• Energy Research

Abstract This paper presents two approaches based on different artificial intelligence techniques (AI), genetic algorithms (GA) and neural networks (NN), for pattern recognition on the set of transient reactor signals. The first approach describes a possibilistic analysis based on the minimum centroids set method optimized by genetic algorithm. The use of a possibilistic classification provides a set of natural and consistent classification rules leading naturally to a good heuristic, based on the cluster size, to handle unlabeled transients by providing a ‘don't know’ response. In situations where the safety is the decisive parameter, this characteristic is fairly desirable in transient classification systems, since wrong or unreliable classifications can be catastrophic. Application of the proposed approach to a nuclear transient identification problem shows good capability of the genetic algorithm in learning optimized possibilistic classification rules for efficient diagnosis including ‘don't know’ response. The second approach uses two multilayer neural networks trained with the backpropagation algorithm. The first NN does a preliminary transient type identification using as inputs a short set (in a moving time window) of recent measurements of each variable. These inputs were presented to the NN, for each variable, as a set of consecutive values in temporal sequence. The second NN is used to validate the preliminary identification (from the first NN) through a self associative neural simulator used to estimate the value of every state variable. In order to validate both methods, a nuclear power plant (NPP) transient identification problem, comprising postulated accidents, is proposed. Noisy data was used in order to evaluate the method robustness. The results obtained reveal the ability of the methods in dealing with both dynamic identification of transients and correct ‘don't know’ response.

Nuclear power plant systems are comprised of both on-line and standby components. Standby components differ from on-line ones, as they might be unavailable due to unrevealed failures. The usual procedure employed to reveal failures before real demands is to submit the component to surveillance tests. Surveillance test policies might deal with two conflicting scenarios: the test frequency must be sufficiently high in order to reveal failures before demands, but, on the other hand, it must be low enough due to its influence on the component unavailability. Standard surveillance test policies for typical nuclear power plants usually consist of periodic tests for which the frequencies are often higher than necessary for obtaining the optimal availability. In this work, a new surveillance test policy optimization method, based on genetic algorithms, is applied to the Angra-I (Brazilian PWR) auxiliary feedwater system. The new probabilistic model has been developed in order to comprise the following features: (1) aging effects on standby components when they undergo surveillance tests; (2) revealing failures during the surveillance tests implies corrective maintenance, and, consequently, increasing outage times; (3) components are distinct (i.e., each has distinct test parameters, such as outage time, aging factors, etc); (4) tests are not necessarily periodic. The results, when compared to those obtained by standard test policies, show improved overall availability at the system level. © 2002 Wiley Periodicals, Inc.

Abstract Small Modular Reactors, SMRs, are variants of compact modular reactors of generation III+ that present attractive features such as simplicity, enhanced safety and are economically competitive. In an iPWR (integral pressurized water reactor), the pressurizer is located at the top of the reactor vessel; this configuration requires the investigation of the mechanisms adopted for the homogenization of boron. The present work shows the contributions to the design of an experimental installation proposed to provide relevant data to study the boron homogenization phenomena in a pressurizer of an SMR. For this, an experimental installation was built with a test section made of stainless steel with a volumetric scale equal to 1:200, representing one-fourth of the pressurizer of an iPWR. The experimental installation has a test section, a heat exchanger, two pumps, five flow meters and three stainless steel tanks with capacities of 0.143, 0.069 and 0.216 m 3 . The test facility layout was mounted so that the test section was fed with water or saline solutions with different concentrations, representing scenarios of boration or deboration. To determine the concentration at the inlet and outlet of the test section, samples have been collected every 10 min during the experiments representing some scenarios (boration/deboration). The electrical conductivity of these samples was determined by a conductivity meter and converted to the concentration of the tracer element (NaCl) used. Three representative experiments were conducted and it was possible to prove the viability of using the experimental setup in a reliable form, for the range of values occurring in the boron homogenization phenomenon in a pressurizer of a compact modular reactor.

Abstract A new approach to the study of ballooning that causes cladding failure in fuel rods using an adaptive neural fuzzy inference system (ANFIS) is presented in this paper. By mapping input/output patterns describing cladding failure phenomena through average inner cladding temperature and fuel rod gas pressure, ANFIS shows a great potential to modeling this problem in alternative to the traditional approach. A typical pressurized water reactor fuel rod data was used to this application. The results confirm the potential of ANFIS comparatively to experimental calculations.

This article presents an uncertainty modeling study using a fuzzy approach applied to the Westinghouse advanced nuclear reactor. The AP1000 Westinghouse Nuclear Power Plant (NPP) is provided of passive safety systems, based on thermo physics phenomenon, that require no operating actions, soon after an incident has been detected. The use of advanced passive safety systems in NPP has limited operational experience. As it occurs in any reliability study, statistically non-significant events report introduces a significant uncertainty level about the failure rates and basic events probabilities used on the fault tree analysis (FTA). In order to model this uncertainty, a fuzzy approach was employed to reliability analysis of the AP1000 large break Loss of Coolant Accident (LOCA). The final results have revealed that the proposed approach may be successfully applied to modeling of uncertainties in safety studies.

This paper presents a preliminary study on the use of adaptive neural fuzzy inference system (ANFIS) to determine the fragility curves in degraded nuclear power plant (NPP) passive components. The goal of this approach is to allow the direct association, using a mapping of input/output patterns, between measurable beam deflections and the structure probability of failure for a severe degradation condition. The present study consists of an Artificial Intelligence framework application considering the information obtained from an original Nuclear Regulatory Commission (NRC) research program. The results indicate that the ANFIS modeling is a promising alternative to traditional approach in nuclear studies of structure degradation in passive components.

Abstract This paper presents a fuzzy inference system (FIS) as an approach to estimate Nuclear Power Plant (NPP) performance indicators. The performance indicators for this study are the energy availability factor (EAF) and the planned (PUF) and unplanned unavailability factor (UUF). These indicators are obtained from a non analytical combination among the same operational parameters. Such parameters are, for example, environment impacts, industrial safety, radiological protection, safety indicators, scram rate, thermal efficiency, and fuel reliability. This approach uses the concept of a pure fuzzy logic system where the fuzzy rule base consists of a collection of fuzzy IF–THEN rules. The fuzzy inference engine uses these fuzzy IF–THEN rules to determine a mapping from fuzzy sets in the input universe of discourse to fuzzy sets in the output universe of discourse based on fuzzy logic principles. The results demonstrated the potential of the fuzzy inference to generate a knowledge basis that correlate operations occurrences and NPP performance. The inference system became possible the development of the sensitivity studies, future operational condition previsions and may support the eventual corrections on operation of the plant.

Abstract The effective contribution of nuclear energy will depend on various factors related to economics, safety, public acceptance and sustainability. To assure, however, the nuclear energy development, reactor accident impacts, as Fukushima, must be evaluated in a predictive way. Environmental assessment models are used for evaluating the radiological impact of potential releases of radionuclides from nuclear reactors to the environment. It is important to evaluate, to the extent possible, the reliability of the predictions of such models, by comparing with measured values in the environment or by comparing with the predictions of other models. Tritium has a complex environmental behavior once released into the environment. It is essential to establish reference scenarios to allow the simulation of tritium aquatic pathway subsequent to accidental releases. For this purpose, two scenarios for seawater circulation were analyzed by hydrodynamic modeling. An inverse modeling procedure was successfully applied to estimate tide elevations on the borders, which are based on applying the harmonic constants and using the same overestimation percentage produced by model results to correct the border values. Simulations of validated model for postulated accidental releases of tritium inventory from heavy water reactors, whose doses could be relevant, were presented here. It was observed differences between the two scenarios for the transport modeling that were caused by the removal of large volume of polluted waters from the accident site and its dilution in the discharge area, which has minor tritium concentrations. Moreover, the processes involved in the dynamic transfer of tritium in the environment were analyzed in dependence on the environmental conditions of tropical coastal ecosystem.

Abstract This paper presents the concept of “Design by Genetic Algorithms (D by GA)”, applied to a new reduced scale system problem. The design problem of a passive thermal-hydraulic safety system, considering dimensional and operational constraints, has been solved. Taking into account the passive safety characteristics of the last nuclear reactor generation, a PWR core under natural circulation is used in order to demonstrate the methodology applicability. The results revealed that some solutions (reduced scale system D by GA) are capable of reproducing, both accurately and simultaneously, much of the physical phenomena that occur in real scale and operating conditions. However, some aspects, revealed by studies of cases, pointed important possibilities to D by GA methodological performance improvement.