Fire emergency management is an extremely complex and challenging process that requires the integration of real-time data to produce optimal solutions. Accordingly, this review specifically conceptualizes fire as a natural disaster and examines AI-supported decision support systems across the fire emergency management lifecycle. In this context, developments in artificial intelligence technology in recent years have been extremely effective in solving this problem, especially in real-time data transfer. Most existing studies in the literature focus on technological approaches and operational processes. However, the effective integration of decision support systems into emergency management remains lacking. This study presents a systematic literature review aimed at analyzing how the integration of decision-making methods with the fire management lifecycle is performed in the literature. The 69 articles examined in our study were clustered based on their themes and methods used. Thus, the relation of each cluster in the fire management lifecycle was determined. The results show that there are numerous technology-supported studies in the detection, monitoring, and prevention phases. Conversely, it is clearly revealed that there are a limited number of studies in the improvement phase. Furthermore, it was concluded that studies integrating AI and decision-making methods throughout the entire fire management lifecycle are limited. This study proposes a conceptual perspective for the development of end-to-end smart fire emergency management systems. The findings are guiding for future research into more proactive, integrated, and decision-oriented AI frameworks that can support holistic fire emergency management.
Gypsum boards are widely used as passive fire protection materials due to their ability to delay heat transfer through endothermic dehydration of calcium sulfate dihydrate (CaSO4·2 H2O). Understanding the temperature-dependent evolution of porosity and pore size distributions in gypsum board is essential for defining microstructural parameters used in fire-scale thermal transport models. In particular, the evolution of pore structure during dehydration and high-temperature degradation influences heat transfer within the material and therefore affects its fire protection capability. Quantitative data describing how the pore structure evolves at high temperatures remain scarce, limiting the accuracy of multi-physics fire simulations. This work experimentally characterizes the temperature-dependent evolution of porosity and pore size distributions in gypsum board exposed to temperatures up to 1000°C. Samples subjected to controlled thermal treatments were analyzed using mercury intrusion porosimetry, optical microscopy, and large-area optical scanning in order to capture pore structures across multiple length scales. The resulting dataset provides a comprehensive description of pore structure evolution associated with dehydration, fragmentation, and high-temperature sintering processes. The experiments reveal a bimodal pore system consisting of micro-pores (< 10 μm) formed during dehydration and stable meso-/macro-pores (10–100 μm) generated by foaming agents. Total porosity increases with temperature, while the average pore diameter decreases, indicating pore multiplication rather than coalescence. The experimentally measured pore characteristics were incorporated into a microstructure-informed thermal transport model (GPRO: Gypsum PROperties) in order to evaluate the influence of representative pore length scales on the radiative component of heat transfer and to assess their impact on the effective thermal conductivity (keff). Model predictions were compared with literature data for the effective thermal conductivity of gypsum board at elevated temperatures to assess the consistency of the derived microstructural parameters with independently measured macroscopic properties. The results show that the evolution of total porosity strongly affects the conductive component of heat transfer, while the radiative conductivity predicted by diffusion-based radiation models is primarily controlled by pore lengths in the meso-scale range identified in the experimental measurements. Sensitivity analysis indicates that variations in representative pore diameter between 1 μm and 100 μm lead to changes in predicted effective thermal conductivity below 3
Fires or explosions caused by crude oil tank leaks are prone to triggering domino effects, expanding the scope of the incident and forming a catastrophic chain of accidents. To enable the rapid prediction of accident consequences in the event of a domino effect in crude oil storage tank areas, this paper proposes a Bayesian Network-based consequence assessment method for pool fire domino effects in crude oil storage tank areas. First, the thermal radiation hazard range of pool fires is analyzed, and a harm radius correction model considering wind speed effects is constructed. Then, based on Bayesian Network analysis, the probabilities of pool fires occurring between tanks and triggering domino accidents are evaluated. Next, combining the harm radius model of pool fires with the tank failure probabilities under the domino effect, individual risk is assessed, and risk contour maps are drawn. Finally, the harm radius of pool fires is simulated using FDS software. A comparative analysis reveals that the average relative errors of the proposed harm radius correction model compared with the simulation values in the upwind and downwind directions are 0.15 and 0.11, respectively, indicating a high level of accuracy. The consequence assessment method proposed in this paper avoids cumbersome calculations by comprehensively analyzing the accident propagation path, facilitating rapid and effective risk analysis of tank area fires and the formulation of corresponding response measures.
Lithium-ion battery systems are increasingly deployed across vehicles, buildings, and energy infrastructure, raising urgent fire safety challenges that are not yet fully addressed by existing codes and standards. This article presents findings from an international survey of fire safety professionals with experience in battery-related projects, designed to capture current practice, perceived gaps, and methodological limitations. The survey reveals strong reliance on heat release rate as a principal design parameter, yet limited recognition of the uncertainties associated with its measurement. Respondents emphasized the lack of validated, system-level test data, particularly on gas composition (primarily toxicity but also flammability), and concerns related to UL9540A goals, quality and interpretation. Concerns about transparency and unequal access to test data were repeatedly noted, alongside the absence of harmonized regulatory guidance across regions. Taken together, the results provide one of the first systematic insights into how fire safety engineers approach battery fire safety design/strategy in projects involving battery systems in practice and identify areas where research, regulation, and industry practice must converge to enable safe large-scale deployment of battery systems.
Thermal runaway propagation (TRP) in electric-vehicle battery modules is strongly influenced by leakage pathways at seams and interfacial gaps, where hot gases and flame jets can convectively transfer heat to adjacent cells. In this work, an intumescent-coated textile wrap was developed to autonomously self-seal such discontinuities and thereby suppress TRP. A para-aramid plain-weave fabric was coated using a silicone-binder intumescent system to produce a mass-matched set comprising uncoated (T0), binder-only (T1), and intumescent-coated (T2) textiles. Thermogravimetric analysis in air showed a substantially higher residue for T2 (21.8 ± 1.9
Automated storage systems and large-scale freezers can be challenging for conventional fire protection systems. Instead, an oxygen reduction system (ORS) could be a suitable option. In an ORS the design oxygen concentration is the key parameter for the level of protection achieved by the system. However, current methods for determining this concentration yield varying results and either require large and complex test setups or arrive at overly optimistic levels for protection. In this paper, an experimental method for using the controlled atmosphere cone calorimeter (CACC) to find the limiting oxygen concentration at extinction (LOCext) is proposed. The results from the CACC showed good agreement with limiting oxygen concentrations (LOC) determined by large-scale; indicating that the CACC could serve as a cost-effective alternative for determining the LOC of ordinary combustibles for ORS applications. Furthermore, thermal analysis and experiments using both nitrogen and carbon dioxide highlights that the thermal properties of the diluent gas, rather than oxygen displacement and chemical quenching, are the primary factors influencing the LOC. Hence, the findings in this paper increase the understanding of ORS systems and offer a way to determine the LOC for ORS based on small-scale testing.
Oxygen availability at a pyrolyzing surface during turbulent flame spread remains poorly understood, and the governing fire dynamics are not yet fundamentally resolved. A series of experiments was conducted using a wall-corner configuration to measure surface oxygen concentrations during flame spread at two distinct locations under a 300 kW hood, using both burner-generated and solid-fuel-generated flame conditions. In burner tests, a propane burner was positioned at the base corner, with mass flow adjusted to achieve steady HRRs. In solid-fuel tests (PMMA, Douglas Fir), ignition was initiated at a 30 kW propane HRR. Gas sampling tubes were flush-mounted to one of the panels at a vertical height of 100 cm above the burner and at lateral distances of 10 cm and 20 cm from the corner to extract combustion gases. A novel technique was introduced to non-intrusively estimate surface oxygen availability using a correlation between the measured oxygen volume fraction and the normalized image intensities. Radiation intensity measurements were used to quantify fire growth rates, which were, on average, 3.5 times greater in the vertical direction than in the horizontal. Flame sheet growth and acceleration were found to directly influence the measured surface oxygen volume fractions. Time-resolved surface oxygen fields were reconstructed for PMMA, providing new insight into oxygen depletion dynamics across distinct pyrolysis regimes during a turbulent fire spread.
Automated storage systems and large-scale freezers can be challenging for conventional fire protection systems. Instead, an oxygen reduction system (ORS) could be a suitable option. In an ORS the design oxygen concentration is the key parameter for the level of protection achieved by the system. However, current methods for determining this concentration yield varying results and either require large and complex test setups or arrive at overly optimistic levels for protection. In this paper, an experimental method for using the controlled atmosphere cone calorimeter (CACC) to find the limiting oxygen concentration at extinction (LOCext) is proposed. The results from the CACC showed good agreement with limiting oxygen concentrations (LOC) determined by large-scale; indicating that the CACC could serve as a cost-effective alternative for determining the LOC of ordinary combustibles for ORS applications. Furthermore, thermal analysis and experiments using both nitrogen and carbon dioxide highlights that the thermal properties of the diluent gas, rather than oxygen displacement and chemical quenching, are the primary factors influencing the LOC. Hence, the findings in this paper increase the understanding of ORS systems and offer a way to determine the LOC for ORS based on small-scale testing.
Over the past decade, there have been numerous thermal runaway events in outdoor, large-scale, lithium-ion (Li-ion) battery energy storage system (BESS) installations. BESS thermal runaway events, which can be several hours to days in duration, have led to shelter-in-place orders, community evacuations, and roadway closures and have raised concerns about impacts on human health. In response to these concerns, some jurisdictions now require plume dispersion modeling to accompany hazard mitigation analyses and risk assessments for new BESS sites. In this study, a methodology for fire and plume dispersion modeling of BESS flaming thermal runaway was developed utilizing Fire Dynamics Simulator (FDS) and AERMOD. The baseline scenario involving a generic BESS unit containing LFP cells was developed for purposes of demonstrating the methodology. Data from the current literature was aggregated to synthesize modeling inputs. A bounding analysis was conducted for meteorological conditions and incident duration. Toxic gas concentrations were compared against established human health impact guidelines to determine exposure limit distances. When evaluating human health impact in both FDS and AERMOD, HF produced the furthest exposure limit distances. Gases such as benzene, CO, CO2, and NH3 have high PACs limits, so regardless of the emission rates, these constituents did not produce the furthest exposure distances. The proximity of the plume to breathing height was significantly impacted by wind speed. The higher wind speeds pushed the plume closer to the ground, resulting in further exposure limit distances. All worst-case scenarios were produced at the 10 m/s wind speed, regardless of temperature. Increases in incident duration tended to decrease exposure limits distances but did not have as pronounced effect compared to wind speed. The results showed that, while AERMOD may be capable of high-level predictions of long-range plume dispersion and grounding, FDS is likely more suitable for near-field analysis (up to 100 m) in scenarios involving combustion (and buoyancy driven flow). The accuracy of the plume dispersion models is dependent on credible failure events and modeling inputs, such as heat release rate, mass loss rate, emissions, and meteorological conditions. Moreover, the engineer must ensure that the model is appropriate for the given scenarios being evaluated and the desired outputs.
Fires have always posed a threat to human life and attributes. Traditional firefighting methods have several shortcomings that need to be addressed, such as low efficiency, high risk, and high cost. The development of robotics has garnered greater attention across many areas recently, especially in the design of specific robotic platforms, certain application scenarios, and breakthroughs in key technologies. This study follows PRISMA guidelines and provides a systematic review of 141 articles published from 2016 to 2025, drawn from the Web of Science, Scopus, and IEEE Xplore databases. This review aims to bridge the gap between emerging fire-related technologies and their practical implementation in fire safety engineering. This study reviews various categories of firefighting robotic platforms, technologies, and application scenarios. It discusses current problems, such as the reliability of perception in extreme conditions, the robustness of communication, operational endurance, the efficiency of multi-robot coordination, and the ease of human-robot interaction. By organizing the literature from an engineering decision-making perspective, this review provides a framework for translating research outcomes into practical fire safety applications.
To investigate the mechanical properties of S-PVA blended fiber-reinforced, high-strength, self-compacting concrete after exposure to high temperatures, this study conducted compressive tests on four groups of C60 concrete specimens after optimizing the mix design using a backpropagation (BP) neural network. The study analyzed interactions between temperature and fiber volume ratio and established a model to predict the compressive strength. The results showed that as the temperature increased, the specimen surfaces transitioned from light gray to grayish-white, along with expanded surface cracks and gradual mass loss. Self-compacting concrete exhibited an initial strength increase followed by a decline. Although the use of PVA fibers alone slightly reduced the compressive strength, their combination with steel fibers significantly enhanced the performance, with 0.1
Fire-induced spalling remains an ongoing concern for concrete structures exposed to fire conditions as existing predictive methods struggle to achieve reliable predictions. To bridge this knowledge gap, this study proposes a methodology to develop empirically driven indices to predict the spalling phenomenon. The methodology comprises three approaches: systematic data transformations, logistic regression, and a data-informed boundary decision based on machine learning (ML). The obtained indices, evaluated over a comprehensive dataset of over 1,000 fire tests and 200,000 cases, exhibited an accuracy ranging from 71–82
Firefighters routinely operate in extreme thermal environments where exposure to radiant heat, flames, and flash fires can degrade the protective performance of firefighter protective clothing. The effectiveness of these garments depends on key properties such as thermal insulation, reflectivity, porosity, and moisture management. While thermal insulation limits heat transfer to the skin, moisture accumulation from perspiration or environmental humidity can enhance heat and mass transfer within the clothing system. Moisture evaporation and condensation further exacerbate thermal stress, increasing the likelihood of burn injuries. This review critically examines the thermal protective performance of firefighter garments under high-temperature and moisture-rich conditions. It synthesizes current knowledge on coupled heat and moisture transport, fabric architecture, air-gap effects, material properties, and multilayer garment configurations. Advances in experimental characterization, numerical modeling, and predictive assessment of thermal degradation are also evaluated. Finally, key research gaps and future directions are highlighted to improve garment aging assessment, damage quantification, and service-life prediction, thereby supporting the development of safer, more durable, and high-performance firefighter protective systems.
Two façade insulation systems, a composite External Thermal Insulation Composite System (ETICS) and a ventilated ‘system with High-Pressure Laminate (HPL) cladding, were tested at full scale using the French LEPIR 2 test bench. The characteristic geometric scales of the external flames (height, width, and ejection angle) were determined from front and side-view images. A dedicated image analysis method, adapted to outdoor imaging conditions, was developed and applied to frames acquired a few minutes after ignition and the emergence of external flames, during a stable period of the fire. The results of the characteristic geometric scales are presented and discussed. Based on these parameters, flame volumes were estimated, allowing the calculation of external Heat Release Rate (HRR). The HRR was found to be higher for the HPL ventilated cladding system, suggesting a higher contribution of the HPL material to the overall combustion. Using the HRR values, the flame heights were predicted using a classical correlation, and the comparison with the observed ones is discussed.
Hydrogen-blended natural gas (HCNG, primarily composed of CH₄–H₂ mixtures) holds high greenhouse gas emission reduction potential, but its large-scale application is hindered by fire and explosion risks under high-temperature and high-pressure conditions. To reduce these risks, ReaxFF molecular dynamics (ReaxFF-MD) was conducted to investigate the effects of CF₄ on CH₄–H₂–O₂ combustion at 2400–3200 K and 100 MPa. Results revealed that CF₄ significantly suppressed combustion through multiple mechanisms. Specifically, with CF₄ addition, OH and HO₂ radicals were consumed via combination with intermediates generated from CF₄, such as CH3F. Meanwhile, CH₃ radicals gradually accumulated due to blocked consumption pathways. Moreover, F radicals competed with the CH₄–H₂–O₂ chain reaction through reactions such as H₂ + F → HF + H, thereby reducing the concentration of radicals (e.g., OH). Additionally, quantitative analysis showed that at 10
Cargo-related fires on container vessels have become a critical safety concern in maritime transportation due to their increasing frequency and the severe human, environmental, and economic losses they cause. The multidimensional nature of these incidents requires that fire risk be addressed not only in terms of technical failures but also in relation to system-level weaknesses in human, organisational, and operational processes. The aim of this study is to classify the causes of cargo-related fires in the cargo areas of container vessels using the Human Factors Analysis and Classification System (HFACS) and to identify critical risk patterns using Association Rule Mining (ARM) techniques. Full-text accident investigation reports obtained from official databases were systematically coded according to the HFACS-PV framework. The resulting qualitative data were transformed into a binary dataset and analysed using the Apriori and Predictive Apriori algorithms. The results indicate that misdeclaration and misclassification of cargo under the IMDG Code, inadequate documentation control, stowage planning errors, and deficiencies in information sharing are key contributors to fire initiation. Furthermore, reduced situational awareness, communication failures, and delayed response behaviours were found to intensify fire severity under adverse operational conditions.
Perfluorohexanone (C6F12O, FK-5-1-12) is a fluorinated fire-extinguishing agent with high suppression efficiency and distinctive engineering applicability. However, systematic research on the flow mechanisms and engineering design methods of C6F12O fire-extinguishing pipe networks remains limited. This paper develops a theoretical pipe-network model for gaseous fire-extinguishing systems and conducts C6F12O discharge experiments. The effects of the expansion coefficient on the discharge process and flow characteristics are examined, and a recommended range is proposed. To address practical engineering design needs, a general engineering method for fire-extinguishing pipe networks is proposed without relying on empirical parameters. The results show that the expansion coefficient significantly affects the discharge duration, and γ = 1.15–1.20 is recommended for engineering calculations. The operating process of gaseous fire-extinguishing systems can be divided into four typical stages. Theoretical predictions agree well with experimental measurements qualitatively, and the quantitative deviations are within 20
Low-rise buildings are widely distributed in residential areas and small commercial zones. Therefore, ensuring their fire safety is crucial for protecting lives and property. However, existing fire risk assessment methods involve incomplete index systems and do not sufficiently consider risk factors in fire scenes, thereby failing to identify hazards accurately. Consequently, these methods cannot be used to formulate effective fire prevention strategies. Accordingly, this study developed a fire risk assessment model based on extension cloud theory and numerical simulation for low-rise buildings. It established an evaluation index system covering the dimensions of personnel, equipment, management, building, environment, and dynamic risk in fire scenes and then determined index weights by using a combination of the Decision-Making Trial and Evaluation Laboratory and best–worst methods. Dynamic risk factors were simulated using PyroSim to analyze how temperature, CO concentration, and visibility affected risk. The following fire development patterns were identified by comparing five scenes: (1) smoke spreads preferentially to the top floor and the floor on which the fire originates; (2) stairway fires cause vertical smoke movement that endangers upper-floor evacuation; (3) fires in the middle of a corridor cause heat accumulation and thus hinder evacuation from the corresponding floor; (4) a mechanical smoke exhaust system with a reasonable layout reduces fire risk; and (5) automatic sprinkler systems can cool the surroundings, capture smoke, improve visibility, and increase the time available for evacuation. Overall, this study offers new insights and methods for fire risk assessment in low-rise buildings.
The compressed air foam system (CAFS) has demonstrated significant potential for use in combating flammable liquid fires at ultra-high voltage (UHV) substations in high-altitude environments. However, the system performance and fire extinguishing performance of CAFS at high altitude are unclear. In this study, the performance of the air compressor, water pump, and CAFS at different altitudes was studied. Besides, the fire extinguishing performance of CAFS at different altitudes, different system flow rates, and different expansion ratios (ERs) was also studied. Results indicated that as the altitude increases, the gas pressure decreases, the pressurization time of the air compressor increases, and the pressurization efficiency decreases. The flow and outlet pressure of the water pump increased with increasing altitude. Besides, the fire extinguishing performance of the CAFS is improved as the altitude or the system flow rate increases. The optimal parameter settings for CAFS with excellent system performance and fire extinguishing performance at high altitude are 11.4 L/min of the system flow rate and 8 times the ER. The results of this paper can guide flammable liquid fire prevention and control in UHV substations at high altitude.
