Fire development in multi‐compartment facilities: PRISME 2 project

Fire hazards analyses and probabilistic fire safety analyses have demonstrated that fire can be an important contributor to core damage frequency and other major plant damage states of nuclear power plants.1 In this issue, fire modelling has been undertaken by a diverse set of contributors, including, nuclear licensees, technical safety organizations, and, in some countries, regulators, to assess the consequences of fires. One important aspect of fire related risk-informed and performance-based regulation is undeniably the availability of verified and validated fire models that can reliably estimate the consequences of a fire in confined and mechanically ventilated compartments. A number of members of the Organization for Economic Co-operation and Development (OECD) Nuclear Energy Agency (NEA) expressed their interest in participating in a joint international research project on the topic of fire events to be carried out under the auspices of the NEA. The PRISME (French acronym for “Fire Propagation in Elementary Multi-Room Scenarios”) Project was launched from 2006 to 2010 by the Institut de Radioprotection et de Sûreté Nucléaire (IRSN, France) with their specially designed facilities in Cadarache. The three major research areas addressed by the PRISME Project included the confinement effect on the fire dynamics, the smoke propagation from the fire compartment to adjacent rooms, and the effect of the ventilation network on limiting smoke propagation. In total, five experimental campaigns consisting of more than 35 large-scale fire tests were carried out. The main experimental results and findings were presented in the first OECD/NEA PRISME summary report.2 In parallel to these experimental campaigns, PRISME partners evaluated the capabilities of various fire modelling codes to simulate fire scenarios based on the PRISME results. A number of benchmark exercises were conducted within an analytical working group of PRISME, which further advanced the knowledge on the predictive capabilities of the various fire codes being used.3-10 Some of these studies were published in the Fire Safety Journal, PRISME special issue.11 The experimental findings of the PRISME project and the analytical working group highlighted that confined mechanically ventilated fires were composed of physical complex phenomena to model. The outputs of the PRISME experiments and analysis allowed for the identification of further focused experimentations to address areas of uncertainty. These targeted areas formed the basis of the second PRISME Project. The second phase was launched on July 2011 and ended on December 2016. A total of nine countries signed the agreement to become PRISME 2 members: Belgium (Bel V and Tractebel-ENGIE), Canada (Canadian Nuclear Safety Commission—CNSC), Finland (Technical Research Centre—VTT), France (IRSN as Operating Agent and Électricité de France—EDF), Germany (Gesellschaft für Anlagen-und Reaktorsicherheit—GRS), Japan (Nuclear Regulation Authority—NRA and Central Research Institute of Electric Power Industry—CRIEPI), Spain (Consejo de seguridad nuclear—CSN), Sweden (Strålsäkerhetsmyndigheten—SSM), and the United Kingdom (Office of Nuclear Regulation—ONR). The project focused on advancing the state-of-the-art knowledge in smoke and hot gas propagation through a horizontal opening between two superposed compartments, fire spread on real fire sources such as cable trays or electrical cabinets, and fire extinguishing studies using fixed water-based suppression systems. In total, four experimental campaigns were undertaken comprising more than twenty large-scale fire tests within the IRSN DIVA facility in Cadarache. The last experimental campaign was defined based on the results and analysis of the previous three to ensure a cost-effective experimental campaign. In addition to the large scale fire tests, extra support tests for the characterization of fire sources, in open atmosphere, were performed to provide additional data for validation purposes. As in the previous PRISME project, the analytical working group evaluated the capabilities of various fire modelling codes to simulate fire scenarios based on the PRISME 2 results.12-15 The output of the PRISME 2 project has been summarized in the OECD/NEA application report.15 The objective of the first PRISME 2 experimental campaign, named VSP for vertical smoke propagation, was to investigate the vertical smoke propagation through a horizontal opening for fire scenarios in mechanically and ventilated compartments. A detailed description of the flow at the vent was one key issue of this campaign in order to provide detailed data to validate fire simulation zone models and more complex three-dimensional computational fluid dynamics (CFD) codes. The experimental setup was composed of two rooms (fire compartment and upper room) mechanically ventilated, with the fire source being a liquid pool fire. Two parameters were investigated: the buoyancy forces due to the temperature difference at the vent and the inertia forces due to the difference of pressure between the compartments. The effects of these parameters were investigated through the four VSP experimental tests. The results of these experiments are reported in this special issue in the paper Prétrel and Vaux, Experimental and numerical investigation of the smoke propagation in case of fire event within two confined and ventilated compartments connected with a horizontal opening. The objective of the second campaign, called FES for Fire Extinction System, was to assess the efficiency of two fixed water-based fire suppression systems. This topic is of great interest because such systems with water as extinguishing medium are often used to suppress fires in switchgear and cable rooms of nuclear power plants. Following initial sensitivity study, two key parameters were selected for investigation, droplet size distribution and water flow rate, to determine how they affect the efficiency of the water-based fire suppression systems. The fire tests investigated these effects by testing two types of industrial sprinkler and deluge nozzles with two water flow rates and two different activation times. A set of four fire tests was defined using a liquid pool as fire source. More details are given in the work of Vaux and Prétrel, Experimental and numerical study of the efficacy of water spray application in case of a fire event in a confined and mechanically ventilated compartment. On the basis of this experimental campaign, a theoretical investigation was undertaken, evaluating the concept of repeatability on large-scale fire tests. The results are reported in the paper by Prétrel and Querre, Repeatability assessment of large-scale fire experiment involving water spray system in a forced ventilated compartment. This original work provides important information for the validation of numerical simulations on this type of fire scenario. The third campaign, called CFS for Cable Fire Spreading, focused on studying fire spread for the two following configurations: fire spread over cables trays and fire spread from one electrical cabinet to other electrical targets, in mechanically ventilated fire scenarios. This campaign consisted of eight fire tests in the DIVA facility. The fire tests CFS 1 to 4 were composed of five horizontal cable trays and involved three electrical cable types, provided by the partners (GDF-SUEZ, VTT, and NRA). In addition, two ventilation renewal rates (high and low) were investigated. Some of these fire tests have been analyzed in the work of Zavaleta et al., Cable tray fire tests with halogenated electric cables in a confined and mechanically ventilated facility. Some other analyses can be found in Zavaleta and Audoui.16 The three, CFS-5 to CFS-7, fire tests involved fire sources represented by a real open-door electrical cabinet and three overhead cable trays. Specific objectives of these experimental tests were to investigate how effects of ventilation, fire dampers shutdown, and cable-type–influenced fire growth and spread in a confined and mechanically ventilated environment. The results of these fire tests are presented in the work of Zavaleta et al., Fire spread from an open-doors electrical cabinet to neighboring targets in a confined and mechanically ventilated facility. The final PRISME 2 campaign conducted tests in both open atmosphere and in the confined and mechanically ventilated DIVA facility. The open atmosphere tests were aimed at studying the impact of cable tray configuration on fire spread over multiple cable trays, including protected cable trays and slanted cable trays.17 The fire tests in a confined environment were designed to complement the previous FES and CFS campaigns. The work of Shirai et al., Experimental study of smoke effects on energized electrical cabinets located nearby a lubricant oil pool fire, reports on one of these tests. In addition to the characterization of these experimental fire tests and their physical analysis, the PRISME 2 project members wanted to share with the nuclear fire community the valuable and novel numerical work conducted during the scope of the PRISME 2 project. This learning is presented within this special issue. The modeling tools used in the various papers cover both fire zone models such as SYLVIA, lumped parameter models like COCOSYS, and highly complex three-dimensional models such as ISIS or FDS. The results provided by these different numerical studies undoubtedly make it possible to get further confidence in extending the validation domain of the different fire models. However, although improvements in fire modelling have been demonstrated, the application of these models for complex fire scenarios has illustrated the limitations of current modelling capabilities. This learning must be taken into account in future experimental programmes. In this direction, a number of recommendations have been provided by the PRISME 2 members for targeting further phenomena not exhaustively studied in PRISME 2. These phenomena are smoke stratification and spread, fire propagation between electrical cabinets, and electrical cable tray fires in confined and ventilated conditions. The ongoing follow-on PRISME 3 Project aims at addressing the above mentioned three phenomena and to provide answers to various issues of interest for nuclear fire safety analysis. The authors are grateful for the financial support of the participating OECD/NEA member countries to the joint OECD PRISME 2 Project. Moreover, they want to acknowledge the outstanding support provided by the members of the program review group for the many fruitful discussion during the PRISME 2 meetings as well as by the OECD/NEA secretariat making by their active and valuable contributions this activity a successful one: F. Bonte and C. Fourneau from Bel V (Belgium), E. Gorza and L.P. Kwahou Kesembo from Tractebel-ENGIE, (Belgium), A. Bounagui from CNSC (Canada), A. Matala and T. Sikanen from VTT (Finland), S. Hostikka from Aalto University (Finland), C. Lallemand from DGA (France), L. Gay, B. Sapa, B. Gautier, from EdF (France), T. Morii and S. Fujita from NRA (Japan), K. Shirai from CRIEPI (Japan), J. Peco Espinosa from CSN (Spain), C. Karlsson from SSM (Sweden), S. Ledin and L. Nyogeri from ONR (United Kingdom), A. Kelsey from Health & Safety Laboratory (United Kingdom), and Andrew White (NEA Secretariat).

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.