Fire safety at nuclear installations is a key topic shared by many actors in the nuclear industry such as nuclear licence operators and nuclear Technical Safety Organisations (TSO) with the support of academic research institutes. Fire safety assessments and operational experience gained from fire events have shown that fire contributes significantly to the overall fuel damage frequency for both existing and new-build plant designs. Consequently, there is a constant need to understand the risks inherent to fire for nuclear facilities and the capability of new tools to address these risks.
With fire occurrence frequencies in the order of 10−2/reactor year, fire at nuclear power plants remains a significant contributor to the overall risk with respect to the operation of these facilities. In fact, fire safety analysis studies supplemented by international operating experience demonstrate that fire contributes significantly to the overall Core Damage Frequency (CDF) for both existing and new plants. The actual contribution of fire to the overall CDF (typically in the same order as that of internal events) requires further analysis given the high levels of uncertainty resulting from the lack of knowledge of the different fire-specific phenomena with the required degree of precision. Furthermore, fire modelling is generally used in a number of fire safety assessment studies and to support regulatory decisionmaking. Nevertheless, challenges still exist when using validated fire models that can reasonably and reliably predict the consequences of a fire at a nuclear installation. Validation studies for fire models for nuclear installations have shown areas where the fire models must be improved, but they have also highlighted areas where engineers can be confident when using the current generation of fire models. Furthermore, these studies have emphasised the need for more large-scale experiments specifically developed for the purpose of validating fire models. Following the work carried out in support of OECD/NEA/CSNI Report, the international nuclear safety community realised that it was important to further experimental fire studies as well as to improve the predictive capabilities of fire models and computer codes.
The design of nuclear plants makes the study of fire development and propagation complex since there is a strong correlation between fire and ventilation conditions. One of the main technical issues in fire safety assessment relates to smoke and hot gas propagation at nuclear plants. This investigation takes into account the disturbance of the ventilation network, negative room pressure due to the fire itself and the behaviour of active fire barrier elements (such as fire dampers, fire-break doors, etc.).
The PRISME project (PRISME is the French acronym for “Fire Propagation in Elementary Multi-Room Scenarios”) was initiated by IRSN under the auspices of OECD/AEN. The proposal was to use their specially-designed facilities at the GALAXIE site in Cadarache (France) to carry out various fire test scenarios of interest for nuclear fire safety. The three major research areas addressed by the PRISME project included propagation of heat and smoke from the fire room to adjacent rooms, impact of heat and smoke on safety–critical systems (such as cable malfunction) and the impact of the ventilation network on limiting heat and smoke propagation. In all, five experimental campaigns consisting of more than 35 large-scale fire tests were carried out using the DIVA facility, part of the GALAXIE site. In parallel to these tests, PRISME partners evaluated the capability of various fire computer codes to simulate fire scenarios based on the PRISME data results. A number of benchmark exercises were conducted within the analytical working group of PRISME with the aim to further advance the predictive capabilities of the computer codes.
The project was formally launched in January 2006 and was concluded in June 2011. A total of 12 countries and more than 20 organisations have ultimately participated in the PRISME Project: Belgium (GDF Suez, Bel V, University of Gent); Canada (AECL); Finland (STUK, VTT); France (IRSN as operating agent, EDF, DGA); Germany (GRS, BFS, iBMB); Japan (JNES); the Republic of Korea (KINS, KOPEC, KAERI, KEPRI); the Netherlands (VROM-KFD, NRG); Spain (CSN); Sweden (VATTENFALL, University of Lund); the United Kingdom (HSE) and the United States (US-NRC).
We decided that it would be appropriate for Fire Safety Journal to publish a special issue dedicated to the PRISME project. The special issue compiles contributions from many of the PRISME participants and focuses on the main outcomes of the project subsequent to five years of experimental research and modelling.
 M. Röwekamp, Al, State of the art report on fire risk analysis, fire simulation, fire spreading and impact of smoke and heat on instrumentation electronics—state-of-the-art report, Nuclear Energy Agency (OECD), NEA/CSNI/R(99)27, February 2000.
 W. Werner, Al, The OECD Fire Database—Conclusions from Phase 2 and Outlook, in: Proceedings of the 20th International Conference on Structural Mechanics in Reactor Technology (SMIRT 20), 11th International Post-Conference Seminar on Fire Safety at Nuclear Power Plants and Installations, 2009.