Multi-objective optimization of maintenance programs in nuclear power plants using Genetic Algorithm and Sensitivity Index decision making

2016 ◽  
Vol 88 ◽  
pp. 95-99 ◽  
Author(s):  
Navid Ayoobian ◽  
Massoud Mohsendokht
2014 ◽  
Vol 494-495 ◽  
pp. 1715-1718
Author(s):  
Gui Li Yuan ◽  
Tong Yu ◽  
Juan Du

The classic multi-objective optimization method of sub goals multiplication and division theory is applied to solve optimal load distribution problem in thermal power plants. A multi-objective optimization model is built which comprehensively reflects the economy, environmental protection and speediness. The proposed model effectively avoids the target normalization and weights determination existing in the process of changing the multi-objective optimization problem into a single objective optimization problem. Since genetic algorithm (GA) has the drawback of falling into local optimum, adaptive immune vaccines algorithm (AIVA) is applied to optimize the constructed model and the results are compared with that optimized by genetic algorithm. Simulation shows this method can complete multi-objective optimal load distribution quickly and efficiently.


Author(s):  
G. SRINIVAS ◽  
A. K. VERMA ◽  
A. SRIVIDYA ◽  
SANJAY KUMAR KHATTRI

Technical Specifications define the limiting conditions of operation, maintenance and surveillance test requirements for the various Nuclear Power plant systems in order to meet the safety requirements to fulfill regulatory criteria. These specifications impact even the economics of the plant. The regulatory approach addresses only the safety criteria, while the plant operators would like to balance the cost criteria too. The attempt to optimize both the conflicting requirements presents a case to use Multi-objective optimization. Evolutionary algorithms (EAs) mimic natural evolutionary principles to constitute search and optimization procedures. Genetic algorithms are a particular class of EA's that use techniques inspired by evolutionary biology such as inheritance, mutation, natural selection and recombination (or cross-over). In this paper we have used the plant insights obtained through a detailed Probabilistic Safety Assessment with the Genetic Algorithm approach for Multi-objective optimization of Surveillance test intervals. The optimization of Technical Specifications of three front line systems is performed using the Genetic Algorithm Approach. The selection of these systems is based on their importance to the mitigation of possible accident sequences which are significant to potential core damage of the nuclear power plant.


Author(s):  
Arnold Gad-Briggs ◽  
Pericles Pilidis ◽  
Theoklis Nikolaidis

A framework – NuTERA (Nuclear Techno-Economic and Risk Assessment) has been developed to set out the requirements for evaluating Generation IV (Gen IV) Nuclear Power Plants (NPPs) at the design conceptual stage. The purpose of the framework is to provide guidelines for future tools that are required to support the decision-making process on the choice of Gen IV concepts and cycle configurations. In this paper, the underpinning of the framework has been demonstrated to enable the creation of an analyses tool, which evaluates the design of an NPP that utilises helium closed Brayton gas turbine cycles. The tool at the broad spectrum focuses on the component and cycle design, Design Point (DP) and Off-Design Point (ODP) performance, part power and load following operations. Specifically, the design model has been created to provide functionalities that look at the in-depth sensitivities of the design factors and operation that affect the efficiency of an NPP such as temperature and pressure ratios, inlet cycle temperatures, component efficiencies, pressure losses. The ODP performance capabilities include newly derived component maps for the reactor, intercooler and recuperator for long term Off-Design (OD) operation. With regard to short term OD, which is typically driven by changes in ambient conditions, the ability to analyse the cycle load following capabilities are possible. An economic model has also been created, which calculates the component costs and the baseline economic evaluation. An incorporated risk model quantifies the performance, operational, financial and design impact risks. However, the tool is able to optimise the NPP cycle configuration based on the best economics using the Levelised Unit Electricity Cost (LUEC) as a measure. The tool has been used to demonstrate a typical decision-making process on 2 Gen IV helium closed gas turbine cycles, which apply to the Gas-cooled Fast Reactors (GFRs) and Very-High Temperature Reactors (VHTRs). The cycles are the Simple Cycle Recuperator (SCR) and Intercooled Cycle Recuperator (ICR). The tool was able to derive the most efficient cycle configurations for the ICR (53% cycle efficiency) and SCR (50% cycle efficiency). Based on these efficiency figures, the baseline LUEC ($/MWh) for the year 2020 is $62.13 for the ICR and $61.84 for the SCR. However, the inclusion of the cost of contingencies due to risks and the subsequent economic optimisation resulted in a cost of $69.70 and $69.80 for the ICR and SCR respectively.


2006 ◽  
Vol 27 (7) ◽  
pp. 1037-1057 ◽  
Author(s):  
John S. Carroll ◽  
Sachi Hatakenaka ◽  
Jenny W. Rudolph

We explore the linkages between naturalistic decision making, which examines decisions in context, and team and organizational learning, which examines how feedback from decisions affects context. We study 27 problem investigation teams in three nuclear power plants, a setting that combines complex team decisions with organizational learning. Further, managers who commission the teams and receive team reports are a key aspect of context for the teams and a critical conduit for organizational learning and change. Questionnaires were given to both team members and manager recipients of written team reports, and team reports were coded for qualities of their analyses and recommendations. We find that team members value reports in which the team discovered causes or lessons that could be used in other contexts, whereas managers appreciate reports with logical corrective actions from teams with investigation experience. Teams with managers or supervisors as team members are better able to reach shared understanding with their manager customers. Teams with more diverse departmental backgrounds produce deeper and more creative analyses. Teams need access to information and analytical skills in order to learn effectively, but they also need management support and boundary-spanning skills in order to diffuse their learning.


Sign in / Sign up

Export Citation Format

Share Document