Preventing failures by in-service inspection

The concept of defence in depth is fundamental to the safety of nuclear installations. Defence in depth consists in a hierarchical deployment of different levels of equipment and procedures in order to maintain the effectiveness of physical barriers placed between radioactive material and workers, the public or environment in normal operation, anticipated operational occurrences and in accidents at the plant. The primary way of preventing accidents is to achieve a high quality in design, construction and operation of the plant, and thereby to ensure that deviations from normal operation are infrequent. According to international guidance on defence in depth design relies on deterministic assumptions and procedures without explicit consideration of probabilities. Conservatism, including safety margins, is a basic prerequisite, which apply to the first three level of the defence. In-service inspection and testing are important means on the second level to prevent incidents and accidents by detecting any degradation of components and equipment before it can affect the safety of the plant. Moreover, in-service inspection and testing of structures, systems and components important to safety need to be of such a standard and frequency as to ensure that levels of reliability and effectiveness remain in accordance with the design assumptions and intent, and that the safety of the plant has not been compromised since beginning of operation. The basic purpose of an in-service inspection programme is to decrease the plant risk by decreasing the probability of failure, inspection having obviously no impact on the consequences of a failure. Development of in-service inspection strategies for passive components in nuclear power plants has always involved implied, but un-quantified, risk assessments. In the early inservice inspection strategies components were consequence ranked by a safety or quality grade classification system. In these strategies influential factors for degradation considered, were more or less focussed on design stresses. During the late 1970's and early 1980's component failures due to different types of cracking and other types of degradation were observed in many nuclear power plants in several countries. The cracking and degradation problems were often discovered by leakage, and not by in-service inspections. These observations led to concerns about the efficiency of the inspection programmes that were based on ASME Section XI or similar codes. Since the introduction of augmented inspections, and other programmes developed for specific degradation mechanisms, together with more stringent requirements for demonstration of NDT performance relatively few cases of serious degradation not detected by inspection have been reported. Substantial knowledge has been accumulated over the years about the degradation mechanisms that can cause component failures and how this type of degradation can occur. Developments and sophistication levels of PRA have increased the risk insights. The experience and knowledge gained have successively been used to further improve the efficiency of the in-service inspection programs. A number of risk-informed approaches are now evolving in the nuclear industry in different countries. Opportunities for further improvements and optimisations clearly exist. Such improvements, based either on qualitative or quantitative risk approaches, have however to consider the fact that in-service inspection and testing are important means in the plants defence in depth strategy. Any risk-informed approach should therefore, as far as possible, consider the fact that unexpected degradation also will occur in the future. Any risk informed in-service inspection approach also has to be balanced with other optimisation efforts in the plants and that are based on risk considerations. (author)