Expert systems for emergency alarms analysis during accident situations in nuclear reactors

Expert system development is an important application of artificial intelligence. During the last few years, many successful expert systems have been developed in various fields like medical diagnosis, geological exploration, office management, etc. Expert systems is a computer software which solves reasonably complex problems where normally one needs an expert in that field to solve them.Recently efforts are being made in developing expert systems to support the operating staff in nuclear reactors. Nuclear power plant is one of the most complex engineering systems and safe and reliable operation is of primary importance. In spite of many automatic and redundant safety systems there are some occasions when the operating staff have to analyse the alarms and take further safety actions. Few of the severe accidents like Three Mile Island in USA and Chernobyl in USSR are attributed to the errors made by the operating staff and/or management. Nuclear engineers or systems analysts who may be expert in analysing an accident situation and advise corrective safety actions may not be readily available during the accident situations in the nuclear power plants. It is possible to model such expert knowledge in expert systems and this can be applied in diagnosing an accident situation like the loss of coolant accident (LOCA) and act as an additional confirmatory aid to the operating staff.Two small expert systems examples have been developed and are explained in this paper. One identifies a spurious LOCA alarm in a heavy water research reactor. The second example identifies the type of medium/small leakage (LOCA) in a coolant circuit of a PWR and suggests the subsequent safety actions. Both the examples have been developed using the expert systems shell VP-expert. They are off-line usable and user interactive. We do not propose expert system application for very fast response safety actions like reactor scram. These two small expert system examples are essentially to support the feasibility study in their applications during accident situations in nuclear power plants.     

LOCA工况下锆合金包壳的行为概述

LOCA作为反应堆运行过程中比较严重的事故,是反应堆基准设计事故;而作为确保裂变产物不泄露的第一道屏障,锆合金优异的性能对于保障LOCA工况下的核安全具有重要意义。阐述了LOCA工况下锆合金的高温氧化行为、抗热冲击性能和力学性能及显微组织等方面的内容,为反应堆用锆合金的研发提供了技术支持。     

An operator support system for research reactor operations and fault diagnosis through a connectionist framework and PSA based knowledge based systems

During reactor upset/abnormal conditions, emphasis is placed on the plant operator's ability to quickly identify the problem and perform diagnosis and initiate recovery action to ensure the safety of the plant. However, the reliability of human action is adversely affected at the time of crisis due to time stress and psychological factors. The availability of operational aids capable of monitoring the status of the plant and quickly identifying the deviation from normal operation is expected to significantly improve the operator reliability.The development of operator support systems using probabilistic safety assessment (PSA) techniques and information is finding wide application in nuclear plant operation. Often it is observed that most of the applications use a rule-based approach for diagnosis as well as safety status/transient conditions monitoring. A more efficient approach using artificial neural networks for safety status/transient condition monitoring and rule-based systems for diagnosis and emergency procedure generation has been applied for the development of a prototype operator adviser (OPAD) system for a 100 MW(th) heavy water moderated, cooled and natural uranium fueled research reactor. The development objective of this system is to improve the reliability of operator action and hence the reactor safety at the time of crisis as well as in normal operation. In order to address safety objectives at various stages of development of OPAD, the PSA techniques and tools have been used for knowledge representation. It has been demonstrated, with recall tests on the artificial neural network, that it can efficiently identify the reactor status in real-time scenario. This paper discusses various issues related to the development of an operator support system in a comprehensive way, right from the study of safety objectives, to data collection, to implementation of such a system.     

Representing context, cognition, and crew performance in a shutdown risk assessment

This paper presents a potentially practical treatment of dynamic operator-system interactions. The approach employs a dynamic event tree framework to explicitly address plant dynamics, system indications, crew-plant interactions, time available, crew cognition, errors of commission (mistakes), and multiple planning and diagnosis possibilities. The approach is applied in an analysis of a hypothesize medium break loss of coolant accident for a test reactor that occurs during plant shutdown. This transient was selected on the basis of: a significant cognitive component being present, high consequences being associated with operator actions, and the importance of event timing to scenario progression. The results presented show how quantitative risk predictions are affected by the treatment of dynamics, and demonstrate how non-proceduralized recovery actions and a number of performance shaping factors (e.g., crew experience, stress, and confidence) can be explicitly treated. Insights and lessons learned regarding the performance of a dynamic assessment are also presented.     

On the chemical compatibility between Zr-4 substrate and well-bonded Cr2AlC coating

Accident tolerant fuel(ATF) for the light water reactor has gained wide attentions after the Fukushima accident. To enhance the accident-tolerance of the nuclear system, one strategy is to modify the Zr-based alloy cladding surface with advanced ceramic coating. In this work, monolithic and dense Cr_2AlC coating has been synthesized by magnetron sputtering. The as-grown Cr_2AlC coating exhibits good chemical compatibility with Zr-based alloy substrate as well as mechanical integrity under both pull-off and scratch tests. The coating system also presents moderate thermochemical compatibility at 800℃ but degrades above 1000℃ under simulated loss-of-coolant accident(LOCA) conditions.     

Application of artificial neural networks to nuclear power plant transient diagnosis

A study on various artificial neural network (ANN) algorithms for selecting a best suitable algorithm for diagnosing the transients of a typical nuclear power plant (NPP) is presented. NPP experiences a number of transients during its operations. These transients may be due to equipment failure, malfunctioning of process systems, etc. In case of any undesired plant condition generally known as initiating event (IE), the operator has to carry out diagnostic and corrective actions. The objective of this study is to develop a neural network based framework that will assist the operator to identify such initiating events quickly and to take corrective actions. Optimization study on several neural network algorithms has been carried out. These algorithms have been trained and tested for several initiating events of a typical nuclear power plant. The study shows that the resilient-back propagation algorithm is best suitable for this application. This algorithm has been adopted in the development of operator support system. The performance of ANN for several IEs is also presented.     

Symptom based diagnostic system for nuclear power plant operations using artificial neural networks

Nuclear power plant experiences a number of transients during its operations. These transients may be due to equipment failure, malfunctioning of process systems and unavailability of safety systems. In such a situation, the plant may result into an abnormal state which is undesired. In case of an undesired plant condition generally known as an initiating event (IE), the operator has to carry out diagnostic and corrective actions. The operator's response may be too late to mitigate or minimize the negative consequences in such scenarios. The objective of this work is to develop an operator support system based on artificial neural networks that will assist the operator to identify the IEs at the earliest stages of their developments. These abnormal plant conditions must be diagnosed and identified through the process instrument readings. A symptom based diagnostic system has been developed to investigate the IEs. The event identification is carried out by using resilient back propagation neural network algorithm. Whenever an event is detected, the system will display the necessary operator actions in addition to the type of IE. The system will also show the graphical trend of relevant parameters. The developed system is able to identify the eight IEs of Narora Atomic Power Station. This paper describes the features of the diagnostic system taking one of the IEs as a case study.     

Level-1 probability safety assessment of the Iranian heavy water reactor using SAPHIRE software

The main goal of this review paper is to analyze the total frequency of the core damage of the Iranian Heavy Water Research Reactor (IHWRR) compared with standard criteria and to determine the strengths and the weaknesses of the reactor safety systems towards improving its design and operation. The PSA has been considered for full-power state of the reactor and this article represents a level-1 PSA analysis using System Analysis Programs for Hands-On Integrated Reliability Evaluations (SAPHIRE) software. It is specifically designed to permit a listing of the potential accident sequences, compute their frequencies of occurrence and assign each sequence to a consequence. The method used for modeling the systems and accident sequences, is Large Fault Tree/Small Event Tree method. This PSA level-1 for IHWRR indicates that, based on conservative assumptions, the total frequency of accidents that would lead to core damage from internal initiating events is 4.44E−05 per year of reactor operation.     

A systematic approach to analysing errors of commission from diagnosis failure in accident progression

The accident scenarios of a nuclear power plant are composed of an initiating event (IE), additional events/failures and human inappropriate actions, the combinations of which lead to irreversible consequences. In such a dynamic situation, operators should diagnose the occurring events/failures (including an initiating event and additional events) and assess the related situations utilising the available resources such as operating procedures or human–machine systems to control and maintain the plant in a stable condition. The misdiagnosis or diagnosis failure of the occurring events could cause critical human inappropriate actions that aggravate the plant condition, which is termed as errors of commission (EOCs). This paper presents a methodology for analysing the potential for diagnosis failure of the initiating and additional events and the consequent EOC events, based on the operating procedures, in the accident scenarios of nuclear power plants. The method to be presented categorizes the diagnostic situations in the accident scenarios into three cases according to the structure of the emergency operating procedures (EOPs) and the time of the occurring events: (1) the diagnosis of an initiating event, (2) the diagnosis of both an initiating event and an additional event when an additional event occurs prior to the performance of the diagnosis procedure, and (3) the diagnosis of an additional event when an additional events occurs after the performance of the diagnosis procedure. The application of the method is illustrated through three case example scenarios: (1) the power-operated relief valve (PORV) or the pressurizer safety valve (PSV) LOCA, (2) the loss of all feedwater (LOAF) event (loss of main feedwater*loss of auxiliary feedwater), (3) the sequence of<the station blackout (SBO)*loss of turbine-driven (or diesel-driven) auxiliary feedwater*PSV stuck-open*recovery of AC power>.     

The impact of assessing lower head integrity on boiling water reactor probabilistic risk assessments

Over the last several decades, much effort has been directed at estimating the likelihood of a large early release of radioactivity during a nuclear accident. This effort has culminated in the Individual Plant Examinations (IPEs) for the over 100 US nuclear power plants and the NUREG 1150 study. The large early release of radioactivity requires core damage with loss of primary containment integrity during the accident. Given a successful reactor scram, early containment failure coupled with a large release of radioactivity will only occur if the reactor core vessel is breached by core debris. Most IPE/PRA studies performed to date have not considered the possibility of quenching core debris in the lower plenum. Consequently, lower head failure is presumed to closely follow the onset of core damage. Therefore, these assessments did not address the role that in-vessel debris retention plays in preserving primary containment integrity, nor do they propose a criterion for evaluating the integrity of the vessel lower head given that core damage has occurred. Yet preserving the vessel lower head integrity is a necessary condition for satisfying the plant design and licensing basis. Therefore, a more complete treatment of the risk associated with nuclear plant operation includes an evaluation of the ability to retain the core debris in-vessel. This paper presents a performance requirement for vessel integrity to be used in probabilistic risk assessments; evaluates the impact the core damage progression and lower plenum quenching models have on the likelihood of terminating the damage progression in-vessel; documents the significant reduction in BWR containment failure probability that can occur when appropriate core damage and lower head quenching models are used; reviews the implications of core debris quenching in the lower head on BWR PRA modeling; argues why crediting the capability to maintain vessel integrity is necessary from a safety point of view. These results and conclusions are derived from consideration of a BWR 4 plant with a 251 inch vessel. However, the concepts are generally applicable and results specific to other BWR designs can be developed using the methodology presented in this paper.     

Research on the prediction of LOCA condition in nuclear power plants based on GRU recurrent neural network and its variants

Considering the development of deep learning and the emergence of intelligent control demands in nuclear reactors, along with the presence of plant-level real-time information monitoring systems in most nuclear power plants, there is a considerable accumulation of sensor measurements from long-term operation. This makes it feasible to conduct medium to long-term predictions for various real-time conditions in nuclear power plants. Therefore, this paper proposes the utilization of a gate-based recurrent neural network called GRU (Gated Recurrent Unit) and its variants for parameter prediction of LOCA (Loss of Coolant Accident) scenarios. The main content of this paper consists of two parts: (1) Experimental verification is conducted to demonstrate that GRU has excellent capability in capturing long-term sequential information and generalization ability, making it suitable for predicting accident conditions in nuclear power plants. Two accident trend prediction methods based on the GRU network are proposed for scenarios with limited data. The results show that these methods can effectively provide short-term development trends for accident conditions. Additionally, by considering the feature extraction capacity of CNN, the fusion of CNN and GRU models is employed for parameter prediction under different sizes of broken area. The results indicate an improvement in the model's generalization ability. (2) In scenarios with limited and incomplete data, a more robust variant of GRU called GRU-D model is utilized for both univariate and multivariate synchronous prediction of accident conditions with different missing values. Experimental results demonstrate that even with a data missing rate of 90%, the GRU-D network exhibits excellent predictive accuracy and generalization ability in parameter prediction for the given conditions.     

An artificial neural network-based expert system for the appraisal of two-car crash accidents

This paper employs artificial neural network (ANN) to develop an accident appraisal expert system. Two ANN models-- party-based and case-based-- with different hidden neurons are trained and validated by k-fold (k=3) cross validation method. A total of 537 two-car crash accidents (1074 parties involved) are randomly and equally divided into three subsets. For the comparison, a discrimination analysis (DA) model is also calibrated. The results show that the ANN model can achieve a high correctness rate of 85.72% in training and 77.91% in validation and a low Schwarz's Bayesian information criterion (SBC) of -0.82 in training and 0.13 in validation, which indicates that the ANN model is suitable for accident appraisal. Furthermore, in order to measure the importance of each explanatory variable, a general influence (GI) index is computed based on the trained weights of ANN. It is found that the most influential variable is right-of-way, followed by location and alcoholic use. This finding concurs with the prior knowledge in accident appraisal. Thus, for the fair assessment of accident liabilities the correctness of these three key variables is of critical importance to police investigation reports.     

Wind turbine fault detection based on SCADA data analysis using ANN

Wind energy is one of the fast growing sources of power production currently, and there is a great demand to reduce the cost of operation and maintenance. Most wind farms have installed supervisory control and data acquisition(SCADA) systems for system control and logging data. However, the collected data are not used effectively. This paper proposes a fault detection method for main bearing wind turbine based on existing SCADA data using an artificial neural network(ANN). The ANN model for the normal behavior is established, and the difference between theoretical and actual values of the parameters is then calculated. Thus the early stage of main bearing fault can be identified to let the operator have sufficient time to make more informed decisions for maintenance.     

Protective action recommendations based upon plant conditions

Analyses conducted by the Nuclear Regulatory Commission (NRC) indicate that timely and effective protective action would be necessary to protect the public in a major nuclear power plant accident. Given the large amount of time required to implement an evacuation around most reactor sites, protective action recommendations (PARs) must be based upon specific plant indicators regarding the status of the core and systems that protect the core. This article describes the assumptions made, and the analyses conducted, by the NRC in developing its procedures for PARs based upon plant conditions.     

Dose evaluation in criticality accidents using response of panasonic TL personal dosemeters (UD-809/UD-802)

This study gives the results of dosimetry measurements carried out in the Silène reactor at Valduc (France) with neutron and photon personal thermoluminescence dosemeters (TLDs) in mixed neutron and gamma radiation fields, in the frame of the international accident dosimetry intercomparison programme in 2002. The intercomparison consisted of a series of three irradiation scenarios. The scenarios took place at the Valduc site (France) by using the Silène experimental reactor. For neutron and photon dosimetry, Panasonic model UD-809 and UD-802 personal TLDs were used together.     

Incorporating Artificial Neural Networks in the dynamic thermal–hydraulic model of a controlled cryogenic circuit

A model based on Artificial Neural Networks (ANNs) is developed for the heated line portion of a cryogenic circuit, where supercritical helium (SHe) flows and that also includes a cold circulator, valves, pipes/cryolines and heat exchangers between the main loop and a saturated liquid helium (LHe) bath. The heated line mimics the heat load coming from the superconducting magnets to their cryogenic cooling circuits during the operation of a tokamak fusion reactor. An ANN is trained, using the output from simulations of the circuit performed with the 4C thermal–hydraulic (TH) code, to reproduce the dynamic behavior of the heated line, including for the first time also scenarios where different types of controls act on the circuit. The ANN is then implemented in the 4C circuit model as a new component, which substitutes the original 4C heated line model. For different operational scenarios and control strategies, a good agreement is shown between the simplified ANN model results and the original 4C results, as well as with experimental data from the HELIOS facility confirming the suitability of this new approach which, extended to an entire magnet systems, can lead to real-time control of the cooling loops and fast assessment of control strategies for heat load smoothing to the cryoplant.     

Accident sequences sensitive to human errors

A Technical Committee Meeting was held by the International Atomic Energy Agency. The meeting was intended to review current operating experience to identify accident sequences involving human errors: to survey, based on PSA results, dominant core damage accident sequences where human errors have a relevant contribution; and to collect information on accident sequences in operator training, including experience with simulator training.It was agreed that more comprehensive reporting schemes have to be considered to collect human error information. PSA insights combined with operating experience have been considered a valuable source of information for identifying some of the human interactions and for appraising their plant impact. Finally, it was stated that there are still major difficulties to use simulators to train operators for severe accidents.     

事故工况下医院中子照射器I型堆瞬态特性研究

采用RELAPS/SCDAP/MOD3.4程序对医院中子照射器Ⅰ型堆(IHNI-1)在事故工况下的瞬态特性进行研究,对意外大反应性引入和池水丧失事故工况进行了计算和分析,计算结果表明:IHNI-1堆具有良好的固有安全性,在发生大反应性引入和池水丧失事故时,最终能够稳定在较低功率,确保反应堆安全.