accurately assess the different display interfaces when considering the aspect of the operators' MWL.

• Nuclear Engineering and Technology
• /
• v.53 no.7
• /
• pp.2396-2406
• /
• 2021

The Pacific Earthquake Engineering Research (PEER) Center has been developing a performance-based earthquake engineering (PBEE) methodology, which is based on explicit determination of performance, e.g., monetary losses, in a probabilistic manner where uncertainties in earthquake ground motion, structural response, damage estimation, and losses are explicitly considered. To carry out the PEER PBEE procedure for a component of the nuclear power plant (NPP) such as the cable tray system, hazard curve and spectra were defined for two hazard levels of the ground motions, namely, operation basis earthquake, and safe shutdown earthquake. Accordingly, two sets of spectral compatible ground motions were selected for dynamic analysis of the cable tray system. In general, the PBEE analysis of the cable tray in NPP was introduced where the resulting floor motions from the time history analysis (THA) of the NPP structure should be used as the input motion to the cable tray. However, for simplicity, a finite element model of the cable tray was developed for THA under the effect of the selected ground motions. Based on the structural analysis results, fragility curves were generated in terms of specific engineering demand parameters. Loss analysis was performed considering monetary losses corresponding to the predefined damage states. Then, overall losses were evaluated for different damage groups using the PEER PBEE methodology.

• Nuclear Engineering and Technology
• /
• v.56 no.6
• /
• pp.2266-2273
• /
• 2024

A machine learning platform is proposed for the diagnosis of a severe accident progression in a nuclear power plant. To predict the key parameters for accident management including lost signals, a long short term memory (LSTM) network is proposed, where multiple accident scenarios are used for training. Training and test data were produced by MELCOR simulation of the Fukushima Daiichi Nuclear Power Plant (FDNPP) accident at unit 3. Feature variables were selected among plant parameters, where the importance ranking was determined by a recursive feature elimination technique using RandomForestRegressor. To answer the question of whether a reduced order ML model could predict the complex transient response, we performed a systematic sensitivity study for the choices of target variables, the combination of training and test data, the number of feature variables, and the number of neurons to evaluate the performance of the proposed ML platform. The number of sensitivity cases was chosen to guarantee a 95 % tolerance limit with a 95 % confidence level based on Wilks' formula to quantify the uncertainty of predictions. The results of investigations indicate that the proposed ML platform consistently predicts the target variable. The median and mean predictions were close to the true value.