Fires are devastating events that may harm humans, properties and the environment. Authorities, organisations, companies and societies should be able to learn from fire incidents to reduce the probability and impact of future fires. To achieve a reduction in fires and their consequences, an effort is needed from multiple actors and both technical, organisational and individual changes would be necessary. Importantly, we therefore consider change as a prerequisite for learning. So how can we as a society change or modify our efforts for prevention and mitigation of fires? A learning approach often starts with some form of inquiry about the occurred accidents – an investigation. This investigation can take many forms: the fire services’ own evaluations of the response to a fire, an authority’s assessments of the compliance and fit of their regulations, a company’s analysis of internal rules and organisation, and the police’s investigation of criminal issues. Investigations require highly skilled professionals using often multidisciplinary skills such as knowledge in human behaviour, fire dynamics, electrical systems, mechanical processes and many more. A fire investigator may use many different techniques and tactics, to figure out how the fire started, what fault led to the fire, what made the fire develop the way it did and, also what factors and measures that worked well in the fire. The investigator can work for the fire service, the police, insurance companies, hired private fire investigators or in larger companies, to mention a few. However, not all fires are investigated in Norway, and there is also a large number of incidents that is concluded with an unknown fire cause. The aim of our research has been to increase the society’s capacity to learn from fires. We have two main objectives contributing to the aim: 1. Obtain knowledge on the preconditions for learning from fires in Norway. 2. Provide recommendations to increase learning from fires in Norway.

This is the abstract from Nordic Fire and Safety Days conference, from the FRIC project "Learning from fires in Norway". This abstract focuses on the investigation phase of that process.

Modern buildings are being built with increasingly complex technical installations and energy systems. The introduction of renewable energy production, like photovoltaic (PV) panels on building roofs and facades and an increasing number of connected electric appliances, changes the way the electric power is distributed from production to end-user. The difference in production and demand for energy over time also gives incentives for installing energy storage systems. Electric energy can be stored in batteries, transferred into hydrogen gas via electrolysis or stored as thermal energy for use later. The current work presents an overview of an ongoing study in the Fire Research and Innovation Centre (FRIC), on fire safety implications related to implementing new technology for energy storage and production. The focus is on the built environment such as dwellings and office buildings situated in the Nordic countries. This study builds on previous studies of related topics.

Based on an experimental study of self-sustained smol-dering in a granular biomass fuel bed [1], a compre-hensive three-dimensional numerical model was developed. The model (computational fluid dynamics - discrete element method solver) considers the com-plex heterogeneous bed structure of wood pellets storage (Fig. 1), describing the interaction between the granular fuel itself and the transfer of heat and oxy-gen. It includes a Lagrangian descrip-tion of the fuel bed (instead of the com-monly employed volume averaged continuum assump-tion), which allows the individual treat-ment of particle shrinkage and, hence, the model-ling of the changing fuel bed structure. Results and discussion Preliminary results show that the newly developed model, is able to simulate self-sustained smoldering after 6 h, when external heating is turned off (see Fig. 2). Furthermore, the model has shown the poten-tial to describe the most important characteristic of smoldering in granular fuel beds, including localized hotspots, and its random nature (see Fig 3). Since the particles are not resolved a coarse grid for the fluid flow can be employed reducing the numerical costs significantly compared to micro-scale models. The promising initial results motivated ongoing and future model improvements and validations focusing on: • different convective heat transfer correlations, • the overlap between different stages (drying, de-volatilization , fuel oxidation and char oxidation), • heat of reaction (devolatilization, fuel oxidation,and char oxidation), • kinetics (devolatilization, fuel oxidation, and charoxidation) and • a parameter study of numerical parameters.

Presentation given to GjensidigeStiftelsen (in Norwegian)