[en] A probabilistic risk assessment and consequence analysis for a BWR with a Mark-II containment design has shown that overpressurization is the dominant containment failure mode for a wide range of potential core meltdown accidents. This failure mode is a major contributor to the predicted off-site health consequences. Mitigation of this failure mode is aimed at maintaining the containment integrity by using a wetwell venting system and a drywell spray system. Detailed analyses with the MARCH 2 code were used to determine the effectiveness of the mitigating features. The results show that the wetwell venting system is capable of preventing overpressurization in the containment and the drywell spray system is needed to reduce the corium/concrete interactions. However, additional mitigating systems are required for long-term cooling of the corium and ultimate removal of the decay heat for the ATWS case

[en] The fault tree technique has become a standard tool for the analysis of safety and reliability of complex system. In spite of the costs, which may be high for a complete and detailed analysis of a complex plant, the fault tree technique is popular and its benefits are fully recognized. Due to this applications of these codes have mostly been restricted to simple academic examples and rarely concern complex, real world systems. In this paper an interactive approach to fault tree construction is presented. The aim is not to replace the analyst, but to offer him an intelligent tool which can assist him in modeling complex systems. Using the CAFTS-method, the analyst interactively constructs a fault tree in two phases: (1) In a first phase he generates an overall failure logic structure of the system; the macrofault tree. In this phase, CAFTS features an expert system approach to assist the analyst. It makes use of a knowledge base containing generic rules on the behavior of subsystems and components; (2) In a second phase the macrofault tree is further refined and transformed in a fully detailed and quantified fault tree. In this phase a library of plant-specific component failure models is used

[en] Financial risk assessment, where the probability and severity of financial consequences are estimated, offers a logical framework for organizing and evaluating data pertinent to nuclear power plant accidents. Under the sponsorship of the Electric Power Research Institute, General Electric investigated the feasibility of financial risk assessment of nuclear power plants and of applying PRA methods and data in such an assessment. This paper summarizes the main findings of this investigation. Specifically, the paper discussed the following topics: definition of financial consequences and financial risk; overall approach for financial risk assessment and how it compares with the approach for PRA used in the Reactor Safety Study; and specific financial risk assessment procedures for defining initiating events, plant response sequences, institutional scenarios, and financial consequences and how they compare to analogous procedures for PRA

[en] In France a probabilistic risk assessment (PRA) is presently in progress in the Safety Assessment Department of CEA for a standardized 900 MW PWR plant. Indeed, the French safety approach for severe accidents management is the implementation of a set of special emergency procedures which identify the optimal actions even for out of design situations. So by accounting for these procedures (correctly or incorrectly applied) an important set of operator actions can be included in the PRA. This approach implies the introduction of the notion of procedure failure at the level of the Event Trees. Quantification of the corresponding probabilities leads to several problems as physical efficiency of the procedures, systems availability (including instrumentation), and human behavior. Among these problems the physics of the sequence is a primary question, both to state the efficiency of the procedure to prevent a core damage, and to assess the allowable time for the operator actions. So a set of thermohydraulic studies was initiated for various accident sequences and procedures. The authors present the example of the small LOCA Event Trees and the studies related to the introduction of procedure actions in case of HPSI failure. The results illustrate the interest of the approach and its significant impact on the PRA

[en] The author discusses the urgent need for better systems and procedures for evaluating actual or suspected construction deficiencies in nuclear power plants. The following topics of interest are discussed: summary of tools available, use of plant-specific probabilistic risk assessments, general process for the rational management of construction deficiencies, rationales for the timing of required corrective actions, example of deficiency management in France, proposed screening process, deficiencies calling for corrective actions, institutional obstacles, and specific recommendations

[en] A joint Nordic project was initiated in 1981 by the Nordic Liaison Committee for Atomic Energy. The project is one of several joint projects: it was running over a four year period, and was funded by the Nordic Council of Ministers. The project is entitled Large Reactor Accidents - Consequences and Mitigating Actions, and various subprojects are described herein. Data-related subprojects include: (1) terrestrial transfer functions; (2) fresh-water pathways; (3) comparison of dynamic and static calculation models for fish; and (4) shielding effect of buildings. Experimental subprojects include (1) natural decontamination of roofs; (2) winter conditions; (3) deposition in urban areas; and (4) the filter effect of buildings. Three subprojects related to accident consequence assessment models are described, and mitigating actions related to health consequences of reactor accidents are discussed

[en] A methodology for the analysis of human actions under accident conditions has been developed, which uses information from plant simulator runs, plant procedures, and plant systems information. The objective is to enhance the completeness of the event sequence model (event trees) with respect to both favorable and unfavorable operator actions. Routine human actions that impact the plant at or below the systems level, such as test and maintenance actions, are handled in the systems analysis. Types of dynamic operator actions analyzed in this paper are actions taken during an event sequence that: supplement the automatic response of plant systems for event mitigation, change or detract from the automatic response of plant systems, or lead to recovery of failed systems. The derived results can be used directly in a probabilistic risk assessment. It is judged that the major cause of possible error is misdiagnosis, which can lead to either errors of omission or errors of commission. Operator mistakes may occur when a situation is misclassified or when inappropriate decisions and response selections are made in the operator action sequences. The operator action sequences are modeled in a natural progression of human response, including observation of plant parameters and the diagnosis of the event or the decision to take action

[en] The Electric Power Research Institute sponsored the review and comparison of five PRA studies: Arkansas Nuclear One - Unit 1, Big Rock Point, Grand Gulf, Limerick, and Zion - Unit 1. The review has been conducted in two phases. The Phase I review may be characterized as a qualitative look into many aspects of a PRA study. The Phase II review was performed to quantify the extent that differences in analytical techniques or key assumptions in these areas affect the differences in study results. In each of the PRA studies reviewed, the general descriptions of analytical approaches and descriptions of the analyses of event tree, fault tree and human interaction analyses that affected the dominant core damage sequences were reviewed. When these descriptions aroused interest because of seeming inconsistencies within the study or with other studies, they were pursued in some depth. The approaches or assumptions were contrasted to similar elements from other studies, and sensitivity analyses were performed in many cases to test the significance of results to the analytical models or assumptions. Inferences were drawn from the results regarding significance of the item to plant-specific results and, where possible, were generalized to other PRAs. This paper describes the results of the review of system dependencies and human interactions

[en] In recent years there has been a trend in Swedish Probabilistic Safety Analysis (PSA) work towards coordination of the tools and methods used, in order to facilitate exchange of information and review. Thus, standardized methods for fault tree drawing and basic event coding have been developed as well as a number of computer codes for fault tree handling. The computer code used by Asea-Atom is called SUPER-TREE. As indicated by the name, the key feature is the concept of one super tree containing all the information necessary in the fault tree analysis, i.e. system fault trees, sequence fault trees and component data base. The code has proved to allow great flexibility in the choice of level of detail in the analysis

[en] The support state model employed in the Progetto Unificato Nucleare (PUN) Probabilistic Safety Study was an effective means for the analysis of intersystem dependencies for modeling and quantifying accident sequences. The model identifies the various plant system interactions and defines the complete spectrum of support system conditions, in such a way as to significantly reduce intersystem dependencies in frontline event trees. The support state model allows for eminent insights into the understanding of plant system dependencies and provides an immediate perception of support system failures that are critical for the safety of the plant. It is a very effective model for an accurate quantification of accident sequence frequencies