[en] Numerous fire PSAs (probabilistic safety assessments) have shown that fire can be a major contributor to nuclear power plant risk. However, there are considerable uncertainties in the results of these assessments, due to significant gaps in current abilities to perform realistic assessments. These gaps involve multiple aspects of fire PSA, including the estimation of the probability of important fire scenarios, the modeling of fire growth and suppression, the prediction of fire-induced damage to equipment (including the effects of smoke), and the treatment of plant and operator responses to the fire. In response to recommendations of /VIR 93/, CSNI/PWG5 established a Task Group to review the present status and maturity of current methods used in fire risk assessments for operating nuclear power plants. The Task Group issued a questionnaire in May 1997 to all nuclear power generating OECD countries. The prime focus of the questionnaire (see Appendix A) was on a number of important issues in fire PSA: Fire PSA methodology and applications; Fire simulation codes; Ignition and damageability data; Modeling of fire spread on cables or other equipment; Modeling of smoke production and spread; Impact of smoke and heat on instrumentation, electronics, or other electrical equipment; Impact of actual cable fires on safety systems. The questionnaire requested specific information on these topics (e.g., computer codes used in fire PSAs, the physical parameters used to model ignition). Responses to the questionnaire were provided by Finland, France, Germany, Hungary, Japan, Spain, Switzerland, United Kingdom, and the USA. This report summarizes the questionnaire responses and thereby: a) provides a perspective on the current fire PSA state of the art (SOAR) with respect to the issues listed above, and b) provides numerous references for more detailed information regarding these issues. The main responsibility for writing different chapters of this report was divided between some OECD countries. The contents of each chapter are based on the writers' knowledge on his or her national practices and on the results of the questionnaire. The emphasis in the descriptions of the national practices also reflects the information supplied by the responding countries. Fire PSA is also used in other OECD countries, but the scope of this report is limited to those countries which responded to the questionnaire. The contents of this report are as follows: Fire PSA methodology overview - Based on a review of fire risk studies performed in the contributing countries, the report addresses different methodology and applications issues. Methodology issues, treated in Chapter 2, include the treatment of physical barriers, fire detection and suppression systems and fire fighting. They also include the treatment of operator actions and dependencies (both direct and indirect) between a fire and the plant's safety systems, definition of initiating events, and screening methods. Key assumptions and the effect of plant operational state (i.e., full power vs. low power operation) are dealt with in the report as well. Fire simulation models and codes applied or available - Chapter 3 of the report identifies which fire simulation codes have been used in actual PSAs. The models and scenarios used in different codes are described. To build confidence on fire simulation models, validation against experimental results in different types of fires is necessary. Fire experiments and the pre- and post experiment calculation used for code validation as well as ongoing fire simulation code development projects are discussed. Examples of fire scenarios and typical modeling assumptions are treated and numerous references are given in Chapter 3. References for experimental case studies and related simulation models and codes used for analyzing the production and spreading of smoke are also provided. The impact of smoke and heat - The immediate consequences of fires are caused by heat, smoke, and soot. The failure thresholds, modes, and attendant disturbances are treated in Chapter 4. The chapter provides references for statistical and phenomenological information relevant to ignition and component damage, and also for fire events. References are given modeling of fire spreading, the ignitability, damageability, and heat release rates of components such as cables, printed circuit boards, and electronics cabinets. Some experiments and simulations on the development of cabinet as well as cable tray fires are also described. Specific examples are given on the impact of smoke and heat on instrumentation electronics, electrical equipment. References are given on relevant experiments studying the impact of smoke and heat on electronics and/or electrical equipment. Some cases where cable fires have resulted in threats to plant safety or caused power and I and C circuit failures are discussed. References concerning the habitability of main control room are also provided. Applications Issues - The essence of fire PSA is discussed in Chapter 5. The scope and completeness of fire PSA may vary according to the strategy and stage of the overall PSA studies. For the identification of the main vulnerabilities of a plant, rough models are sufficient, whereas in a full scope fire PSA for detailed evaluation and comparison of risks more sophisticated methods have to be applied. Purpose and contents of the fire PSAs, as well as the uses made of these studies in different OECD countries are described in Chapter 5. It should be noted that the report is intended to serve as a resource to fire risk analysts but is not intended to serve as a fire PSA procedures guide. References for fire PSA guidance are provided in Chapter 2