[en] The international nuclear community has expressed concern that some changes in existing plants could challenge safety margins while fulfilling all the regulatory requirements. In 1998, NEA published a report by the Committee on Nuclear Regulatory Activities on Future Nuclear Regulatory Challenges. The report recognized 'Safety margins during more exacting operating modes' as a technical issue with potential regulatory impact. Examples of plant changes that can cause such exacting operating modes include power up-rates, life extension or increased fuel burnup. In addition, the community recognized that the cumulative effects of simultaneous changes in a plant could be larger than the accumulation of the individual effects of each change. In response to these concerns, CSNI constituted the safety margins action plan (SMAP) task group with the following objectives: 'To agree on a framework for integrated assessments of the changes to the overall safety of the plant as a result of simultaneous changes in plant operation / condition; To develop a CSNI document which can be used by member countries to assess the effect of plant change on the overall safety of the plant; To share information and experience.' The two approaches to safety analysis, deterministic and probabilistic, use different methods and have been developed mostly independently of each other. This makes it difficult to assure consistency between them. As the trend to use information on risk (where the term risk means results of the PSA/PRA analysis) to support regulatory decisions is growing in many countries, it is necessary to develop a method of evaluating safety margin sufficiency that is applicable to both approaches and, whenever possible, integrated in a consistent way. Chapter 2 elaborates on the traditional view of safety margins and the means by which they are currently treated in deterministic analyses. This chapter also discusses the technical basis for safety limits as they are used today. Chapter 3 looks at techniques for the deterministic calculation of safety margins and discusses the complementary probabilistic risk assessment techniques needed to generalize safety margins beyond design basis accidents. Chapter 4 examines the definition of safety margin, which is noted to take different meanings in different fields. For example, in civil engineering and applications that deal with the load-strength interference concept, safety margin describes the distance between the means of the load and strength probability density functions with regard to the standard deviation in both. However, in the nuclear industry, the term safety margin evolved to describe the goal of assuring the existence of adequate safety margin in deterministic calculations. Specifically, safety margin refers to keeping the value of a given safety variable under a pre-established safety limit in design basis accidents. Implicitly, safety margin in the nuclear industry is the distance from the safety limit to onset of damage. The SMAP task group fulfilled its first objective by adopting a methodology for quantifying safety margins that merges the deterministic and probabilistic approaches. The methodology described in Chapter 5 is consistent with the definition of safety margin commonly used in the nuclear industry. The metrics of this methodology quantify the change in safety over a range of accident sequences that extend beyond the design bases. However, the methodology is not described in this report to a level that would meet guidance document requirements. This is in part because methods and techniques needed to quantify safety margins in a global manner are evolving, and thus specific guidance rendered at this time would shortly become obsolete. This report presents the framework in sufficient detail to serve as the basis of an analysis and, thus, this report meets the second objective established for the SMAP group. A proof-of-concept application to further aid potential applicants of the methodology is included in Chapter 6. As recognized from the beginning, the proposed action plan has been highly multidisciplinary and a wide variety of expertise has been needed for its development. The exchange of information and experience among participating organizations, which was defined as the third objective in the safety margins action plan, has not been only an essential working method, but also a net benefit of the work