Examination of Harmful Substances Emitted to the Environment During an Electric Vehicle Fire with a Full-Scale Fire Experiment and Laboratory Investigations

Abstract

Nowadays, electromobility has a significant role in transportation; different electrically driven vehicles are spreading continuously. Due to this form of drivetrain, fire safety hazards have also changed when compared to those of conventional vehicles. Lately, electric vehicle fires have become more common; thus, we have chosen to investigate the negative impacts of these fires on humans and the environment, in addition to the toxic properties of the resulting combustion products. In our research work, we conducted a full-scale fire experiment on an electric passenger car. Fire extinguishing was executed with fire-fighting foam, and its efficiency was examined. After extinguishing the fire, we took samples from the combustion gases and soil. Samples were subjected to laboratory investigations. Our results and experiences are presented in this article.

1. Introduction

The spreading of electromobility has promoted the use of several different batteries in electric vehicles in traffic, which can be passenger cars, heavy-goods vehicles, buses, e-scooters, and also bicycles or even skateboards. The increase in the number of electric vehicles is due to the quick development of lithium-ion and lithium-polymer technologies, because their application enhances zero-emission drives [1]. Lithium-ion batteries (LIBs) have an outstanding energy density–mass ratio, which makes them superior to other battery technologies, such as lead–acid (PbAB) or nickel–metal hydride (NiMH) batteries. One reason for this is lithium, the lightest metal, which has great electrochemical potential and a specific charge-transferring ability, which offers an advantage regarding selection [2]. In the case of electric vehicles, a very high battery capacity is needed because, depending on performance, their consumption of electric power is significant; as such, batteries must have great energy capacity and excellent performance. Along with this, it is important to eliminate the weight and space limits of vehicles, and they should also remain marketable. Battery systems thus differ from each other not only in structure but also in design. There are three main cell layouts applied in LIBs; these package types are cylindrical, pouch, and prismatic cells [3].

In our article, we assess fires in electric vehicles; further, we put an emphasis on batteries built for electric passenger vehicles. Electric vehicles require a great number of battery cells to produce the necessary performance and amount of energy, and so for the design of their placement, the main goal is to maximize the capacity of the battery whilst maintaining proper safety. Generally, battery packs are mounted on strengthened and rigid compartments or are placed so that their mechanical impacts and any injuries are minimal. This area is called the safety zone of vehicles. This zone is considered the range between the wheelbase and the middle of the chassis, which means batteries are placed in a T-shape or square on the floor (sometimes as rectangular battery packages), or the battery package is situated higher in the back of the car [2]. In newer developed models, a so-called H-shape organization is applied, as represented in Figure 1.

As the size and capacity of battery packs have increased in electric vehicles, so has the probability of fires. Li-ion batteries do not necessarily pose a spontaneous fire hazard if operation conditions are stable and do not change to a great extent; however, changing and extreme operating conditions that cause mechanical impacts and electric and thermal impulses on Li-ion batteries tend to cause them to ignite [4]. The fire hazards of electric vehicles are not zero, even in the case of normal operation, because spontaneous traffic accidents can cause fires as well [5,6]. We are aware from our experience that electric vehicle fires are also started during battery charging and in standby, and they are followed by intense fire spread and heat and smoke production [7,8]. This depends on the materials that are applied and used to build the batteries, chassis, and passenger compartment and how strongly toxic combustion products form and escape into the environment during the burning process [9,10].

Battery electric and internal combustion engine vehicles were tested in laboratory conditions by the authors of [11], and they concluded that total heat release was not affected by traction energy. However, combustion gases showed significant differences, mainly in the concentration of HF. In another study, the heat release rate was monitored for electric, conventional, and fuel-cell electric vehicles in laboratory conditions. The so-called peak heat release rate (pHRR) was found to be between 6.51 and 7.25 MW for EVs; this is a little bit lower than the value for an ICEV (7.66 MW) but higher than for an FCEV (5.99 MW) [12]. Furthermore, in a study about electric and hybrid cars, electric buses were also reviewed for their battery fires. The research results showed that the heat release rate is similar in the case of electric and internal combustion engine vehicles as well; however, toxic gases released during the fire itself occurred more in the case of electric vehicles, for instance, due to fires from Li-ion batteries resulting in a higher concentration of HF [13].

Based on the previous statement, a scientific investigation into the effects of combustion product and heat can help us to more precisely understand these phenomena during electric vehicle fires. In order to achieve more accurate results, a full-scale fire experiment was conducted. Regarding the experiment, we ignited an electric vehicle in open space (on a fire service training area) and not in laboratory conditions; during this time, samples from combustion gases were taken and then were analyzed in a laboratory environment. Another goal of the experiment was to use as little fire extinguisher as possible to put the fire out successfully. After putting out the fire successfully, samples were also taken from the soil and then transported to the laboratory. The results and experiences represented further in this article will contribute to the work of firefighters.

2. Materials and Methods

The experiment was conducted in Slovakia, on the base of the Fire and Rescue Authority of Galanta, featuring restricted conditions and qualified first-response personnel. In the experiment, the restricted burning of a Peugeot e208 EV type passenger car (Peugeot, Paris, France) was conducted. The vehicle was donated to the fire department by PSA Peugeot Citroën Slovakia Trnava for training.

Technical data, according to manufacturer’s description Peugeot e208EV:

Battery type Number of cells Pack configuration Nominal voltage Nominal capacity Useable capacity

Lithium-ion (Modell 2P6S) 216 108s2p 400 V 50.0 kWh 46.3 kWh

The manufacturer donated the vehicle with a 20% state of charge for the fire drill. Before the drill itself, the battery was charged to 60%; the reason for this is that a transporting EV’s average state of charge is this value. To preserve soil under the experiment area, foil with an evenly spread 10 cm soil cover was used. The vehicle towed to the prepared experiment area was lifted 0.6 m in the air and supported under the sill (specifically used for this purpose) with a metal rod. Before the experiment, the protecting shield of the battery was opened at the front and the back to form a 0.3 × 0.4 m area; then, at a distance of 0.4 m, an open-flame ignition source was placed directly in front of the opened gap (Figure 2).

We decided to ignite from underneath, because this is how the battery could be most easily accessed, as we aimed to burn the battery first. An open-flame ignition source was placed directly under the battery pack. Right before the experiment, the whole area was closed; only the first responders could enter, with individual protection. Firefighters moved away from the vehicle after igniting the open-flame source. The first minute after ignition is presented in Figure 3.

The temperature of the open flame upon reaching the battery was 650 °C, which was measured with a thermocamera by a firefighter (as shown in Figure 3, marked with red). The figure above also shows that the package in the back burnt with greater intensity.

In the fifth minute after the ignition combustion became self-sufficient, ignition sources were removed from the experiment area. Fire spread more and more heavily throughout the vehicle; combustion gas sampling started from the 6th minute and lasted until the 11th minute. Samples were taken from the combustion gases; to do so, in the sampling probe of the pipeline, an SKC Anasorb CSC glass pipe (SKC Anasorb, Eighty Four, PA, USA) was mounted, which has active carbon with a large specific surface, also containing fiberglass and a filter. In the sampler, active carbon binds air-polluting materials. The pipeline containing the probe was connected to an AirCheck XR500 (Fluke, Washington, DC, USA) (2 L/min performance) sampler pump, which was mounted on a 2 m long telescopic rod. For a 5 min duration, exhaust gases flowed through the sampling system. Sampling is represented in Figure 4. The SKC Anasorb CSC sampler pipe is marked with a red circle in the top left corner of the figure.

It is clearly visible from the figure above that the fire burned with high intensity; after it spread to the panels of the vehicle, the plastic parts melted. At this time, fumes could also be seen coming from the passenger compartment.

Blank samples had to be taken to validate laboratory investigations; thus, samples from the experiment area were also taken by breaking a glass pipe open to ensure that we had samples from the pure air before the experiment. Since sampling was conducted from the air polluted by the gases that formed during combustion, the following influencing factors must be considered: atmospheric pressure (p₀), wind velocity (v₀) (there was no wind on the day of the experiment in that area), temperature (T₀), and relative humidity. The data measured on site are included in

Table 1. Influencing factors in the experiment.

Examined Factors	Measured Values
p0	1013 hPa
T0	283 K
v0	0 m/s
ρ	65%

.

After taking the samples, the glass pipe was closed, and then a protecting cover was used to preserve its condition until it reached the laboratory. At the same time as the glass pipe sampling, a Dräger X-am^® 5600 gas detector (Drager, Lubeck, Germany) with infrared sensor technology was also used. This device is commonly applied to analyze air content on the spot; it is also capable of analyzing and monitoring six different gases. The device we used was converted in order to measure the concentration of O₂, CO, CO₂, HCN, and H₂S components. The device was calibrated and checked during the experiment, and it was held into the flow of combustion gas by a trained firefighter. The results were recorded and are shown in

Table 2. Dräger X-am^® 5600 gas detector results.

Examined Factors	Measured Values
CO2 [volume%]	0.12–0.15
CH4 [%LEL]	0
O2 [volume%]	20.9
HCN [ppm]	0
H2S [ppm]	0
CO [ppm]	40–150

.

The CO₂ content of air is about 0.040 [v/v %] normally [11]. Our measured values were detected in the smoke flow and exceeded the values above, but further from the fire, this concentration became diluted. CO₂ is an element of the air; in small amounts, it is not poisonous, but if it accumulates in closed spaces, it extrudes oxygen. Since it has no color or odor, without proper sensors, it cannot be detected [11]. In the case of CO 40–150 ppm, calculated as 0.004–0.015 [v/v %] of 0.02 [v/v %], it is toxic, and at a concentration of 1.28 [v/v %], 2–3-min of inhalation is enough to be lethal [14].

During the fire experiment, we observed that the intensity of combustion increased continuously; meanwhile, the color of the smoke changed several times, from light grey to dark black. The reason for this is that fire spread to the plastic and rubber parts of the vehicle and also to its painted surfaces. Soot formation was also observable; this was confirmed by the fact that along with the concentration of oxygen, the presence of CO and CO₂ was significant and changed in accordance with the burnt materials.

In the 12th minute of the experiment, fire extinguishing started with one foamed water mist jet. The extinguisher foam consisted of water, foaming agent (1–6%), and air, which were produced at the site of the fire extinguishing. During the practice, a 3% foam mixture was applied. The combustion process ceased, because in EU countries, firefighters have to reach the furthest point from the operation area 10 min after the alarm. The preparation before putting a fire out can take up to 2 min. At this time, the passenger compartment and the engine bay of the vehicle were burning; moreover, the tires and the bumpers were burning, and the first windows also broke. The extinguishing of the battery package took place first, and then the same treatment was applied to the other parts of the vehicle (Figure 5).

No signs of thermal runaway were detected by firefighters. According to their experience, using foam for firefighting requires less extinguishing agent than using water, meaning less runoff of fire-fighting agent that is contaminated with combustion products enters the environment. Despite the use of fire-fighting foam, the vehicle reignited from the battery pack after a short period of time following the extinguishing. The fire fighting continued with one foamed water mist jet. The battery pack continued to emit fumes after putting the fire out, and multiple instances of cooling with a water mist jet were necessary. In Figure 6, it is clearly visible that strong fumes were steaming out of the battery. We assumed that without proper cooling, the vehicle would likely have reignited again.

After the firefighting was completed, the vehicle could not be dismantled, as examining the structures was not part of the experiment.

After finishing these works, the vehicle was transported, and then a soil sample was taken from a depth of 10 cm beneath the vehicle; this was followed by control soil samples, which were taken from an area 50 m away from the vehicle (within the fire brigade training area), also at a depth of 10 cm. The samples were transported to a specialized testing laboratory for further analysis.

3. Results

3.1. Air Examinations

The laboratory testing of combustion gas samples is accredited for the detection of volatile organic hydrocarbons; therefore, the presence of these compounds was identified.

MSZ EN 13649:2002 [15] describes the “Emissions of stationary sources of air pollution: Determination of the mass concentration of unique gaseous organic compounds”. The instructions of the European standard “Active carbon and solvent desorption method” were followed. The mentioned standard describes the sampling of volatile organic compounds generated during chemical processes on activated carbon and the procedures for sample preparation and analysis. The measurement of individual gaseous organic compounds consists of three steps, which are as follows:

• sampling of the final gas;
• treatment of the sampled material;
• gas chromatography analysis.

When sampling from the final gas, care should be taken to ensure that organic compounds in the gas of a known volume are bound to the activated carbon, that solid particles that interfere with the measurement are removed, and that water condensation does not occur.

The detected samples were analyzed by flame ionization–gas chromatography (GC-FID).

The conditions used for the chromatographic analysis were as follows:

• Device: Agilent Technologies 6890N Network GC System
• Column type: Restek Rtx^®-1 30 m 0.53 mm ID 3 µm
• Mode: Splitless
• Gas: H₂
• Injection temperature: 250 °C
• Heating program: After sample injection, the instrument heated the column space at 35 °C for 5 min, then at a heating rate of 30 °C/min to 250 °C, which was maintained for 1 min.

Chromatography of the sample is necessary to determine the different mass flow rates of the polluting compounds.

Table 3. Basic data for sample mass flow calculations.

Basic Data	Quantities
Measured volume of dry end gas [V]	6 L
Sample air pressure [p]	1011.38 hPa
p0	1013 hPa
T0	273 K
Real temperature of end gas [T]	292 K
mbenzene	0.125 mg
mtoluol	0.017 mg
mxylene	0.0089 mg
mesitylene	0.0099 mg
methyl-acetate	0.021 mg
methyl-acetate	0.008 mg
mn-butyl-acetate	0.005 mg
mC-sum	0.271 mg

shows the data that are required for the determination of the mass flow.

The concentration of the compounds present in the air were determined in mg/m³ according to Formula (1), corresponding to the standard MSZ EN 13649:2002 [15]:

$$c_i={\displaystyle \frac{m_i}{V_{cor}}}\times 1000,$$

(1)

where, c_i is the concentration of the i-th compound [mg/m³]; m_i is the mass of the i-th compound [mg]; V_cor is the volume of the sample in liters, in a normal state (273 K and 1013 hPa), appealing to dry gas. The concentration of the normal sample was determined according to (2):

$$v_{cor}=V\times {\displaystyle \frac{p_0}{p}}\times {\displaystyle \frac{T}{T_0}}$$

(2)

where V is the measured volume of the dry gas sample; p is the pressure of the sampled air [hPa], p₀ = 1013 hPa; and T is the actual temperature of final gas sample [K], T₀ = 273 K.

Substituting the measured data of each component into the above formulae, the quantity of pollutants in the air in mg/m³ can be obtained. The calculation is presented below for benzene, from which the concentration of the other pollutants was determined.

$$\begin{gathered} v_{cor}=\mathrm{V}\times {\displaystyle \frac{{\mathrm{p}}_0}{\mathrm{p}}}\times {\displaystyle \frac{\mathrm{T}}{{\mathrm{T}}_0}}=10\times {\displaystyle \frac{1013}{1011.38}}\times {\displaystyle \frac{292}{273}}=\mathrm{10,713}\mathrm{L} \\ {c}_{benzene}={\displaystyle \frac{{\mathrm{m}}_{\mathrm{i}}}{{\mathrm{V}}_{\mathrm{c}\mathrm{o}\mathrm{r}}}}\times 1000={\displaystyle \frac{0.125}{10.713}}\times 1000=11.67\mathrm{m}\mathrm{g}/{\mathrm{cm}}^3 \end{gathered}$$

The calculated concentrations of the detected compounds are given in

Table 4. Calculated concentration of detected compounds.

Compound	Quantity [mg/m3]	Allowed Limit [μg/m3] *
Cbenzene	11.67	10
Ctoluol	1.59	200
Cxylene	0.84	60
Cmesitylene	0.93	1000
Cmethyl-acetate	2	70
Cethyl-acetate	0.75	100
Cn-butyl-acetate	0.47	100
CC-sum	25.3	-

* Twenty-four-hour limits according to 4/2014. (I. 14.) Regulation introduced by Ministry of Agriculture and Rural Development [16].

.

The volatile organic compounds in the spectrograms shown in the table above were present in the samples, making a total hydrocarbon concentration; this is because the individual components were present in trace amounts in the samples, but in total concentrations, they were also significant. Analysis of the results shows that, with the exception of benzene (which was slightly below the permitted limit), all the compounds identified exceeded the permitted limit. All of these compounds are harmful to health and the environment and can be fatal after prolonged exposure. The harmful effects of these compounds are listed in

Table 5. Health-harming effects of the detected compounds.

Compound	Harmful Effects *
Benzene	Carcinogenic; damages the bone marrow, liver, and immune system
Toluol	Causes eye and respiratory irritation and depression
Xylol	It causes respiratory irritation; damages the liver, kidneys, and central nervous system
Mesitylene	Causes eye, skin and respiratory irritation; damages internal organs
Methyl-acetate	Causes eye and skin irritation, dizziness, and drowsiness
Ethyl-acetate	Causes eye and respiratory irritation, loss of consciousness, and damage to the central nervous system
N-butyl-acetate	Causes eye, skin, and respiratory irritation

* VAX: Based on the Hazardous Materials Manual in the Hungarian language.

.

The combustion gases were analyzed for the detection of N-methylpyrrolidone C₅H₉NO (NMP), an industrial solvent widely used in the production of Li-ion batteries, circuit boards, liquid crystal electronics, and semiconductors. It has serious adverse health effects such as reproductive toxicity, severe eye irritation, skin irritation, and respiratory irritation [17].

The samples were analyzed with a flame ionization detector (GC-FID). The result obtained for the concentration of NMP was below the detection limit, but the chromatogram clearly showed that a well-defined peak in NMP was detected at the signal-to-noise ratio. NMP was clearly released into the air during combustion of the Li-ion battery.

3.2. Soil Examination

The determination of the amount of potentially toxic elements (heavy metals) accumulated in the soil during the burning experiment was conducted with inductively coupled plasma–mass spectrometry (ICP-MS).

This examination consists of three steps:

• Representative subsampling;
• Exploration of subsamples in an acidic medium with microwave assistance;
• Qualitative and quantitative element determination with ICP-MS examination.

We took representative soil samples from the burning area as well as control soil samples. From these samples, three representative soils subsamples were selected and quartered to an individual mass of 0.3 g each. Parallel analytical measurements were performed on the subsamples.

According to the EPA 3051A “Microwave assisted acid digestion of sediments, sludges, soils, and oils” standard, the subsamples were subjected to microwave-assisted acid digestion in a 3:1 mixture of nitric acid (HNO₃) and hydrochloric acid (HCl) in an Ethos EASY-type microwave digester; then, for mass spectrometric measurements, the solutions were diluted to 200 cm³.

During the ICP-MS test, the sample in the solution was injected into argon (Ar) plasma at a temperature of 7000 K, where the elements of the sample were atomized/ionized. The resulting discrete ions were selected according to their mass/charge (m/z) ratio in quadrupole analyzer(s) operating on the electromagnetic principle, and in the case of some devices, in collision cells filled with auxiliary gases. The ICP-MS tests were performed on an Agilent 8900-type triple quadrupole (tandem) inductively coupled plasma–mass spectrometer (ICP-MS/MS, ICP-QQQ), using He and O₂ auxiliary gases, based on the expanded EPA 6020B “Inductively coupled plasma–mass spectrometry” standard.

During evaluation, with the help of the solution order corresponding to the given element, a measurement curve could be compiled. The examination consisted of 47 elements, from which metals and semi-metals were selected according to Joint Decree No. 6/2009 (Joint Decree 6/2009 (IV. 14) KvVM-EüM-FVM on the limit values necessary for the protection of the geological environment and groundwater against pollution and the measurement of pollution) [18] in Annexes 1 and 3. In addition to the potentially toxic elements (PTEs), which are listed in the joint decree, we also measured the concentration of manganese (Mn). The results from three parallel measurements had a deviation of less than 5%. The arithmetic averages of the measurement results, rounded to whole numbers, as well as the limit values, are summarized in

Table 6. Comparative table of the elemental compositions of the soil samples provided at the incineration and control sites, as well as the pollution limit values (B) included in Joint Decree 6/2009.

Pollutant	Soil Sample from the Burning Area (mg/kg)	Control Soil Sample (mg/kg)	(B) Pollution Limit (mg/kg)
Manganese (Mn)	361	232	-
Chromium (sum) (Cr)	32	24	75
Cobalt (Co)	226	27	30
Nickel (Ni)	1501	143	40
Copper (Cu)	748	401	75
Zinc (Zn)	792	401	200
Arsenic (As)	5	4	15
Selenium (Se)	<1	<1	1
Molybdenum (Mo)	N/A *	N/A *	7
Cadmium (Cd)	2	<1	1
Tin (Sn)	N/A *	N/A *	30
Barium (Ba)	1084	435	250
Mercury (Hg)	N/A *	N/A *	0.5
Lead (Pb)	84	127	100
Silver (Ag)	<1	<1	2
Antimony (Sb)	N/A *	N/A *	5
Boron (B)	32	34	1000

* Not applicable.

. We would like to note that the VI. separation of Cr, which is listed separately in the joint decree, was not suitable. No quantitative analysis was performed for molybdenum (Mo), tin (Sn), antimony (Sb), or mercury (Hg). The quantitative analysis of aluminum was not suitable because the aluminum that originated from the burning experiment could not be distinguished from natural aluminum, which occurs in aluminosilicate minerals.

Summarizing the achieved results, it can be concluded that the control samples indicated that the soil was polluted with different metals, such as copper, zinc, and barium; however, during burning, further significant amounts of toxic elements (heavy metals) entered the soil. Nickel stood out, in an amount that was 10 times higher, and for cobalt, 8 times higher concentrations were detected. Copper, zinc, cadmium, and barium were present in 9 times higher concentrations. These elements are used in large amounts in the production of Li-ion batteries, and these entered the soil with the foam fire extinguisher.

Despite being a non-toxic element, the concentration of manganese also rose after the burning experiment. Given that samples had no noticeable microscopic residuals, it can be assumed that pollutants disperse in the soil and are adsorbed on the reactive surface of organic compounds in the soil, sheet silicates (such as clay minerals), and iron oxide/oxyhydroxide minerals.

It is important to mention that the analytical results came from a single experiment in which analytical measurements on car parts (such as paints) were not performed. The origin of the heavy metals whose concentrations rose in the soil should be examined in further burning experiments.

4. Discussion

During the planning and execution of the full-scale fire examination that we conducted, we regarded our experiences of fire incidents as a starting point. A 10-min pre-burning process was determined, corresponding to the recommendations of EU firefighting interventions; in accordance, the qualified interventional firefighting units in the operational area needed 10 min to respond to a fire at the farthest point of the operational area. Preparation for extinguishing a fire usually takes 2 min, which is why the extinguishing of the burning vehicle started in 12 min in our experiment. The results of combustion gas samples and on-site gas detector measurements showed that during the electric vehicle fire, a significant amount of combustion gas entered the environment and posed a threat to human health and the environment. People involved in or very near to the fire incident, if they are directly exposed to it, can suffer from permanent health damage. Even if the concentration of combustion gases dilutes further from the fire, some components remain for longer as pollutants. Almost every volatile organic compound exceeded the allowed concentration limits; however, we must remark that the total hydrocarbon concentration was also significant and could have health- and environment-damaging effects. The soil samples showed that potentially toxic heavy metals such as nickel accumulated at almost 10 times normal values, and cobalt exceeded normal values by 8 times. This can be clearly connected to electric vehicle fires. Copper, zinc, cadmium, and barium approximately doubled in concentration as a consequence of the electric vehicle fire. The experiment provided several useful experiences for firefighting units as well. It was observed that as the intensity of the fire increased and the fire spread, smoke generation changed. Extinguishing the electric vehicle fire also presented challenges, as during the experiment, the extinguished vehicle caught fire several times from the side of the battery cells. Extinguishing the fire itself, including cooling, took more than an hour. It is important to correctly choose the fire extinguishing agent in use (in our case, this was fire extinguishing foam, which is a mixture of water, foaming agent, and air) and extinguishing tactics, because more pollutants enter the air and soil during ineffective fire extinguishing.

5. Conclusions

Thanks to the spread of electromobility, the number of vehicles with various forms of electric driving in traffic is constantly increasing, which has resulted in an increase in fire safety risks when compared to vehicles driven with an internal combustion engine. In order to gain a more accurate understanding of the phenomena and environmental effects related to electric vehicle fires, we conducted a full-scale fire experiment with an electric passenger car and then took samples from the combustion gases produced during the fire and from the soil under the vehicle after the fire was extinguished. The samples were subjected to laboratory tests. We analyzed our results, during which we found that electric vehicle fires produce a significant amount of combustion gas, which poses a threat to human health and the environment (for example, through toxic heavy metals leaking into the soil).

It is important to note that the composition and amount of hazardous substances released into the environment during the combustion of each car type may vary, as there are differences between the types of materials used in the construction of vehicles and their batteries. We should note the passengers travelling in these vehicles and the people in the area of a possible fire, as when they are exposed to combustion products, they are certain to suffer damage to their health due to the harmful substances produced. The experiences and results of the experiment can be used during training exercises carried out by intervening firefighting units. In the future, we plan to carry out further combustion experiments and laboratory tests; we would like to experiment with Li-ion battery packages separately, with an open-flame ignitor and mechanical-abuse ignition, then apply a water-soaking extinguisher and analyze the residual extinguisher water. With our results, we wish to draw attention to the risks associated with electric vehicles.