Earticle

Fires involving metal dust can exhibit rapid flame propagation and complex combustion behavior, particularly when deposited dust becomes suspended and interacts with an existing flame. Under such conditions, abrupt flame expansion and short-duration intense thermal exposure may occur. This study conducted reconstruction experiments to examine the combustion behavior and thermal exposure characteristics of metal dust under controlled conditions. The experiments were designed to compare deposited and suspended dust states, sustained flame environments, and varying disturbance conditions. Flame propagation patterns, upward flame jet formation, and thermal distribution were observed through visual documentation and thermal imaging. The results suggest that suspended metal dust exposed to an existing flame can cause rapid flame expansion and concentrated upward flame jets, accompanied by short-duration but intense thermal exposure. In contrast, deposited dust under identical ignition conditions exhibited limited flame spread and localized combustion. These findings suggest that combustion behavior depends more on dust suspension and flame interaction than on dust quantity alone. These results provide experimental observations that may be used to understand combustion patterns and thermal exposure characteristics in metal dust fire incidents.

This study aims to analyze the causes of the Wang Fuk Court apartment fire tragedy in Hong Kong in November 2025, which resulted in over 160 fatalities, and to propose institutional improvement measures for the fire safety of high-rise buildings in South Korea. This tragedy was a foreseeable man-made disaster caused by a combination of vulnerable building materials and inadequate management of firefighting facilities. The bamboo scaffolding used for exterior repairs, along with combustible protective nets and styrofoam insulation, created a chimney effect that rapidly spread the flames to the upper floors. Notably, the intentional deactivation of sprinkler valves and fire alarm systems—under the pretext of construction convenience and avoiding noise complaints—was identified as the decisive human error that rendered initial fire suppression impossible. Consequently, this study derived policy implications for South Korea by comparatively analyzing the fire administration case of Macau, which successfully deters large-scale fires through stringent regulations and real-time monitoring systems. Based on these findings, this study proposes key countermeasures to strengthen the domestic fire safety net: legislating the use of flame-retardant materials for construction site protective nets, mandating the placement of fire extinguishing equipment in temporary structures, establishing strict penalties for the arbitrary deactivation of firefighting facilities, and implementing a real-time monitoring system. In conclusion, this study reaffirms social safety as an essential value and provides practical guidelines to prevent future large-scale fire tragedies through proactive legislative amendments.

Rapid urbanization and industrialization have increased the complexity of fire scenes, revealing the limitations of traditional investigation methods. Conventional approaches often expose personnel to life-threatening hazards, such as toxic gases and structural collapses, while relying on subjective 2D photography and manual measurements. To address these challenges, this study develops a methodology utilizing drone-based 3D mapping and AI analysis to enhance the speed, accuracy, and safety of fire investigations. By integrating high-resolution thermal imaging and 3D spatial data, the proposed system allows for the digital reconstruction of complex scenes without requiring immediate human entry into dangerous areas. Comparative experiments using the DJI Matrice 30 demonstrated that drone technology significantly outperforms manual methods. The drone-based system achieved over 90% accuracy in tracking ignition points—a 40% improvement over traditional techniques—and reduced investigation time from 90 minutes to just 10 minutes for a 300 $m^2$ site. Furthermore, the 3D mapping analysis yielded a remarkably low error rate of 0.81% in combustible material calculations, compared to 19.35% in manual inspections. These findings suggest that 3D spatial integration and automated data processing drastically improve the scientific reliability and legal utility of fire investigation results.

This study proposes the H.E.L.P (Human Evacuation Location Prediction) model, which reconstructs National Fire Data System (NFDS) data using artificial intelligence to quantitatively predict evacuee behavior patterns and rescue priorities during fires. Despite extensive NFDS records, current use is confined to statistical and administrative reporting, with limited support for real-time decision-making. An AI training dataset integrating individual characteristics (age, gender, disability, movement type) and environmental factors (fire location, smoke spread, emergency stairway accessibility) was developed and applied to deep learning and SARSA-based reinforcement learning simulations. Experimental results demonstrated 96–100% recognition accuracy, over 80% prediction accuracy, and more than 20% reduction in average evacuation time. Additional tests using RandomForest and XGBoost achieved high casualty prediction accuracy, confirming NFDS’s potential for real-time rescue strategy support. The model’s integration with IoT-based recognition and AR/VR command systems can enhance precision and efficiency in firefighting operations, offering a scalable public safety AI platform that unites policy, technology, and data.

Recent fire statistics released by the Korean Fire Agency indicate that a total of 678 lithium-ion battery-related fires occurred in South Korea over the past five years (2020–2024). While electric kickboards and electric bicycles account for the majority of these incidents, fires involving electronic cigarettes and disposable electronic cigarettes have continued to occur persistently. As the use of electronic cigarettes becomes more widespread, particularly among adolescents, concerns regarding fire safety are increasing. This study experimentally investigates ignition mechanisms of electronic cigarettes and disposable electronic cigarettes by reproducing three critical failure conditions: overcharging, mechanical compression, and e-liquid leakage combined with moisture infiltration. Controlled experiments were conducted on lithium-ion and lithium-polymer batteries commonly used in e-cigarette devices. Under overcharge conditions without functional protection circuits, thermal runaway occurred within 45–75 seconds, causing battery rupture or explosion. Mechanical compression tests showed that loads exceeding 150 kgf can induce internal short circuits and thermal deformation, suggesting ignition risks during disposal. Prolonged exposure to e-liquid and moisture further degraded protection circuits, resulting in delayed ignition. These results confirm that fires involving electronic cigarettes can originate not only during usage and charging but also during disposal stages, depending on battery failure modes. The findings underscore the need for improved charging safety practices, enhanced protection circuit designs resistant to liquid ingress, and systematic disposal management for battery-containing electronic cigarette products.

The utilization of electricity, which enhances the convenience of daily life and significantly contributes to scientific and economic advancement, continues to increase. However, improper handling or misuse of high-risk electrical systems can lead to serious incidents such as electric shocks and electrical fires. Among the primary causes of such incidents, electrical leakage has been identified as a critical risk factor. Fires resulting from leakage can originate at any point along the path between the leakage point and the grounding terminal. This study aims to experimentally investigate how environmental conditions surrounding potential ignition points influence the likelihood and characteristics of ignition. By analyzing the ignition and combustion behavior at the origins of electrical leakage fires and identifying the environmental factors that may accelerate combustion, the findings of this study are expected to provide valuable insights for fire investigation methodologies and serve as practical resources for public fire safety education and awareness.

While the demand for high-quality fire investigation reports continues to grow, fire scene visualization remains a significant challenge for practitioners. This study analyzed the current state of visualization practices through a survey of 130 fire investigators at the Gyeongbuk Fire Service Headquarters and explored practical improvement measures through an eight-month case study using Blender, an open-source 3D software. The survey results revealed a significant discrepancy: although 70.3% of respondents acknowledged the importance of visualization, “difficulty in software operation” was identified as the primary barrier (45.7%), and 73.5% expressed a need for systematic professional training. The case study confirmed that Blender is a viable tool capable of overcoming the financial and functional limitations of existing methods. Ultimately, this research identifies that the root cause of the problem lies not in the tools themselves, but in the lack of fundamental education regarding 3D concepts. Accordingly, this study proposes the establishment of a customized education system for fire investigators as a key strategy to enhance their visualization capabilities.