Risk-informed decisions optimization in inspection and maintenance

摘要

The Risk-Informed Approach (RIA) used to support decisions related to inspection and maintenance program is considered. The use of risk-informed methods can help focus the adequate in-service inspections and control on the more important locations of complex dynamic systems. The focus is set on the highest risk measured as conditional core damage frequency, which is produced by the frequencies of degradation and final failure at different locations combined with the conditional failure consequence probability. The probabilities of different degradation states per year and consequences are estimated quantitatively. The investigation of inspection and maintenance process is presented as the combination of deterministic and probabilistic analysis based on general risk-informed model, which includes the inspection and maintenance program features. Such RIA allows an optimization of inspection program while maintaining probabilistic and fundamental deterministic safety requirements. The failure statistics analysis is used as well as the evaluation of reliability of inspections. The assumptions regarding the effectiveness of the inspection methods are based on a classification of the accessibility of the welds during the inspection and on the different techniques used for inspection. The probability of defect detection is assumed to depend on the parameters either through logarithmic or logit transformation. As example the modeling of the pipe systems inspection process is analyzed. The means to reduce a number of inspection sites and the cumulative radiation exposure to the NPP inspection personnel with a reduction of overall risk is presented together with used and developed software. The developed software can perform and administrate all the risk evaluations and ensure the possibilities to compare different options and perform sensitivity analysis. The approaches to define an acceptable level of risk are discussed. These approaches with appropriate software in partial case is used for the construction and research of the models related to inspections and maintenance planning of Ignalina Nuclear Power Plant (RBMK-1500) piping components. The discussed example is related to risk analysis and inspection program improvements for selected pipe systems. The new risk-informed inspection and maintenance program for selected pipe systems are compared with various alternatives. The usage of risk evaluations to optimize the selection of inspection locations, the inspection interval, and the changes in risk and cost due suggested modifications are demonstrated. The proposed integrated modeling methodic and general model of inspection process can be used as a base for other risk-informed models of inspection process control and risk monitors of complex dynamic systems.