ABSTRACT During oil transportation, oil leaks are a common occurrence that can result in the formation of spill fires, potentially causing significant damage to surrounding facilities. This study experimentally investigates the flame height characteristics of small-scale spill fires under varying fuel discharge rates. A series of spill fire tests were conducted, involving the continuous release of n-heptane as fuel on a water cushion surface. The recorded time-varying burning area and the measured stable flame height values indicate that the evolution of the spill fire can be characterized by four distinct phases. By analyzing the relationship between the fuel discharge rate and the pool diameter, a functional correlation between the two variables was established. The flame height of a spill fire in the quasi-steady state phase was compared with that of a pool fire achieving steady combustion under similar scale conditions. The analysis revealed that the primary factor contributing to the difference in flame height was the disparity in the burning rates between the pool fire and the spill fire. Upon comparing the measured flame height from the experiment with the predicted flame height from the pool fire model, it was found that the model lacked satisfactory predictive accuracy. By examining the plume velocity and air entrainment in the spill fire, this study proposes a new functional correlation between the steady fuel discharge rate and the flame height during the quasi-steady state phase. The flame height relationship derived in this work demonstrates superior predictive accuracy compared to the Heskestad and Zukoski models for flame height prediction.

This study examines the influence of thermal treatment on the ignition properties of Norway spruce (Piceaabies (L.) H. Karst.) and sessile oak (Quercus petraea (Matt.) Liebl.) wood. Using a cone calorimeter both untreated and thermally modified samples (180°C for 6 h) were analysed to determine key fire modelling parameters: combustion efficiency, critical heat flux, ignition temperature, thermal inertia, and thermal response parameter. Obtained results reveal that thermally treated wood exhibits higher combustion efficiency than its untreated equivalent, with spruce generally outperforming oak. The effect of thermal treatment on other properties was species-dependent. Thermally treated spruce showed an increase in critical heat flux and a decrease in both thermal inertia and the thermal response parameter. Conversely, thermally treated oak displayed a reduction in critical heat flux and an increase in both thermal inertia and the thermal response parameter. These results highlight the complex, species-specific effects of thermal modification on the fire behaviour of wood.

Purpose The aim of the study is to investigate the applicability of the new Eurocode localized fire model for the derivation of hazard functions. Moreover, transient analysis is introduced in the paper to refer to intensity measures (IM) to design parameters defined in the time domain. Design/methodology/approach The study employs a frequentist approach approximated with the simple Monte Carlo (SMC) technique for the assessment of IM. The new Eurocode localized fire model was broadly discussed and then used as a deterministic model for SMC with net heat flux and steel temperatures as output variables. The sensitivity of the model was evaluated with the Spearman rank. Findings The results reveal that a localized fire model may be useful and, in some topologies, even more appropriate than widely adopted gas temperature curves. The model is highly sensitive to changes in the rate of fire growth and the distance between the fire axis and the structural element. Changes in massivity or utilization factor do not affect the final results that much. Originality/value This research fills some of the unexplored niches in the probabilistic structural fire engineering (PSFE) framework and presents its application. The study offers a new perspective on the assessment of IM in probabilistic frameworks with great practical potential in structural safety engineering.

Abstract This study examines the impact of a low-intensity forest understory fire on turbulent heat transfer using sonic anemometer measurements at lower (3 m), mid- (10 m), and upper (19 m) canopy from five in situ towers within the burn plot and one upwind control tower. The fire induced up to a 50-fold increase in kinematic vertical turbulent heat flux, primarily driven by increased temperature perturbations. The maximum flux occurred at midcanopy due to decreasing temperature and increasing vertical velocity perturbations from the lower to upper canopy. A double peak in turbulent heat flux during fire-front passage resulted from interactions between fire-induced turbulence and atmospheric flow, where a downdraft immediately following the fire front temporarily suppressed flux before the second peak emerged. The flux increase was largely due to intensified ejection events rather than a higher event frequency, indicating enhanced heat transfer efficiency. At mid- and upper canopy levels, sweep events increased at all but one tower, while at the lower canopy, inward interactions increased and outward interactions decreased across all towers. The arrival of a sea-breeze front, bringing cooler air and stronger winds, dampened fire effects. During the fire, spectral energy for temperature and heat flux intensified at higher frequencies, particularly in the lower canopy, flattening the spectral curve in the inertial subrange. Buoyancy-driven turbulence dominated heat transfer, although some mechanically generated turbulence was also present. These findings enhance the understanding of fire-induced turbulence and heat transfer, aiding the development of fire behavior and smoke dispersion models for improved fire management. Significance Statement This study investigates fire-induced turbulence and heat transfer during a low-intensity forest understory fire, using unprecedented sonic anemometer data from multiple in situ towers at lower, mid-, and upper canopy levels, along with an upwind control tower. The findings reveal up to a 50-fold increase in vertical turbulent heat flux, peaking at midcanopy, and highlight complex fire–atmosphere interactions. Key dynamics include a strong downdraft trailing the fire front, producing a double peak in heat flux, a substantial rise in ejection event contributions across the canopy, increased inward interactions in the lower canopy, and interactions with an incoming sea-breeze front. These insights enhance fire behavior and smoke dispersion models, informing more effective fire management strategies.

Modeling of urban fire spread with integrated cellular automata (CA)-unity framework: Case study on 2025 wildfire in Altadena

An efficient multi-task forest fire and smoke detection model

Pyrolysis of 24 samples of foliage from three U.S. regions with frequent wildland fires (Southeastern U.S., northern Utah and Southern California) was studied in a fuel-rich flat-flame burner system at 765 °C (for Southeastern U.S. samples) and 725 °C (for northern Utah and Southern California species), with a heating rate of approximately 180 °C/s. These conditions were selected to mimic the conditions of wildland fires. Individual plant samples were introduced to the high temperature zone in a flat-flame burner and pyrolysis products were collected. Tar was extracted and later analyzed by GC/MS. Light gases were collected and analyzed by GC/TCD. The estimated range for the average yields of tar and light gases were 48 to 62 wt% and 18 to 31 wt%, respectively. Apart from Eastwood’s manzanita (Arctostaphylos glandulosa Eastw.), aromatics were the major constituents of tar. The variations in the concentrations of tar compounds likely resulted from differences in biomass composition and physical characteristics of the foliage. The four major components of light gases from pyrolysis (wt% basis) were CO, CO2, CH4 and H2. Tar contributed more than 82% of the high heating value of volatiles. These data can be used to improve physical-based fire propagation models.

Fire in urban areas such as Surabaya City is a non-natural disaster that can have a significant impact on public safety, economic stability, and the environment. This study aims to develop a fire risk level classification model using Extreme Gradient Boosting (XGBoost) algorithm based on selected predictor variables, namely response time, fire subtype, and number of victims affected. The dataset consists of 859 fire events throughout 2024, enriched with spatial and demographic attributes. The research methodology involved data preprocessing (including label coding and normalization), class imbalance handling with Synthetic Minority Over-sampling Technique (SMOTE), model training with XGBoost, and evaluation using metrics such as accuracy, precision, recall, and f1-score. The classification model achieved excellent performance, with an overall accuracy of 1.00% and perfect precision, recall, and f1-score of 1.00 across all risk categories (low, medium, and high). Confusion matrix and ROC curve analysis confirmed the high predictive ability of this model. In addition, the results were visualized using a Streamlit-based interactive dashboard to enhance the usability of the model for decision-making. These findings highlight the potential of XGBoost as a powerful tool for fire risk classification and emphasize its relevance in supporting early warning systems and evidence-based disaster mitigation policies in urban environments.

Background Prevailing American wildland fire modelling systems fail to predict fire growth in urban areas due to the absence of burnable urban fuels. Aims This research aims to identify fuel models that optimise fire spread in urban areas relative to a hypothetical fire spread model derived from observations of recent urban fires. Methods A target Rate of Spread (RoS) is derived from observations of seven urban conflagrations to anchor the model to absolute RoS. Exhaustive parameter sweeps are used to identify combinations of fuel variables that result in optimal performance. Key results The target RoS is 0.81 km/h. Parameter sweeps converge on unique sets of fuel parameters including (1) BU0, an unconstrained custom fuel model; (2) BU1, a custom fuel model that operates within the constraints of current US modelling systems; and (3) Anderson Fuel Model 9, a best-performing standard fuel model. Conclusions & implications Although this approach stretches current modelling systems beyond their intended design, the resultant fuel models provide a necessary stopgap for emergency management until urban-specific fire spread models find their way into operational use.

Remote sensing of wildland fires has become an integral part of fire science. Airborne sensors provide high spatial resolution and can provide high temporal resolution, enabling fire behavior monitoring at fine scales. Fire agencies frequently use airborne long-wave infrared (LWIR) imagery for fire monitoring and to aid in operational decision-making. While tactical remote sensing systems may differ from scientific instruments, our objective is to illustrate that operational support data has the capacity to aid scientific fire behavior studies and to facilitate the data analysis. We present an image processing algorithm that automatically delineates active fire edges in tactical LWIR orthomosaics. Several thresholding and edge detection methodologies were investigated and combined into a new algorithm. Our proposed method was tested on tactical LWIR imagery acquired during several fires in California in 2020 and compared to manually annotated mosaics. Jaccard index values ranged from 0.725 to 0.928. The semi-automated algorithm successfully extracted active fire edges over a wide range of image complexity. These results contribute to the integration of infrared fire observations captured during firefighting operations into scientific studies of fire spread and support landscape-scale fire behavior modeling efforts.

Amid the growing frequency of forest fires in southwestern China, this study aims to quantify pollutant emissions and identify key meteorological drivers using multi-source satellite data. Active fire data from Himawari-8/9, MODIS, and VIIRS were integrated to construct a top-down emission inventory for 2016–2023, while the Geodetector method was applied to evaluate meteorological influences. Results indicate mean annual emissions (×103 t·a−1) of 5623.58 (±1554.33) for CO2, 356.84 (±98.63) for CO, and substantial amounts of particulate and gaseous pollutants. Spatially, Yunnan and Sichuan were the dominant emitters; temporally, emissions peaked in January–April and November–December, with daytime levels surpassing nighttime levels. Relative humidity was identified as the dominant meteorological driver (Q = 0.1223), while the interaction between temperature and relative humidity (Q = 0.1486) further enhanced explanatory power. These findings improve the precision of emission inventories and provide essential support for regional fire management and air quality modeling in complex environments.

Electrical cable insulation, mainly composed of polymeric materials, progressively deteriorates under thermal, electrical, mechanical, and environmental stress factors. This degradation reduces dielectric strength, thermal stability, and mechanical integrity, thereby increasing susceptibility to failure modes such as partial discharges, arcing, and surface tracking—recognized precursors of fire ignition. This review consolidates current knowledge on the degradation pathways of cable insulation and their direct link to fire hazards. Emphasis is placed on mechanisms including thermal-oxidative aging, electrical treeing, surface tracking, and thermal conductivity decline, as well as the complex interactions introduced by flame-retardant additives. A bibliometric analysis of 217 publications reveals strong clustering around material degradation phenomena, while underlining underexplored areas such as ignition mechanisms, diagnostic monitoring, and system-level fire modeling. Comparative experimental findings further demonstrate how insulation aging modifies ignition thresholds, heat release rates, and smoke toxicity. By integrating perspectives from materials science, electrical engineering, and fire dynamics, this review establishes the nexus between aging mechanisms and fire hazards.

Abstract - This comprehensive review integrates recent evolution in the development of fire-resistant concrete and mortar materials, focussing the addition of inventive additives, fibers, and geopolymer technologies. The studies focuses the advanced fire resistance, mechanical performance, and durability of geopolymer concrete (GPC) in comparison to traditional Ordinary Portland Cement (OPC), with GPC revealing little cracking, spalling and better residual strength at higher temperatures. The integration of silica-fume, metakaolin, fly-ash and polypropylene fibers remarkably increases thermal insulation, decrease spalling, and enhance overall structural strength when exposed to fire . Hybrid systems integrating carbon nanotubes and polypropylene fibers indicate synergetic benefits adjusting fire safety with post fire structural stability. Furthermore advanced composite panels containing basalt fiber reinforced geopolymers and geopolymer coatings shows lightweight, high-temperature resistant solutions suitable for fireproofing and tunnel protection. The review emphasize the importance of performance-based, risk-informed design strategies, using advanced fire modelling and material innovations to enhance safety, sustainability, and resilience in fire prone structures. Overall, the addition of fibers, geopolymers, and composite materials shows a promising pathway toward safer, more durable, and environmentally sustainable fire-resistant construction materials. Keywords: Spalling, Metakaolin, Polypropylene,Fire Resistance

Fire modelling of a glove box for use in a two-zone fire model

Abstract As an important grassland ecological function area on the Tibetan Plateau, the risk of grassland fires in Qinghai Province has gradually increased due to climate warming and human activities. To quantitatively assess changes in grassland fire susceptibility under future climate scenarios, this study used historical fire data and CMIP6 model data, combined with multiple regression and the MaxEnt model, to simulate the distribution and trend of NDVI and fire susceptibility. Results showed that NDVI decreased under the low emission scenario (SSP119), and NDVI of grassland with medium and low coverage increased under medium and high emission scenarios (SSP245 and SSP585), while that of high coverage grassland decreased slightly. Fire susceptibility was higher in the east and south, and lower in the Qaidam Basin and northwest, with wind speed, distance from settlements, NDVI, slope, and human footprint as main driving factors. Mann‐Kendall and Theil‐Sen slope analyses showed that future fire susceptibility areas under medium‐ and high‐emission scenarios increased significantly and fluctuated, concentrating in the periphery of the Qaidam Basin and Southern Qinghai Plateau. Risk varied significantly among grasslands of different coverage. The study reveals the impact of global emission pathways on regional fire risk, emphasizing the need to strengthen adaptation, mitigation, and optimize grassland fire prevention to safeguard ecological security of the Qinghai‐Tibetan Plateau.

GreenFireNet: A real-time fire detection algorithm model for green hangars based on multi-modal fusion

Artificial intelligence (AI) offers several opportunities in wildfire management, particularly for improving short- and long-term fire occurrence forecasting, spread modeling, and decision-making. When properly adapted beyond research into real-world settings, AI can significantly reduce risks to human life, as well as ecological and economic damages. However, despite increasingly sophisticated research, the operational use of AI in wildfire contexts remains limited. In this article, we review the main domains of wildfire management where AI has been applied—susceptibility mapping, prediction, detection, simulation, and impact assessment—and highlight critical limitations that hinder practical adoption. These include challenges with dataset imbalance and accessibility, the inadequacy of commonly used metrics, the choice of prediction formats, and the computational costs of large-scale models, all of which reduce model trustworthiness and applicability. Beyond synthesizing existing work, our survey makes four explicit contributions: (1) we provide a reproducible taxonomy supported by detailed dataset tables, emphasizing both the reliability and shortcomings of frequently used data sources; (2) we propose evaluation guidance tailored to imbalanced and spatial tasks, stressing the importance of using accurate metrics and format; (3) we provide a complete state of the art, highlighting important issues and recommendations to enhance models’ performances and reliability from susceptibility to damage analysis; (4) we introduce a deployment checklist that considers cost, latency, required expertise, and integration with decision-support and optimization systems. By bridging the gap between laboratory-oriented models and real-world validation, our work advances prior reviews and aims to strengthen confidence in AI-driven wildfire management while guiding future research toward operational applicability.

Purpose To evaluate the influence of heating and cooling phases of fire on structural behaviour through cross-sectional analysis. The core objective is to develop parametric fire load density (FLDs) through field survey accounting role of modern materials and firefighting measures in evaluating the adequacy of fire-resistance rating (FRR) for reinforced concrete (RC) elements. The structural (axial and deflection) capacities indicative of performance are also assessed during both the heating and decay phases to quantify the resistance duration. Design/methodology/approach A field survey of 184 office rooms and 191 hostel rooms was conducted to determine realistic FLDs, including materials not addressed by code, occupancy type, floor area and ventilation, which influences fire dynamics. Utilising these data, Eurocode-based parametric fire curves were developed for thermo-mechanical analysis to examine the thermal response and subsequent degradation of RC components. Findings Flashover time has been reduced by more than 50% for the maximum FLD curve compared to standard curves. Firefighting measures reduce burning time to 80 min, causing 75–82% reduction in fire progression for the maximum FLD compartment, preventing structural degradation. The temperature field ratio of structural components revealed significant recovery of strength and stiffness in the cooling phase, which implicates structural capacity. Originality/value Highlights inadequacy of FRR by prescriptive procedures and advocates development of compartment-specific parametric fire models for building fire safety. Strongly emphasises performance-based assessment tailored to building topologies and their occupancy to assist in enhancing evacuation planning, fire safety contributing to resilience.

Abstract Wildfire risk and the extent of burned areas have been increasing in the Mediterranean region over recent decades, driven by high temperatures, low humidity, and extreme weather conditions intensified by climate change. Especially Greece, due to the country's diverse natural environment, has recently faced increasingly intense wildfires which have challenged firefighting efforts. The prediction of smoke dispersion can provide crucial information for enhancing fire management operations. In this study, an integrated Air Quality and Fire Modeling System based on WRF-SFIRE-Chem model has been developed to predict the wildfire spread, smoke dispersion and the distribution of PM10 and CO under specific weather conditions that significantly affect the fire by performing simulation scenarios on selected days—based on the Fire Weather Index (FWI) estimation—and specified fire ignition points. Two areas with specific interest were selected: i) the Samaria Gorge in Crete and ii) the cross-border forest area of Skopos in Western Macedonia. Results showed that the large-scale weather conditions and the local weather created by the fire, play a significant role in the pollutants’ dispersion and ground level concentrations. In the case of Samaria gorge, PM10 can reach up 11,238 μg/m3 at Ksiloskalo and 4629 μg/m3 at Agia Roumeli, and the ignition locations of the fire, even if in close distances, can result in different fire behavior and further to dissimilar smoke dispersion in the wider area. In the case of cross border area of Western Macedonia, PM10 can reach up to 2500μg/m3 at Skopos village and 3000μg/m3 at Agios Athanasios in the case of a wildfire event in the attached forest area. The developed fire and air quality modeling system, provides specialised understanding on the prevailing dispersion patterns of PM₁₀ and CO in the regions surrounding the areas of interest, for planning future fire risk management systems, designing efficiency cooperative prevention and evacuation strategies. Graphical Abstract

Effect of fire propagation modelling on structural elements temperature of steel racks