ID 118.

High-Confinement Explosion Analysis in a

Residential Building: A Comparative Study of TNO
Model Estimates and Real Gas Consumption Data

Vincenzo Puccia 1
1
Corpo Nazionale Vigili del Fuoco, Comando dei Vigili del Fuoco di Verona,
Verona, Italy

INTRODUCTION
On April 15, 2023, a significant explosion occurred in a residence complex in
the North East of Italy. The incident, attributed to a natural gas leak, resulted in
extensive structural damage and injuries. This study presents a detailed forensic
analysis of the explosion dynamics, with a focus on the impact of the confined
environment on pressure wave propagation and structural damage. The analysis
includes a comparison between the estimated overpressures using the TNO
multi-energy method and the actual gas consumption data provided by the
network operator.

Fig. 1. The effects of the high confinement natural gas explosion.

The investigation combined physical site assessment, computational modeling,
and quantitative data analysis. Structural damage, particularly the displacement
and destruction of interior walls and doors, was evaluated to estimate the peak
overpressure and its distribution. The TNO multi-energy method was employed
to estimate the overall mass of natural gas with regard to the explosion damages
under high-confinement conditions, comparing the literature pressure build-up.
On the other hand, gas consumption data from the operator indicated an
anomalous peak of 10.85 m³ on the day of the incident, compared to a typical
daily average of 1.81 m³. The quantitative analysis of the gas release, based on

11th International Seminar on Fire and Explosion Hazards, 15-20 June 2025, Rome, Italy.
Book of Abstracts.
247
the observed damage and estimated overpressures, was conducted to validate
these findings against the TNO model predictions.

RESULTS AND DISCUSSION

The results indicated that the explosion, originating from a gas leak in the
kitchen, occurred within a confined space, resulting in a highly overpressurized
environment. The overpressure, as calculated using the TNO method, reached up
to 0.3 bar in localized areas, correlating with the observed structural damage.
This included the expulsion of partition walls and the apartment door where the
premixed combustion was initiated, as well as the inward projection of doors in
the apartments above.
A detailed comparison with the actual gas consumption data revealed a strong
correlation between the estimated gas volume involved in the explosion
(approximately 8.85 m³) and the total recorded consumption spike. The high-
confinement nature of the explosion significantly amplified the damage, as
indicated by the failure of non-load-bearing elements and the partial
displacement of the roof structure.
The concentration vs. time analysis suggested that the gas mixture reached
explosive limits (5-15% by volume) within approximately 200 minutes of the
leak onset, assuming minimal ventilation. This slow build-up was likely
undetectable by the residents, contributing to a delayed but intense deflagration
upon ignition.
This case study highlights the critical role of high-confinement conditions in
amplifying the effects of gas explosions in residential settings. The alignment
between the TNO model estimates and real consumption data underscores the
importance of accurate predictive models in forensic investigations. The findings
suggest a need for enhanced safety measures, such gas detection systems, to
mitigate the risks of high-confinement explosions in similar environments.

REFERENCES
[1] R.J.Harris, The Investigation and Control of Gas Explosions in Buildings and Heating
Plants, E&F Spon, 1983.
[2] D.Bjerketvedt, J.R.Bakke, K.van Wingerden, Gas Explosion Handbook, Journal of
Hazardous Materials 52 (1997) 1-150.
[3] Z.Jiang H.Teng, Gaseous Detonation Physics and Its Universal Framework Theory,
Springer 2022.
[4] M.E.Weyler, Suncam Forensic Analysis involving fugitive Natural Gas And Propane,
Online Continuing Education Course 2014
https://www.suncam.com/miva/downloads/docs/199.pdf.
[5] M.A. Netileton, Gaseous Detonations: their nature, effects and control, Chapman and
Hall Ltd, London 1987.

11th International Seminar on Fire and Explosion Hazards, 15-20 June 2025, Rome, Italy.
Book of Abstracts.
248
 ID 126. CFD simulations of H2/CH4/H2S dispersion and
consequences for the risk analysis of large scale H2
storage

Domenico Enicchiaro 1, Maria Portarapillo 1, Franco Polidoro 2,

Almerinda Di Benedetto 1
1
University of Naples Federico II, Department of Chemical, Materials and Production
Engineering, Napoli, Italy
2
Ricerca sul Sistema Energetico - RSE S.p.A., Ricerca sul Sistema Energetico - RSE
S.p.A., Milan, Italy

INTRODUCTION
Salt caverns are considered as one of the most suitable options for large-scale,
high-pressure hydrogen storage due to their high energy density (ranging from
200 to 450 kWh/m³), significant storage capacity, and relatively low construction
costs. Their use is not without challenges as methanogenic microorganisms
present at the bottom of salt cavities can produce H2S, CO2 and CH4 by
consuming H2 in the presence of sulfates and carbonates. This bacterial activity
poses H2 purity and safety issues. From a quality perspective, hydrogen must
meet strict purity standards for many advanced applications. On the safety front,
microbial activity can generate hazardous (toxic) gases like H₂S and the possible
formation of combustion products such as SO2 in case of fires.
A comprehensive risk assessment of underground hydrogen storage is essential
to determine the feasibility and safety of large scale H2 storage.
In a previous work, Portarapillo and Di Benedetto (2021) [1] evaluated gas
dispersion from large-scale salt caverns into the atmosphere, applying empirical
models to assess the behaviour of gas dispersal, to quantify the effects of thermal
radiation in fires, and to evaluate the overpressure impacts in case of explosions.
However, empirical models have proven inadequate for hydrogen analysis due
to its high dispersion coefficients and positive buoyancy, highlighting the need
for new and more precise models [2]. A detailed model based on the momentum,
energy and specie mass balance equations (CFD) could overcome these issues,
offering better understanding of safety aspects, particularly related to H2/H2S
release and jet fires. While few CFD studies have addressed large-scale hydrogen
jet fires, none have focused on salt caverns. In this context, a comprehensive
methodology for risk analysis of large-scale underground hydrogen storage sites
has been performed through CFD simulations applying the Unsteady Reynolds-
Averaged Navier-Stokes approach. It has been simulated the dispersion of H2,
H2S, and CH4 mixtures as well as the effects of immediate ignition. The mixture
composition was varied to consider the effect of bacterial metabolism. The CFD
results were compared to the results obtained with the aid of empirical models
to assess the strengths, limitations, and potential of each approach. Simulations
were carried out starting from the rupture of the riser pipe connecting the salt

11th International Seminar on Fire and Explosion Hazards, 15-20 June 2025, Rome, Italy.
Book of Abstracts.
249
cavern to the ground as initiating event, considering the storage of 500,000 m3
of hydrogen at 80 °C and 160 bar.

RESULTS AND DISCUSSION

Results have shown that changes in the gas mole fractions (Fig. 1), driven by H2
consumption and production of H₂S and CH4, can significantly affect the risk
analysis. As hydrogen content decreases, the flammable region where ignition is
possible reduces. Consequently, the potential damage from an Unconfined Vapor
Cloud Explosion becomes negligible. The spatial maps of the specie
concentration obtained by CFD simulations differ significantly from those
obtained using empirical models [1]. The effect zones for various incident
outcomes calculated with empirical models result smaller than those obtained by
CFD simulation.

Fig. 6. 2D Contour plots of the mole fraction for different gas mixtures over
time. (A) 100% H2, (B) 25% H2, 70% CH4, 5% H2S, (C) 50% H2, 47% CH4,
3% H2S.
Regarding H₂S and SO₂, CFD simulations have shown that within the first
minute these gases have not reached the ground despite their negative buoyancy,
due to the initial momentum. Comparison of jet fire lengths calculated with
empirical models and CFD simulations revealed small differences. Significant
discrepancies were found in the values of the radiative flux calculated by means
of the solid flame model and the CFD simulations. More precisely, the solid
flame model overestimates the flux.

11th International Seminar on Fire and Explosion Hazards, 15-20 June 2025, Rome, Italy.
Book of Abstracts.
250
REFERENCES
[1] M. Portarapillo, A. Di Benedetto, Risk assessment of the large-scale hydrogen storage
in salt caverns, Energies (Basel) 14 (2021).
[2] M. Calabrese, M. Portarapillo, A. Di Nardo, V. Venezia, M. Turco, G. Luciani, A. Di
Benedetto, Hydrogen Safety Challenges: A Comprehensive Review on Production,
Storage, Transport, Utilization, and CFD-Based Consequence and Risk Assessment,
Energies (Basel) 17 (2024).

11th International Seminar on Fire and Explosion Hazards, 15-20 June 2025, Rome, Italy.
Book of Abstracts.
251
 ID 128. Risk assessment of Integrated Fuel Cell
Systems for Rail Transport

Maria Portarapillo 1, Augusto Bellucci Sessa 1, Almerinda Di

Benedetto 1
1
University of Naples Federico II, Department of Chemical, Materials and
Production Engineering, NAPOLI, Italy

INTRODUCTION
Due to the depletion of fossil fuels and the deterioration of environmental
conditions, global attention is increasingly focusing on the use of renewable
energy sources and the improvement of energy efficiency. The use of hydrogen-
based technologies requires infrastructures for production, transport, storage and
distribution on a small, medium and large scale. The main focus is on optimising
technical performance. However, when developing hydrogen infrastructure,
safety must be considered first and foremost. In this context, the proton exchange
membrane fuel cell (PEMFC), which efficiently converts hydrogen into
electricity, is becoming increasingly important, especially in mobile systems
(e.g. vehicles). In particular, the high temperature PEMFC (HT-PEMFC) is very
promising thanks to the operating temperatures (120-200 °C) and the resulting
CO tolerance (4-5 % at 200 °C). The higher operating temperature leads to
increase of reaction kinetics, catalysts tolerance, heat dissipation and water
management eventually obtaining higher energy efficiency. The use of hydrogen
fo fuel cell requires safe and efficient systems for hydrogen storage.
Alternatively, the in-situ production of hydrogen from "safer" chemicals can be
explored and compared. The aim of this work is to investigate the feasibility of
an integrated railway traction system in terms of volume and weight based on
hydrogen storage in the form of liquid methanol, ethanol, glycerol or ammonia,
hydrogen production by oxidative steam reforming (OSR) or decomposition and
HT-PEMFC. Design of configurations energy self-sufficiency was performed
and the risk assessment of all the systems was carried out. Results were
compared with those obtained for a configuration with on-board hydrogen
storage at 350 bar.

RESULTS AND DISCUSSION

The energy self-sufficiency assessment of the integrated system was carried out
using AspenPlus®. The configuration consists on the stage to preheat the
reactants, the reaction unit and the unit for cooling the outlet stream to reduce
the temperature to a level acceptable for HT-PEM. The heat released by the fuel
cell was compared with the heat required in the heating phase. The heat released
in the the cooling step was taken into account to achieve the autothermal
operating conditions. . The volumes of the alcohol/water or ammonia tanks were
calculted by taking into account the hydrogen production yields given in the
literature ([1]). The electrical system, consisting of high-temperature PEM fuel

11th International Seminar on Fire and Explosion Hazards, 15-20 June 2025, Rome, Italy.
Book of Abstracts.
252
cells and batteries, was dimensioned by simulating the Brescia-Edolo line (Italy)
and evaluating the required electrical power, motor traction and travelling speed
using conventional train kinematics equations [2]. Results in terms of weight of
stored material, storage tank and traction system are reported in Table 1In the
case of glycerol, the optimum operating temperature of the reactor is 600 °C.
Under these conditions, a system weight of around 22 tonnes was determined. In
the case of ammonia, liquid storage at room temperature and 9 bar as well as a
reactor operating temperature of 300 °C are required. Under these conditions,
the system has a weight of 14 tonnes. All volumes are compatible with the size
of a railway wagon and the weights are below those permitted on railway lines
such as the Brescia-Edolo line.
Table 4 - Weight of stored material, storage tank and traction system.
Mass of stored Mass of storage Total weight of the
System
material (kg) tank/s (kg) traction system (kg)
Hydrogen 438 10400 20000
Methanol 3582 467 14000
Ethanol 6749 860 17000
Glycerol 12114 1145 22000
Ammonia 2147 2000 14000
In terms of risk assessment, different outcomes may occur depending on the
system studied (Table 2). In the case of ethanol, the only possible outcome in the
case of hole formation in the storage tank is the formation of a pool fire in the
rail car. Due to the mitigating effect of water, the formation of vapor cloud
explosions is not possible. In the case of methanol and ammonia, the most
serious effects are related to their toxicity. For ammonia in particular,
concentration levels at ERPG-1 are found up to 6 km away, values much higher
than for methanol, which is diluted with water even over long distances. Both in
terms of probability of immediate and delayed ignition and in terms of severity,
the most critical system is the reference system based on hydrogen storage on
the rail roof. Results show that 50% probability of fatalities related to thermal
radiation heat flux from jet fire and to overpressure from VCE can be found at
distances of 43 m and 10 m, respectively.
Table 2 - Outcomes description in the case of hole generation on the storage
tank.
System Outcomes
Jet fire, Vapor cloud explosion, Flash
Hydrogen
fire
Methanol Pool fire, Toxic dispersion
Ethanol Pool fire

11th International Seminar on Fire and Explosion Hazards, 15-20 June 2025, Rome, Italy.
Book of Abstracts.
253
 Glycerol -
Ammonia Pool fire, Jet fire, Toxic dispersion

REFERENCES
[1] S.J. Kim, T.S. Nguyen, J. Mahmood, C.T. Yavuz, Sintering-free catalytic ammonia
cracking by vertically standing 2D porous framework supported Ru nanocatalysts, Chem.
Eng. J. 463 (2023) 142474.
[2] Joshua Peli, APPLICAZIONE DELLE FUEL CELL PER L’ALIMENTAZIONE DI
VEICOLI FERROVIARI, Politecnico Milano, 2018.

This study was carried out within the MOST – Sustainable Mobility National
Research Center and received funding from the European Union Next-
GenerationEU (PIANO NAZIONALE DI RIPRESA E RESILIENZA (PNRR) –
MISSIONE 4 COMPONENTE 2, INVESTIMENTO 1.4 – D.D. 1033
17/06/2022, CN00000023). This manuscript reflects only the authors’ views and
opinions, neither the European Union nor the European Commission can be
considered responsible for them.

11th International Seminar on Fire and Explosion Hazards, 15-20 June 2025, Rome, Italy.
Book of Abstracts.
254
 ID 142. Investigating the Thermal Hazard to the
Immediate Surroundings of a Burning Electric Vehicle

Parham Dehghani 1, Matthew DiDomizio 1, Nathaniel Sauer 1,

Adam Barowy 1
1
Fire Safety Research Institute, ULRI, COLUMBIA, United States

INTRODUCTION
The advent of electric vehicles (EVs) has marked a significant advancement in
automotive technology to reduce carbon emissions and decrease dependence on
fossil fuels. However, this transition also introduces new challenges, notably the
risk of fires associated with their lithium-ion batteries [1], and whether the
involvement of Li-ion batteries in a vehicle fire substantially impacts thermal
hazards and fire spread to the vehicle’s surroundings. This study investigates the
thermal hazards posed by EV fires, focusing on the battery pack and passenger
cabin. It examines the timing, location, and magnitude of their contributions to
the overall thermal exposure of the immediate surroundings.
The vehicles examined in this study [2] included a Tesla Model 3 and a Nissan
Leaf. Steel panels were positioned on each side of the car, set at approximately
1 m from the vehicle sides and centered on the vehicle’s length. The panels were
utilized to measure the spatially and temporally varying heat flux with high
resolution to the immediate surroundings of the burning EV. The 22 ga stainless
steel panels were 1.5 m tall and spanned the length of the vehicle. The panels
were coated with high-emissivity (0.94) black paint [3] and were insulated at
their peripheral edges to create an adiabatic boundary. Four Schmidt-Boelter
water-cooled combination (total, and radiometer) heat flux gauges were installed
at the mid-height of the walls, distributed evenly along their lengths. Heat flux
data were recorded at 2 Hz, and the temperature of the steel panels was measured
at 30 Hz using infrared cameras positioned on the un-exposed sides. An energy
conservation-based Inverse Heat Transfer (IHT) model [4] was used to calculate
the heat exchange between the EV fire and the panels. A new open-source
software, HFITS [5], was developed to calculate gauge and incident radiative
heat flux from the measured thermograms. The discrete total heat flux gauge
measurements were utilized to validate the two-dimensional heat flux field
measurements at nearby locations. The heat flux field measurements were then
used to evaluate the thermal hazard from the EV fires to their surroundings.

RESULTS AND DISCUSSION

The transient two-dimensional heat flux fields measured in this study provided
meaningful insight into the dynamics of an EV fire and the associated thermal
hazards. It was demonstrated that the cabin fire was the dominant source of
thermal exposure to the sides of the vehicles when both the cabin and battery
were burning, obfuscating any thermal hazard specifically attributable to battery
jets unless jetting occurred in the early stages of the fire (before spreading to the

11th International Seminar on Fire and Explosion Hazards, 15-20 June 2025, Rome, Italy.
Book of Abstracts.
255
cabin), or in an otherwise isolated area. Fig. 1 shows the Tesla Model 3 Long
Range fire along with the incident radiative heat flux near the rear driver's side
of the vehicle. Flame jetting was observed near the tire, and the associated
thermal hazard is apparent in the measured heat flux field.
The gauge heat flux calculated with the IHT model showed agreement with
measurements from the Schmidt-Boelter radiometer heat flux gauges. The
accuracy was quantified as the normalized Root Mean Squared Error (NRMSE)
between the measured and IHT-inferred heat flux values at nearby locations, with
an overall NRMSE of 7.6% across both vehicles and all eight gauges. The
findings of this study address the unique thermal hazards of EV fires, including
the spatial and temporal aspects of thermal exposure to the immediate
surroundings, as well as the identification of thermal runaway events in terms of
their location, timing, and contribution to thermal exposure. In EV fires, once
the passenger cabin catches fire, it becomes the dominant heat source.
This work not only advances the understanding of EV fire hazards but also
demonstrates the application of a novel measurement technique that provides
high-resolution (centimeter scale) 2D field measurements of heat flux,
something impractical (if not impossible) with traditional heat flux gauges.
Obtaining this type of high-resolution data allows researchers to capture the
spatial variation of thermal exposures from a fire. This technique has potential
applications across various fields of fire safety science.

Fig. 1. Tesla Model 3 Long Range thermal runaway jetting (Left), 2-D incident
radiative heat flux obtained by HFITS [5] (IHT)(Right).

REFERENCES
[1] S. Peiyi, R. Bisschop, H. Niu, and X. Huang. A review of battery fires in electric
vehicles. Fire technology 56, no. 4 (2020) 1361-1410.
[2] A. Barowy, Fire Safety of Batteries and Electric Vehicles, Fire Safety Research
Institute, https://fsri.org/research/fire-safety-batteries-and-electric-vehicles (2024).
[3] G. T. Bellamy, M. J. DiDomizio, M. K. Patel, M. B. McKinnon, Characterization of
high-temperature paints for infrared thermography in fire research, Fire Safety Journal
137 (2023).

11th International Seminar on Fire and Explosion Hazards, 15-20 June 2025, Rome, Italy.
Book of Abstracts.
256
[4] O. M. Alifanov, Inverse Heat Transfer Problems, Springer Berlin, Heidelberg,
Germany, 2012.
[5] P. Dehghani, M. J. DiDomizio, HFITS: Heat Flux Measurements using Infrared
Thermography and a Plate Sensor, Github. https://github.com/ulfsri/HFITS (2024).

11th International Seminar on Fire and Explosion Hazards, 15-20 June 2025, Rome, Italy.
Book of Abstracts.
257
 ID 143. On Measurements, and Modeling of Diffusion
Flames Temperatures Aboard the International Space
Station

Parham Dehghani 1
1
Fire Safety Research Institute, ULRI, COLUMBIA, United States

INTRODUCTION
Human explorations in the universe bring new uncertainties and hence new risks
to the human health and man-made structures such as spacecrafts. One of these
risks is caused by fire mainly in the form of pool fire diffusion flames caused by
condensed phase materials. Fire hazards in microgravity have not been
sufficiently studied to ensure the safety of astronauts and spacecrafts. Early
studies on candle flames showed that microgravity diffusion flames can burn for
long durations [1]. Research by Fernandez Pello et al. focused on flame spread
behavior across varying flow conditions and gravity levels [2]. Several studies
by Quintiere et al. have been performed on material burning in-place (i.e., before
spreading) in microgravity simulation of condensed phase materials burning
using a gas-supplied burner (the Burning Rate Emulator (BRE)) and by matching
material properties (heat of gasification (L), heat of combustion (∆ℎ' ), laminar
SS
smoke point (ls), and surface re-radiation (𝑞̇ 77 ) with those deducted from the
burner flames. The supplied gas generally consisted of a Hydrocarbon fuel and
a neutral component (Nitrogen), and the flame properties were deducted for the
mixture and not just the Hydrocarbon. It was observed that microgravity flames
did not yield soot, and the burner surface temperature could not be controlled so
the simulation material properties were limited to heat of gasification and heat
of combustion. A general theory was developed for microgravity condensed
phase material flames that led to flammability maps as a function of material
properties (∆ℎ' , and L) and ambient conditions (XO2, and p) [3]. Sunderland et.
al. utilized Silicon-Carbide thin filaments to measure the hot gas temperature
inside microgravity non-sooty flames. As the thin filaments were inserted into
the hot gas, they started glowing and the temperatures were deducted from the
recorded images [4]. As of today, no study has been focused on microgravity
temperature measurements resulting from condensed phase materials to the best
of the author’s knowledge. In this study temperature measurements of two
flames were examined from ignition until the test endpoint. The experimental
apparatus consisted of the BRE burner [5], two non-water-cooled Medtherm heat
flux gauges flushed with the burner surface. The remote radiometers measured
flame radiative emission which led to the modelling of flame radiative loss
fraction (Xr), crucial for the flame temperature modelling. The experiments took
place inside a chamber with 100 L of free volume filled with a synthetic
atmosphere consisting of a pre-defined species mixture. The detail of the
experimental apparatus is discussed in ref. [5]. The results were compared
against the Chemical Equilibrium with Applications (CEA) model developed by
NASA which is used to calculate chemical equilibrium compositions [6].

11th International Seminar on Fire and Explosion Hazards, 15-20 June 2025, Rome, Italy.
Book of Abstracts.
258
RESULTS AND DISCUSSION
The flame images from 3 s after ignition and before extinction with the filaments
inserted in them are shown in Fig. 1. The mass flux and fuel-mixture composition
were kept constant, and the main variable was the ambient condition. It was
shown in refs. [3,5] that pressure has a negligible effect on the flammability of
materials in microgravity, and therefore the main variable here was the ambient
XO2. The extinction time was reduced from 74.34 s to 32.03 s as a result of a
3% reduction in XO2. Both tests were ignited having flame temperatures above
1200 K with the first one having a higher temperature as a result of more intense
burning in higher concentration of Oxygen in the ambient and as expected the
lower XO2 led to lower flame temperatures that reached the extinction
temperature threshold significantly faster. The main objective of this study is to
obtain the temperature of extinction onset for these flames which was found to
be 1168 K. This value was close to the value modeled by Snegirev et. Al. [7] in
their 3D CFD study of BRE flames, and the theoretical predictions of Dehghani,
Quintiere et. Al. [8]. This temperature represents the gas temperature of the onset
of flame oscillations that led to extinction. The second objective was to examine
the accuracy of the predictions made by an available-to-public model (CEA). It
was shown that the discrepancy between the measured and CEA modeled values
were 17%, and 33% of the measured values. This discrepancy suggests that
further analysis is required to establish a simplified theoretical model to predict
flame temperatures of condensed phase fuels simulated flames in microgravity.

t=3s t = 17 s t = 65 s

Fig. 1. Image sequence of a microgravity flame with inserted thin filaments.

REFERENCES
[1] D.L. DIETRICH, H.D. ROSS, Y. SHU, P. CHANG, J.S. T’IEN, Candle Flames in
Non-Buoyant Atmospheres, Combust. Sci. Technol. 156 (2000) 1–24.
[2] McAllister, S, et al. "Modeling microgravity and normal gravity opposed flame spread
over polymer/glass composites." 45th AIA. 2007.
[3] Dehghani, P., de Ris J., and Quintiere, J.. Demonstrating steady burning for small flat
materials in microgravity in a quiescent ambient. Proc. Comb. Inst.(2023): 3949-3958.
[4] Maun, Jignesh D., Peter B. Sunderland, and David L. Urban. "Thin-filament
pyrometry with a digital still camera." Applied optics 46.4 (2007): 483-488.
[5] Dehghani, Parham, et al. "Burning in microgravity: Experimental results and
analysis." Combustion and Flame 228 (2021): 315-330.
[6]Chemical Equilibrium with Applications. https://www1.grc.nasa.gov/research-and-
engineering/ceaweb (accessed 24 February 2024).

11th International Seminar on Fire and Explosion Hazards, 15-20 June 2025, Rome, Italy.
Book of Abstracts.
259
[7] Snegirev, A., et al. Transient dynamics of radiative extinction in low-momentum
microgravity diffusion flames. Proc. Comb. Inst. 38.3 (2021): 4815-4823.
[8] Dehghani, P., and Quintiere J., Theoretical analysis and predictions of burning in
microgravity using a burning emulator. Combustion and Flame 233 (2021): 111572.

11th International Seminar on Fire and Explosion Hazards, 15-20 June 2025, Rome, Italy.
Book of Abstracts.
260
 ID 148. Hydrogen release and dispersion in a
underground car park under natural ventilation

Paola Russo 1, Maria Grazia Meo 1, Luca Nassi 2

1
Sapienza University of Rome, Department of Chemical Engineering Materials
Environment, Rome, Italy
2
Italian National Fire Department, Siena Fire Command, Siena, Italy

INTRODUCTION
As a result of increasing environmental concerns, the reduction of emissions in
the transport sector is currently being promoted worldwide through the
promotion of alternative fuel vehicles. In particular, the possibility of using
hydrogen as a zero-carbon emission fuel is of great interest. On the other hand,
the introduction of hydrogen-powered vehicles needs to be carefully assessed
from a safety point of view in comparison with conventionally fuelled vehicles.
The risk of an accidental release needs to be considered. A release could result
from the activation of the safety thermal pressure relief device (TPRD) or from
component failure. In fact, while a release of hydrogen from a vehicle outdoors
can be considered almost harmless due to its rapid dispersion, a release of
hydrogen in a confined space, such as from a vehicle in a car park, can lead to a
potentially dangerous situation. Due to the wide range of flammability of
hydrogen in air (4-75% v/v), accidental releases are likely to ignite and become
jet flames or even explode. Current regulations in each country provide safety
guidelines that have been developed and updated mainly for common
conventional vehicle scenarios in confined situations. Ventilation systems in car
parks are also specified by national standard codes.
In this study, a CFD was used to investigate the unignited release and dispersion
of hydrogen from a car in an underground car park with different ventilation
conditions. In accordance with the current guidelines, the car park was equipped
with vents for natural ventilation.
RESULTS AND DISCUSSION
The TPRD is located under the car and, when activated, provides a rapid release
of hydrogen (700 bar). In this study, a downward release direction was assumed
with a release angle of 45° to the rear through a 2 mm diameter TPRD. The
adiabatic blowdown model of storage tank dynamics [1] was used to calculate
the blowdown of hydrogen from the high-pressure storage tank through the
TPRD, as it estimates the under-expanded jet formed during release and the rapid
expansion to atmospheric pressure and allows the calculation of the mass flow
rate during release through an effective notional nozzle at atmospheric pressure.
The volumetric source method was then used to calculate all the conditions at
the effective jet exit, and finally set up for CFD simulations for all the scenarios
evaluated. The simulated underground car park has a surface area of 737 m2 with
air vents that can be set from a minimum of 1.8 m2 to 56 m2.

11th International Seminar on Fire and Explosion Hazards, 15-20 June 2025, Rome, Italy.
Book of Abstracts.
261
The simulations performed showed that the dispersion of the flammable
hydrogen-air mixture depends on the car position in the park and the ventilation
conditions. In the worst scenarios studied, natural ventilation was shown to be
unable to control and reduce hydrogen dispersion, even when air vent areas were
assumed to be larger than recommended from current national regulation.
Current guidelines and standards for car parks should be appropriately revised
to meet the requirements needed to reduce risk also for hydrogen-powered
vehicles, including mechanical forced ventilation systems and specific
instructions on parking position for hydrogen cars.

Fig. 1. Dispersion of the flammable volume with a smaller vent area (1.8 m2).

Fig. 2. Dispersion of the flammable volume with a larger vent area (56 m2).

REFERENCES
[1] Hussein, H., Brennan, S., & Molkov, V. (2020). Dispersion of hydrogen release in a
naturally ventilated covered car park. International Journal of Hydrogen Energy, 45(43),
23882-23897. https://doi.org/10.1016/j.ijhydene.2020.06.

11th International Seminar on Fire and Explosion Hazards, 15-20 June 2025, Rome, Italy.
Book of Abstracts.
262
 ID 156. CFD-based Risk Analysis in Natech Scenarios
for Hythane distribution infrastructure

Elena Capasso 1, Maria Portarapillo 1, Almerinda Di Benedetto 1

1
University of Naples "Federico II", Department of Chemical, Materials and
Production Engineering, Naples, Italy

INTRODUCTION
A technological accident triggered by a natural disaster is known as Natech
(natural hazard triggering technological disasters) and is a topic of great concern
due to the increasing intensity of weather-related phenomena, mainly caused by
climate change. The Multi-Risk sciEnce for resilienT commUnities undeR a
changiNg climate (RETURN) project aims to improve the entire disaster risk
management cycle. The harmonization of energy production and climate
protection relies on hydrogen as a promising energy source. However, one of the
main barriers to its use is the lack of an adequate distribution infrastructure. One
idea is the use of natural gas pipelines. But replacing natural gas with hydrogen
is currently out of the question, mainly for safety reasons due to hydrogen
properties, but also because of the economic investment required for the
necessary changes in the existing pipelines. For these reasons, a potential interim
solution to reduce CO2 emissions could be the implementation of hythane, a
hydrogen-enriched methane mixture. Considering the impact of natural
phenomena and the challenges of integrating hydrogen into the existing natural
gas infrastructure, this study focuses on a CFD-based risk assessment of the
current gas distribution infrastructure. The novelty of this work is the use of CFD
simulations to overcome limitations of empirical modeling to allow an improved
representation of flow evolution and of large-scale turbulence dispersion of gas
mixtures. Release and dispersion of mixtures with different CH4/H2 ratios (0 to
1), using a specific framework for Natech scenarios were performed [1].
Earthquakes, floods and lightning strikes were evaluated as natural events and
Loss of containment (LOC) triggers. Literature vulnerability models were used
to assess the frequency of pipeline damage associated with natural events [2].
After evaluating the frequency of occurrence of LOC, an event tree analysis
(ETA) was developed to determine the frequency of occurrence of the different
outcomes. The consequence analysis was performed by developing a CFD model
based on the Unsteady Reynolds Averaged Navier-Stokes (URANS) approach
coupled with a standard k-ε turbulence model to simulate the dispersion of
released gases. In the seismic event, the ETA results in atmospheric dispersion,
unconfined vapor cloud explosion (UVCE), and jet fire.

RESULTS AND DISCUSSION

CFD results show that when increasing the hydrogen content in the
methane/hydrogen mixtures, the plume changes both qualitatively (plume shape)
and quantitatively (plume length and height) and the combustible mass of the

11th International Seminar on Fire and Explosion Hazards, 15-20 June 2025, Rome, Italy.
Book of Abstracts.
263
plume increases. Moreover, the comparison between hythane/hydrogen and pure
methane releases showed a significant increase in UVCE zones as the hydrogen
content increased.

Fig. 1. Molar fraction profile of the gas released in the plane z-x at y=50m.
Jet fire impacts were calculated using empirical models that relate the radiative
heat flux received at different distances from the point source for different
CH4/H2 ratios. Results show that the size of the isorisks zone decreases as the
mole fraction of hydrogen in the hythane increases. The risk analysis results
show that the addition of hydrogen affect the individual risk values, in
particularly the individual risk of 1 min-long jet fire shifts from 10-6 to 10-5
years-1, while for UVCE and 5 min-long jet fire from 10-8 to 10-7 years-1,
demonstrating an increase of the risk related to the addition of hydrogen to actual
infrastructures. The knowledge gained will contribute to the development of
proactive strategies to improve the pipelines safety and resilience to unforeseen
natural disasters, with a view to implementing hydrogen as a future energy
carrier.

REFERENCES
[1] V. Cozzani, G. Antonioni, G. Landucci, A. Tugnoli, S. Bonvicini, G. Spadoni,
Quantitative assessment of domino and NaTech scenarios in complex industrial areas, J.
Loss Prev. Process Ind. 28 (2014) 10–22. https://doi.org/10.1016/j.jlp.2013.07.009.
[2] G. Lanzano, E. Salzano, F.S. de Magistris, G. Fabbrocino, Seismic vulnerability of
natural gas pipelines, Reliab. Eng. Syst. Saf. 117 (2013) 73–80.
https://doi.org/10.1016/j.ress.2013.03.019.
PIANO NAZIONALE DI RIPRESA E RESILIENZA (PNRR) MISSIONE 4 –
COMPONENETE 2 – INVESTIMENTO 1.3 – TITOLO DELL’INVESTIMENTO:
Creazione di “Partenariati estesi alle università, ai centri di ricerca, alle aziende per il
finanziamento di progetti di ricerca di base” - D.D. n. 341 del 15 marzo 2022 –
PE_00000005 RETURN “Multi-risk science for resilient communities under a changing
climate”.

11th International Seminar on Fire and Explosion Hazards, 15-20 June 2025, Rome, Italy.
Book of Abstracts.
264
 ID 157. Experimental study on the toxicity of different
halogenated olefins to mice

Zong Ruowen1,2, *, Yao Fuyao 1, Wang Kaitao 1

1
State Key Laboratory of Fire Science, University of Science and Technology
of China, Hefei, Anhui, China
2
Suzhou Key Laboratory of Urban Public Safety, Suzhou Institute for
Advanced Research, University of Science and Technology of China, Suzhou,
Jiangsu, China
*
Corresponding author email:[email protected]

INTRODUCTION
The application of halogenated olefins in firefighting has increased but its
biological toxicity particularly the acute toxicity has not been well studied [1].
This study aims to assess the acute toxicity of halogenated olefins in mice and
compare the impact of halogen substituents on acute toxicity. In this study, the
acute toxicity evaluation was conducted by HE staining and physiological
changes of mice. The extent of lung damage was assessed to analyze the toxic
effects of different halogenated olefins based on the histopathological changes
observed in the lung tissue.

RESULTS AND DISCUSSION

In this study, four halogenated olefins were tested for acute toxicity in mice. The
halogenated olefins were mixed with air to control its release concentration.
Mice were exposed for 20 mins in each test. Mice that died after exposure were
dissected and the lungs were stained with hematoxylin-eosin (HE). For mice that
survived, behavioral observations and physiological changes were monitored for
72 hours [2].

Fig. 1. Results of HE-staining of the lungs from mice in four toxicity tests.
As shown in Figs.1, lungs of mice were extracted and HE staining was carried
out after the toxicity test. The four selected halogenated olefins in the test all
caused damage to the lungs of mice, but the extent of lung damage they caused
varied. Of the four compounds, 2-BTP caused the most severe damage to the

11th International Seminar on Fire and Explosion Hazards, 15-20 June 2025, Rome, Italy.
Book of Abstracts.
265
lungs of mice, while HFO-1234yf and HFO-1336mzz(Z) caused the least.
According to the staining results, 2-BTP and HCFO-1233xf both induced
inflammatory cell infiltration and multiple hemorrhagic foci within the
interstitial tissue, caused alveolar septal thickening and compensatory
pulmonary emphysema [3].
#CF! #CF!
#*F*
!#>,9?!!!! 0@AB>"BBCWCCTD80 #*F*
!#>,9?!!!! 0@AB>"BBCWCCTD80
#*F!
0@EAB>"#BBFd!!!! 0eP< #*F!
0@EAB>"#BBFd!!!! 0eP<
#AF* #AF*
0@AB>"#BA/d 0@AB>"#BA/d
#AF! #AF!
#BF* #BF*
,-./01OP456T48

#BF!

,-./01OP456T48
#BF!
##F* ##F*
##F! ##F!
#"F* #"F*
#"F! #"F!
#!F* #!F*
#!F! #!F!
"HF* "HF*
"HF! "HF!
"'F* "'F*
"'F! "'F!
! "# #A BC A' C! (# ! "# #A BC A' C! (#
9PWOT5-;<Z8 9PWOT5-;<Z8

Fig. 2. Diagram of body weight changes in mice after toxicity tests

Figs.2.shows the body weight changes of mice following exposure to the four
tested halogenated olefins (males on the left, females on the right). 2-BTP and
HCFO-1233xf caused a reduction of body weight in mice. This experiment
revealed that mice exhibited severe loss of appetite after inhaling 2-BTP and
HCFO-1233xf. Mice that exposed to 2-BTP all died within the first 24 hours.
HFO-1234yf and HFO-1336mzz(Z) had little effect on the appetite of the mice,
but the body weight of mice increased more slowly than that of the control group.
Mice show many physical reactions after toxicity testing.In addition to dyspnea,
2-BTP and HCFO-1233xf also caused ophthalmopathy, hypersomnia, and belly
swollen in mice.
Brominated olefins exhibited higher acute toxicity than chlorinated olefins, and
fluorinated olefins showed the lowest acute toxicity. The three types of
halogenated olefins all cause lung damage in mice and redult in dyspnea. And
they all led to a decreased rate of body weight gain by suppressing appetite. The
acute toxicity of halogenated olefinfire extinguishing agents warrants serious
attention.

REFERENCES
[1] Zhou S, Yang Q, Zhang H, Zhou X. Theoretical and experimental insights into the
effects of halogen composition on the thermal decomposition details, as well as the fire-
suppressing mechanism and performance of CF3CX[double bond, length as m-dash]CH2
(X = F, Cl, Br). Phys Chem Chem Phys. 2021;23(19):11411-11423.
[2] Wang, T., Kuang Qin, H. Wang, and R.-M. Pan. Toxicology Evaluation of Thermal
Decomposition Products of Pentafluoroethane (HFC-125) with n-Heptane/Air Pool Fire.
Combustion Science and Technology.2017; 189 (12): 2083–94
[3] Aboubakr M, Elbadawy M, Ibrahim S S. Allicin and Lycopene Possesses a Protective
Effect against Methotrexate-induced Testicular Toxicity in Rats[J]. Pakistan Veterinary
Journal, 2023, 43(3).

11th International Seminar on Fire and Explosion Hazards, 15-20 June 2025, Rome, Italy.
Book of Abstracts.
266
 ID 158. Quantitative Analysis of Flames Generated by
Li-ion Battery Thermal Runaway

Sponem L.1, Bengaouer A.1, Dubourg S. 2, Kawka S.1, Koudriakov

S. 3, Reytier M. 1, *
1
Université Grenoble Alpes, CEA, LITEN, DEHT, F-38000 Grenoble, France
2
CEA, DAM, Le Ripault, F-37260 Monts, France
3
Université Paris-Saclay, CEA, ISAS, F-91191 Gif-sur-Yvette, France
*Corresponding author email : [email protected]

INTRODUCTION
The growing demand for electrical energy and autonomy in transportation has
made Li-ion batteries ideal for electrical vehicles, due to their high energy
densities. However, increasing the energy density can lead to severe safety
issues. Thermal runaway (TR) occurs when a cell generates exothermically more
energy than it can dissipate, resulting in an uncontroled temperature increased
and the ejection of hot combustible gases and particles. In a battery module, TR
can cause the casing rutpure, releasing gases into the surroundings, and may also
propagate to adjacent cells and modules.
It is crucial to characterize ejection during TR as well as the resulting combustion
flame, to predict and mitigate the risks. Gas ejection velocities are typically
measured via canister experiments [1], while flame length is commonly
measured optically [2, 3] and can be used to assess impingement risk. These
aspects are well documented, yet few studies address heat transfer from the Li-
ion cell to surfaces [4]. Furthermore, while flame extension is well correlated to
ejection parameters for typical gases [5, 6], allowing for simple prediction
models, direct correlation from experimental data are limited for Li-ion cells [3].
This study intends to link measured ejection parameters to flame extension
length, provide new data on radiative heat flux from TR flames and to assess the
applicability of the existing semi-empirical correlation to Li-ion flames [5].

RESULTS AND DISCUSSION

The thermal runaway (TR) of 53 A.h NMC Li-ion batteries at 100% state of
charge, triggered by overtemperature, is studied in two configurations. The cells
are first tested in a calorimeter under vacuum, with pressure and temperature
measurements used to compute the gas flow rate. Then open-air experiments are
conducted, where temperature and radiative heat flux are measured along the
flame axis. High-speed and HD cameras are used to measure the flame
morphology and the particles ejection velocities through tracking.
Velocity measurements from both the optical and calorimeter methods yielded
consistent results (figure1). The mean velocity is the highest at the onset of TR,
then stabilized to a smaller value. TR images reveal two distinct flame regimes:
an unstable, lifted jet with ignitions and extinctions followed by the formation of

11th International Seminar on Fire and Explosion Hazards, 15-20 June 2025, Rome, Italy.
Book of Abstracts.
267
a stable flame attached to the cell vent. In this stable regime, measured ejection
velocities, temperatures and flame length are compared with Delichatsios’
correlation for diffusion flame without particles, showing good agreement with
predicted values from both velocity measurement methods.

Fig. 1. Velocity measurement methods comparison.

Radiative heat transfer data indicates that the flames are not highly radiative,
with measured radiative heat flux at 50 cm from the flame below 4 kW/m2 and
an estimated radiative fraction ranging from 0.11 to 0.15.
Additionally, current developments focused on characterizing more realistic
configurations, where jets interact with structures and air availability is limited,
will be discussed.

REFERENCES
[1] A. W. Golubkov, D. Fuchs, J. Wagner, H. Wiltsche, C. Stangl, G. Fauler, G. Voitic, A.
Thaler, V. Hacker, Thermal-Runaway Experiments on Consumer Li-Ion Batteries with
Metal-Oxide and Olivin-Type Cathodes , RSC Advances 4(7) (2013) 3633‑42.
[2] H. Chen, J. E. H. Buston, J. Gill, D. Howard, R. C. E. Williams, C. M. Rao Vendra,
A. Shelke, J. X. Wen, An experimental study on thermal runaway characteristics of
lithium-ion batteries with high specific energy and prediction of heat release rate, Journal
of Power Sources 472 (2020) 228585.
[3] B. Mao, C. Zhao, H. Chen, Q. Wang, J. Sun, Experimental and Modeling Analysis of
Jet Flow and Fire Dynamics of 18650-Type Lithium-Ion Battery, Applied Energy 281
(2021) 116054.
[4] Y. Zhang, D. Kong, P. Ping, H. Zhao, X. Dai, X. Chen, Effect of a Plate Obstacle on
Fire Behavior of 18650 Lithium Ion Battery: An Experimental Study, Journal of Energy
Storage 54 (2022) 105283.
[5] M. A. Delichatsios, Transition from momentum to buoyancy-controlled turbulent jet
diffusion flames and flame height relationships, Combustion and Flame 92(4) (1993)
349‑64.
[6] R. W. Schefer, W. G. Houf, T. C. Williams, B. Bourne, J. Colton. 2007.
Characterization of high-pressure, underexpanded hydrogen-jet flames, International
Journal of Hydrogen Energy 32(12) (2007) 2081‑93.

11th International Seminar on Fire and Explosion Hazards, 15-20 June 2025, Rome, Italy.
Book of Abstracts.
268
ID 159. A simplified methodology to enhance efficiency
in numerical simulation of thermal runaway in Li-ion
batteries for safety applications

Yhuel E.1, *, Bengaouer A.1, Kawka S.1

1
Université Grenoble Alpes, CEA, LITEN, DEHT, F-38000 Grenoble, France
*Corresponding author email: [email protected]

INTRODUCTION
The proliferation of electric vehicles (EV) in the transportation sector is driven
by the prevalence of lithium-ion batteries, particularly NMC and LFP. The rise
of EV promise substantial environmental benefits, including reduced carbon
emissions and decreased reliance on fossil fuels. However, to capture a larger
automotive market share, it is essential that the EV must match or surpass the
performance of internal combustion engine vehicles, necessitating
improvements in efficiency, safety, and cost-effectiveness. Considering safety
into the battery design process from the outset is crucial. This involves defining
key safety parameters and conducting thermal runaway (TR) experiements and
simulations, which are complex and costly. Therefore, ther is an urgent need to
reduce costs and speed up design tools to facilitate safer and more efficient EV
battery development.
In this study, a simplified methodology is proposed to simulate the TR in battery
modules, based on the experimental characterization of a single cell in TR. From
one single TR experiment, we measure six initial parameters that allow
calculating all the input parameters of the model. In this work, we apply our
methodology on a test case conducted by Schöberl et al. [1], which
experimentally compare the TR behavior NMC-811 and LFP batteries. One of
the hypotheses we aim to validate is the temporal profile of the ejection during
TR, an aspect that is challenging to evaluate experimentally due to rapid venting
and obstruction from smoke and dust. Furthermore, we observed a significant
influence of the temporal profile of ejection on the cells measured temperatures
and TR propagation. Upon validation, this model can serve as a design tool to
investigate the effects of confinement, geometry, insulation material, or even
chemistry.

RESULTS AND DISCUSSION

Simulations have been performed with finite element method. First, only
conduction between five prismatic cells is considered. A uniform heating power
density of 6.2 W/cm² heats one side of a cell, to reproduce overheating.
According to [1-4], TR starts in a cell when the temperature rise rate exceeds
T > 10 K/s. Then, an unconfined configuration is considered to account for the
impact of gas ejection and convection on TR propagation.

11th International Seminar on Fire and Explosion Hazards, 15-20 June 2025, Rome, Italy.
Book of Abstracts.
269
Several hypotheses regarding the temporal profile of gas ejection are tested,
altering the power distribution during the TR. One considers that the ejection
starts with the TR, splitting the power between the cell and the ejectas. Another
one supposes that the ejection starts when the TR ends, putting at the first stage
all the power into the cell. Consequently, when considering the second
hypothesis, the energy vented is higher.

REFERENCES
[1] J.Schöberl, M. Ank, M. Schreiber, N. Wassiliadis, M. Lienkamp. Thermal runaway
propagation in automotive lithium-ion batteries with NMC-811 and LFP cathodes: Safety
requirements and impact on system integration. eTransportation (2024), 100305.
[2] S. Hoelle, S. Scharner, S. Asanin, O. Hinrichsen. Analysis on Thermal Runaway
Behavior of Prismatic Lithium-Ion Batteries with Autoclave Calorimetry. J. Electrochem.
Soc. 168(2) (2021) 120515.
[3] S. Hoelle, S. Haberl, A. Rheinfeld, P. Osswald, S. Zimmermann, O. Hinrichsen.
Experimental investigation on thermal runaway propagation in lithium-ion battery cell
stack. In: 2022 IEEE transportation electrification conference & expo. IEEE; (2022),
1174-1179.
[4] X. Feng, J. Sun, M. Ouyang, F. Wang, X. He, L. Lu et al. Characterization of
penetration induced thermal runaway propagation process within a large format lithium
ion battery module. J Power Sources 73 (2015), 275-261.

11th International Seminar on Fire and Explosion Hazards, 15-20 June 2025, Rome, Italy.
Book of Abstracts.
270
 ID 160. When Lightning Strikes: Risk Analysis of a
Major Accident at the Matanzas Hydrocarbon Facility
Dueñas Santana J.A.1,4, *,Salzano E.2,Di Benedetto A.3,Van Coile R.4
1
Scuola Superiore Meridionale. School for Advanced Studies. Università degli
Studi di Napoli Federico II. Largo S. Marcellino, 10, 80138, Napoli, Italia
2
Dipartimento di Ingegneria Civile, Chimica, Ambientale e dei Materiali,
Università degli studi di Bologna, Via Terracini 28, 40131, Bologna, Italia
3
Dipartimento di Ingegneria Chimica, dei Materiali e della Produzione
Industriale, Università degli Studi di Napoli Federico II, P.le Vincenzo Tecchio
80, 80125, Napoli, Italia
4
Department of Structural Engineering and Building Materials, Ghent
University, Technologiepark-Zwijnaarde 60, 9052 Ghent, Belgium
*
Corresponding author email: [email protected] or
[email protected]

INTRODUCTION
On August 5-10, 2022, a severe industrial accident occurred at a hydrocarbon
storage facility in Matanzas, Cuba. Initiated by a lightning strike impacting a
50,000 m³ fuel-oil storage tank, the incident escalated into a catastrophic domino
effect, resulting in four large tank failures, significant human casualties, severe
environmental degradation, and extensive economic losses. The complexity of
such domino effects highlights the critical need for advanced methods to assess
the economic and environmental impacts of industrial accidents [1].

RESULTS AND DISCUSSION

Sequence and description of events
The initiating event was lightning striking the geodesic dome roof of tank 52 at
the Matanzas oil facility on August 5, triggering a fire in a tank holding 26,000
m³ of fuel oil. On August 6, approximately at 5:00 a.m., a second tank exploded
at full capacity containing fuel oil. By August 8, all four storage tanks (each with
a capacity of 50,000 m³ containing crude oil and fuel oil) were involved, leading
to severe escalation of the event (Fig. 1).
Thermal radiation and human impact
The thermal radiation emitted reached extreme intensities, significantly affecting
the surrounding areas. The incident resulted in 16 fatalities and 146 injuries.

11th International Seminar on Fire and Explosion Hazards, 15-20 June 2025, Rome, Italy.
Book of Abstracts.
271
After the fireball occurrence, thermal radiation intensities peaked at
approximately 70 kW/m² within a radius of 500 meters, affecting nearly all
nearby industrial units and significantly exacerbating consequences (Fig. 2).

Fig. 1. Escalation pattern occurred in the analysed area.

Fig. 2. Thermal radiation effects from the first fireball in tank 52.
Environmental consequences
An extensive toxic cloud continuously produced over five days, significantly
deteriorated air quality across a vast area. Containing hazardous substances
caused acid rain as far as 100 km away and travelled approximately 300 km.
Economic Impact Assessment
Fire And Explosion Economic Losses Index (FEEL-Index) (based on the DOW
Fire and Explosion Index) quantifies economic hazardous potential by assessing
each process unit's risk based on factors such as accident probabilities, economic
damages, and the potential for domino effects [2]. Considering vessel costs,
storage material losses, and business interruption, estimated damages were
around 56 million USD.

REFERENCES
[1] Cozzani V, Reniers G (2021) Dynamic Risk Assessment and Management of Domino
Effects and Cascading events in the process industry. ISBN: 978-0-08-102838-4.
[2] Dueñas Santana JA, Cuba Arana Y, et al. (2022) Fire and Explosion Economic Losses
(FEEL) Index. Process Safety and Environmental Protection 165:.
https://doi.org/10.1016/j.psep.2022.07.007.

11th International Seminar on Fire and Explosion Hazards, 15-20 June 2025, Rome, Italy.
Book of Abstracts.
272
 ID 163. Investigating the Soil Temperature in Wildfires

Arnold S.1, *, Nagel M.1, Schultze T. 2

1
University of Applied Sciences Dortmund, Institute for Electronics,
Sonnenstrasse 96, 44139 Dortmund, Germany
2
University of Duisburg-Essen, Chair of Communication Systems, Bismarckstr.
81, 47057 Duisburg, Germany
*
Corresponding and presenting author email: [email protected]

INTRODUCTION
This poster presents the results of a preliminary experiment conducted as a
foundation for a larger study on wildland fires. The upcoming experiment will
focus on recording 3D soil temperature data during fir tree branch fires using
thermocouples and pt100 sensors to capture real-time temperature variations.
Experiments measuring the soil temperature have been done before [e.g.1,2], but
not with a high special resolution which is needed as a basis for numerical
simulations. The preliminary experiment aimed to address the following key
questions:
- Evaluating the Setup: Assessing whether the selected thermocouples
and pt100 are suitable for measuring soil temperature during fir tree
fires.
- Testing pt100 Sensors: Determining if pt100 sensors can reliably
measure soil temperature in a certain soil depth, ensuring that the
sensors are deep enough so that the temperature does not exceed 600˚C.
- Comparing Sensor Performance: Analysing the usability of
thermocouples and pt100 sensors in terms of temperature accuracy and
time resolution during a fire to diversify the data collection.
The findings from this initial experiment will guide the design of the final
experimental setup, including sensor arrangement, required soil depth, and the
optimal number of sensors for accurate data collection. The basic structure of the
future experiment has already been completed and can now be finalised with the
help of the results presented here. In addition to the temperature distribution, the
mass loss rate will also be determined in the future experiment. In addition, a
qualitative measurement of the surface temperature distribution will be recorded
using a thermal camera.

RESULTS AND DISCUSSION

Figure 1 shows the setup and the measurement during the fire. Data was recorded
with data loggers and via mqtt. Figure 2 shows the data recorded during three
experiments. It could be shown that with such an arrangement a 3D image of the
soil temperature distribution can be recorded with a time resolution of 1 second.

11th International Seminar on Fire and Explosion Hazards, 15-20 June 2025, Rome, Italy.
Book of Abstracts.
273
Fig. 1. Setup of Sensors during the fir tree fire experiments. Thermocouples with
different diameters and pt100 are arranged in a certain order to compare their
performance.

Fig. 2. Sensor data recorded during three experiments. Left: small fire with wood
chips, Center: dry fir tree fire, Right: Green fir tree fire. Pt100 were calibrated in
the climate chamber up to 180˚C, therefore data only up to this value is
presented.
For further analysis of the sensor performance the data from the five sensors in
the middle (thermocouples 1, 2, 3 and pt100 3, 7) was compared. It was
demonstrated that both measurement systems are suitable for measuring soil
temperature. Thermocouples respond more quickly and are therefore preferable
for dynamic measurements. However, pt100 sensors also provide data at a
sufficient speed, making them a viable option for complementary measurements.
The soil depth should exceed 3 cm, as even during smaller fires, temperatures of
up to 600°C can be reached at this depth, which is the upper limit of pt100
sensors. For future experiments the pt100 sensors must also be calibrated for
higher temperatures.
REFERENCES
[1] E. Enninful: Predicting Temperature Profiles in Soil During Simulated Forest Fires,
University of Saskatchewan, July 2006. Published: International Journal of Wildland
Fire, 17(2) 205-213, 2007.
[2] Sacket, S.S., Haase, S.M.: Measuring soil and tree temperatures during prescribed
fires with thermocouple probes, United States Department of Agriculture, Forest
Service, Pacific Southwest Research Station, General Technical Report PSW-GTR-
131,1992.

11th International Seminar on Fire and Explosion Hazards, 15-20 June 2025, Rome, Italy.
Book of Abstracts.
274