Formal–statistical methods - Probabilistic Safety Assessment for Optimum Nuclear Power Plant Life Management (PLiM) - page 51

with ageing, may significantly reduce plant safety, as they may damage one
or more levels of protection. Ageing can lead to large-scale degradation
of the physical safety barriers and their components and increase the
probability of common cause failure. It may also lead to a reduction in
safety margins provided by the design and statutory requirements and,
ultimately, reduce safety.
Mechanical elements also pose a safety threat associated with
ageing because they can be damaged by unnoticed, almost catastrophic
failure. Mechanical items do not have enough failure statistics for
analysis. Acceptable data for statistical analysis could be obtained from the
monitoring data. Since the collection of data for mechanical components is
very costly and uninformative, they are sometimes studied by simulation.
Taking into account the fact that each nuclear power plant is composed
of thousands of items, reliability and ageing cannot be assessed for each
of them, so that the elements for analysis and management should be very
carefully selected. This also applies to probabilistic safety analysis (PSA),
which deals with the components important for safety. In this case, the
choice must be made so as to take into account the effects of ageing to
assess the level of risk. This can ensure greater efficiency of analysis and
reduce the costs associated with safety.
There are three qualitatively different approaches to data collection and
analysis:
(1) frequent periodic data collection over a long period of time. The
purpose of the analysis is to identify the existence and magnitude of the
trend.
(2) A single data collection under current conditions. The obtained
measurement data are used as input data for reliability and ageing
analysis. The purpose of the analysis is to estimate the probability of
failure in the future.
(3) data collection, such as the results of non-destructive testing of
materials. Analysis may have two objectives: to assess the current state
of the material or, as in (1), the detection and quantitative assessment of
current trends.
Approach (1) is not new and includes developed methods of analysis.
Approach (2) requires the development of mechanical models. Approach (3)
requires special consideration since the gradual development of defects such
as cracks can be detected by periodic in-service inspection (ISI). ISI data
are extremely complex and difficult to interpret. A probabilistic approach
to evaluate a random error must be used. Probabilistic models and methods
of solutions considered are Refs. 16, 20-27.
It is evident that the data that should be collected and stored are
determined by the tasks of controlling the reliability, ageing and operating
life of nuclear power plant 19 :
• Prediction of residual life;
• Preventative maintenance of equipment and pipelines;

￿ ￿ ￿ ￿ ￿