Effects of Ambient Pressure on Burning Characteristics of Gasoline: A Pilot Study

A fire can pose a significant threat to a building’s occupants and leads to property damage. The burning characteristics usually determine the severity of the accident. Environments in high-altitude areas feature low oxygen content and ambient pressure, which can influence the burning characteristics of combustibles. In this paper, a series of field experiments were conducted to investigate the burning characteristics of gasoline at different altitudes considering heat release rate (HRR), flame height, and smoke release rate. Results show that the combustion process can be divided into three stages: initial stage, stable stage, and attenuation stage. Lower oxygen content and ambient pressure reduce the HRR; for example, the HRR at an altitude of 4150 m is nearly half at an altitude of 500 m, contributing to a lower smoke release rate. The HRR is proportional to 1.3 power of atmospheric pressure, and a fitting equation was brought out in this paper. Flame height increases with the increase in altitude due to the demand for more oxygen during the combustion process since the oxygen content is low in high-altitude areas.


Introduction
With the development of urbanization and the economy in China, increasingly more industrial and civil architecture and traffic facilities have been constructed to satisfy all kinds of resident demands [1,2]. Fire in structures like buildings or tunnels can pose a significant threat to the occupants and lead to property damage, especially in a place where the population density is quite large [3][4][5].
A fire is a process that involves the ignition and combustion of materials, which generates heat and smoke, endangering peoples' lives and the safety of the structures [6,7]. Burning characteristics such as heat release rate (HRR)-the most important factor in a fire-fire growth speed, smoke release rate, flame height, and smoke toxic directly determine the severity of fire accidents [8][9][10]. Numerous researchers have investigated the burning characteristics of a fire [11,12]. HRR is influenced by many factors, such as the type and amount of combustibles, environmental parameters (i.e., ventilation conditions and oxygen supply), and space type [13][14][15]. For a heavy-duty vehicle carrying a tank full of dangerous combustible goods, the HRR could reach 100 MW [16]. Higher HRR can lead to fire spread between compartments and vehicles [17][18][19]. The influence of wind speed conditions on these burning characteristics, especially for the HRR, can be quite complex [15,20]. The HRR increases or decreases depending on the burning conditions, such as the size of the fire source and wind speed [21,22]. Space type may affect the heat and mass transfer [23][24][25]; for example, a fire inside a road tunnel may be more severe than other engineering structures because it is long and narrow, which is more challenging to discharge smoke and heat that the fire source releases [26][27][28]. The rate of fire growth can be determined by a formula based on exponential growth [29], which varies according to the fire growth coefficient of the burning material. For a fire induced by synthetic materials, such as polyurethane [30], the fire growth rate is quite high. When the fire develops rather quickly, it can be difficult to extinguish or reach a safe zone for stranded people. Little time is left for people to evacuate once they perceive and interpret signs of the fire [31], which has been validated by many fatal incidents featuring rapid fire development after its initial discovery [32,33]. Flame height is an important sign in the initial stage of detecting and confirming the existence of a fire, which can be affected by fire source dimensions and ventilation conditions. Smoke yield and toxicity are a hot topic in fire research since many casualties have resulted from smoke instead of the fire itself [34,35]. It is reported that stranded people who leave for a safe zone by passing through a smoky environment often have to change direction or even retrace their steps since the existence of smoke can cause breathing problems, low visibility, psychological panic, and so on [36,37].
From the above content, we can conclude that exploring burning characteristics during a fire may be of vital importance. Few studies focus on fires in high-altitude areas where environmental parameters, such as ambient pressure and oxygen content, differ from those in plain areas [38][39][40]. However, with the Western Development Strategy and improvement of traffic facilities in western China [41,42], increasingly more buildings and traffic facilities have been constructed in high-altitude areas. Therefore, it is necessary to explore burning characteristics, such as smoke release rate and HRR in fire accidents, to provide a reference for fire prevention and treatment [43][44][45].
This work aims to analyze the effects of ambient pressure on the burning characteristics of gasoline. A series of field experiments were conducted in different areas at different altitudes that consider the HRR, flame height, and smoke release rate. We hope our work can shed light on the characteristics of fire accidents in high-altitude areas and develop strategies to control fires and reduce economic losses and casualties. The structure of the paper is as follows: In Section 2, a description of the research sites is provided, and details on the measurement procedure are listed. Experimental results are listed in Section 3; analyses and discussions of the results are presented in Section 4. Finally, we reach a set of clear and concise conclusions in Section 5.

Materials and Methods
In order to explore the burning characteristics of gasoline at different altitudes, three sites listed in Table 1 were selected as experimental sites. The altitude for these sites ranges from 500 m to 4150 m. The burning experiments were conducted in an open environment in these sites. The environmental features, such as ambient pressure, temperature, humidity, and oxygen content vary with altitude. In order to obtain the environmental parameters in different experimental sites, several environmental devices listed in Table 2 were adopted in this study. In order to eliminate the effects of ambient temperature on the experiment results, all field experiments were conducted when the ambient temperature was around 24 • C.  The burning experiments were conducted using the test platform shown in Figure 1. The platform consists of two parts: a burning and measuring system and a recording system. The burning and measuring system includes a fuel plate, two thermal insulation boards, a balance, and a steel ruler. The fuel plate is 40 cm long, 30 cm wide, 4 cm high, and 4 mm thick. A balance with a range of 0-30 kg and accuracy of 0.1 g is set at the bottom to obtain the mass variation of gasoline in the fuel plate. In order to prevent heat damage from the plate to the balance, two thermal insulation boards are placed between the balance and plate. The recording system includes two sets of tripods with mobile phones. One tripod is installed beside the steel ruler to record the flame height during the burning process. The other is placed beside the balance to acquire the mass variation during the burning process. The burning experiments were conducted using the test platform shown in Figure 1. The platform consists of two parts: a burning and measuring system and a recording system. The burning and measuring system includes a fuel plate, two thermal insulation boards, a balance, and a steel ruler. The fuel plate is 40 cm long, 30 cm wide, 4 cm high, and 4 mm thick. A balance with a range of 0-30 kg and accuracy of 0.1 g is set at the bottom to obtain the mass variation of gasoline in the fuel plate. In order to prevent heat damage from the plate to the balance, two thermal insulation boards are placed between the balance and plate. The recording system includes two sets of tripods with mobile phones. One tripod is installed beside the steel ruler to record the flame height during the burning process. The other is placed beside the balance to acquire the mass variation during the burning process.

°C.
The fuel in this study was gasoline. As a common combustible substance, it provides stable HRR during the burning process and has been widely used in numerous fire studies [8,26,46]. The gasoline adopted in this study was No. 93 gasoline under Chinese standard, which has a density of 0.73 g/cm 3 and an effective combustion heat at46 MJ/kg. A total of 0.22 kg of gasoline was burned for each burning. The process of the burning experiments is shown in Figure 2. Before the burning experiments, the environmental parameters were recorded. For each research site, the burning experiment was repeated five times. The fuel in this study was gasoline. As a common combustible substance, it provides stable HRR during the burning process and has been widely used in numerous fire studies [8,26,46]. The gasoline adopted in this study was No. 93 gasoline under Chinese standard, which has a density of 0.73 g/cm 3 and an effective combustion heat at46 MJ/kg. A total of 0.22 kg of gasoline was burned for each burning.
The process of the burning experiments is shown in Figure 2. Before the burning experiments, the environmental parameters were recorded. For each research site, the burning experiment was repeated five times.

Environmental Parameters
The results of environmental parameters before the experiments were conducted for each site are listed in Table 3. With the increase in altitude from 500 m in Sichuan to 4150 m in Tibet, the ambient pressure decreased from approximately 97 kPa to 62 kPa. The ambient pressure in Yunnan at an altitude of 1950 m is 81.1 kPa, which falls between that in Sichuan and Tibet. As a controlled variable, the temperature of the three experimental sites was controlled at approximately 24 °C. The humidity dramatically decreased to 26.4% in high-altitude areas such as Tibet, while it remained at approximately 80% in Sichuan and Yunnan. The oxygen volume percentage is approximately 20.9% regardless of the altitude. The air density, which is related to ambient pressure, decreases in high-altitude areas, so the oxygen content reduces.

HRR
The HRR is one of the most important parameters in fires, so it is necessary to investigate the effects of ambient pressure on HRR. In this paper, the HRR is obtained from the mass variation per unit area of gasoline in the fuel plate, and the value of mass variation is the average of five repeated experiments in each site. Figure 3 illustrates the mass variation of gasoline during the burning process at three experimental sites.

Environmental Parameters
The results of environmental parameters before the experiments were conducted for each site are listed in Table 3. With the increase in altitude from 500 m in Sichuan to 4150 m in Tibet, the ambient pressure decreased from approximately 97 kPa to 62 kPa. The ambient pressure in Yunnan at an altitude of 1950 m is 81.1 kPa, which falls between that in Sichuan and Tibet. As a controlled variable, the temperature of the three experimental sites was controlled at approximately 24 • C. The humidity dramatically decreased to 26.4% in high-altitude areas such as Tibet, while it remained at approximately 80% in Sichuan and Yunnan. The oxygen volume percentage is approximately 20.9% regardless of the altitude. The air density, which is related to ambient pressure, decreases in high-altitude areas, so the oxygen content reduces.

HRR
The HRR is one of the most important parameters in fires, so it is necessary to investigate the effects of ambient pressure on HRR. In this paper, the HRR is obtained from the mass variation per unit area of gasoline in the fuel plate, and the value of mass variation is the average of five repeated experiments in each site. Figure 3 illustrates the mass variation of gasoline during the burning process at three experimental sites. Based on the results of Figure 3, the mass variation rate of gasoline can be obtained by the following equation: M represents the mass variation rate of gasoline, (kg/s); t 1 and t 2 represent two adjacent observation points, (s); M t 1 and M t 2 represent the mass of gasoline at time t 1 and t 2 separately, (kg). The mass variation rate per unit area of gasoline at three experimental sites is shown in Figure 4.  Based on the results of Figure 3, the mass variation rate of gasoline can be obtained by the following equation: where, ̇ represents the mass variation rate of gasoline, (kg/s); 1 and 2 represent two adjacent observation points, (s); 1 and 2 represent the mass of gasoline at time 1 and 2 separately, (kg). The mass variation rate per unit area of gasoline at three experimental sites is shown in Figure 4.    In Figure 4, the burning process has some similar features in the three experiment sites. In sequence, they all experience three stages: a gradual increase in mass variation rate, a fluctuation in mass variation rate, and a gradual decrease in mass variation rate. Consequently, the whole burning process can be divided into three stages according to the aforementioned features: initial stage, stable stage, and attenuation stage with different time durations in different experimental sites. Table 4 lists the time durations of different stages in the three experimental sites. In order to eliminate interference from the environment during the burning process, as shown in Figure 5, the Karman filtering method was adopted to process the original data in Figure 4. In Figure 4, the burning process has some similar features in the three experiment sites. In sequence, they all experience three stages: a gradual increase in mass variation rate, a fluctuation in mass variation rate, and a gradual decrease in mass variation rate. Consequently, the whole burning process can be divided into three stages according to the aforementioned features: initial stage, stable stage, and attenuation stage with different time durations in different experimental sites. Table 4 lists the time durations of different stages in the three experimental sites. In order to eliminate interference from the environment during the burning process, as shown in Figure 5, the Karman filtering method was adopted to process the original data in Figure 4. The HRR can be calculated by the equation below [46]: where, ̇ represents the HRR, (kW); represents the combustion efficiency, (1); ̇ is the mass loss rate per unit area of fuel, (kg/(m 2 s); Δ is the effective heat of combustion The HRR can be calculated by the equation below [46]: Q represents the HRR, (kW); χ represents the combustion efficiency, (1); . m is the mass loss rate per unit area of fuel, (kg/(m 2 s); ∆H f is the effective heat of combustion of fuel, (kJ/kg); A f is the fuel plate area, (m 2 ). The mass loss rate per unit area of fuel . m can be calculated by dividing the maximum mass loss rate at the stable stage in Figure 5 by the fuel plate area. The mass loss rate per unit area of gasoline and HRR at the stable stage are listed in Table 5.  of fuel, (kJ/kg); is the fuel plate area, (m 2 ). The mass loss rate per unit area of fuel ̇ can be calculated by dividing the maximum mass loss rate at the stable stage in Figure 5 by the fuel plate area. The mass loss rate per unit area of gasoline and HRR at the stable stage are listed in Table 5.

Smoke Release Rate
There are many factors influencing the smoke release rate, such as fire source size, type of fuel and ventilation conditions around the fire source. In this study, the smoke diffusion volume and smoke opacity is adopted as the criteria of smoke release rate. Considering the results in Sichuan and Tibet as examples, the corresponding field pictures of the three stages are shown in Figure 7. The smoke release rate decreases in high-altitude areas regardless of which stage because of the sharp decrease in HRR and longer burning time.

Smoke Release Rate
There are many factors influencing the smoke release rate, such as fire source size, type of fuel and ventilation conditions around the fire source. In this study, the smoke diffusion volume and smoke opacity is adopted as the criteria of smoke release rate. Considering the results in Sichuan and Tibet as examples, the corresponding field pictures of the three stages are shown in Figure 7. The smoke release rate decreases in high-altitude areas regardless of which stage because of the sharp decrease in HRR and longer burning time.

Discussion
As shown in Figure 3, the mass of gasoline decreases during the burning process. Burning out 0.22 kg of gasoline in Tibet requires approximately 330 s, while it only requires approximately 140 s in Sichuan and 230 s in Yunnan for the same quality of gasoline. Therefore, ambient pressure has a great impact on the burning duration. Specifically, the time duration for burning the same amount of gasoline increases with reduced ambient pressure due to lower oxygen content, hindering the burning process.
From the mass variation rate per unit area of gasoline presented in Figure 4 and the time duration of different stages listed in Table 4, the characteristics of the three stages are summarized as follows: 1. Initial stage: the mass variation rate per unit area of gasoline increases gradually at this stage. In Sichuan, the mass variation rate per unit area of gasoline increases from 0

Discussion
As shown in Figure 3, the mass of gasoline decreases during the burning process. Burning out 0.22 kg of gasoline in Tibet requires approximately 330 s, while it only requires approximately 140 s in Sichuan and 230 s in Yunnan for the same quality of gasoline. Therefore, ambient pressure has a great impact on the burning duration. Specifically, the time duration for burning the same amount of gasoline increases with reduced ambient pressure due to lower oxygen content, hindering the burning process.
From the mass variation rate per unit area of gasoline presented in Figure 4 and the time duration of different stages listed in Table 4, the characteristics of the three stages are summarized as follows: 1. Initial stage: the mass variation rate per unit area of gasoline increases gradually at this stage. In Sichuan, the mass variation rate per unit area of gasoline increases from 0 kg/(m 2 s) to 0.019 kg/(m 2 s) with a duration of approximately 10 s. In Tibet, the Energies 2021, 14, 4627 9 of 12 final value of mass variation rate per unit area at this stage decreases to approximately 0.0092 kg/(m 2 s) compared with that in Sichuan, and the time duration increases to 25 s due to lower oxygen content. The final mass variation rate per unit area and time duration at this stage in Yunnan is between that in Sichuan and Tibet; 2. Stable stage: this stage is the longest, and the mass variation rate per unit area of gasoline fluctuates. In Sichuan, this stage lasts for 126 s, and the mass variation rate per unit area ranges from 0.0075 kg/(m 2 s) to 0.029 kg/(m 2 s). In Tibet, the time duration at this stage is larger than that in Sichuan, and the mass variation rate per unit area is smaller due to lower oxygen content. The mass variation rate per unit area and time duration in Yunnan is between that in Sichuan and Tibet; 3. Attenuation stage: the fuel is burning out at this stage, and the mass variation rate per unit area of gasoline is decreasing gradually. In Sichuan, the mass variation rate per unit area decreases from 0.011 kg/(m 2 s) to 0 kg/(m 2 s), and this stage lasts for approximately 7 s. As for Tibet, the mass variation rate per unit area decreases from 0.005 kg/(m 2 s) to 0 kg/(m 2 s), and the time duration is approximately 15 s. The results in Yunnan are between that in Sichuan and Tibet.
After the Karman filtering method, the mass loss rate per unit area and HRR were obtained for the burning experiments in three sites, as shown in Table 5. In Sichuan, the fuel mass loss rate per unit area at the stable stage is approximately 0.0125 kg/(m 2 s), and the HRR is 68 kW. The fuel mass loss rate per unit area at the stable stage and HRR are 0.009 kg/(m 2 s) and 49 kW in Yunnan, respectively; meanwhile, they decrease to 0.0063 kg/(m 2 s) and 34 kW in Tibet, respectively. Under the experimental conditions with 0.22 kg of No. 93 gasoline burning in a plate of 30 cm × 40 cm, the relationship between the ambient pressure and HRR follows the equation below: where, .
Q represents the HRR (kW); k represents an experiment constant that can be determined through experiment results; p is the ambient pressure (kPa) and α is 1.3.
In this experiment, k equals 0.172 as determined by non-linear regression. Figure 8 shows the relationship between ambient pressure and HRR. The experiment results correspond well with Equation (3). It can be concluded that the HRR decreases with the increase in ambient pressure due to lower oxygen content.
Energies 2021, 14, x FOR PEER REVIEW 9 of 12 kg/(m 2 s) to 0.019 kg/(m 2 s) with a duration of approximately 10 s. In Tibet, the final value of mass variation rate per unit area at this stage decreases to approximately 0.0092 kg/(m 2 s) compared with that in Sichuan, and the time duration increases to 25 s due to lower oxygen content. The final mass variation rate per unit area and time duration at this stage in Yunnan is between that in Sichuan and Tibet; 2. Stable stage: this stage is the longest, and the mass variation rate per unit area of gasoline fluctuates. In Sichuan, this stage lasts for 126 s, and the mass variation rate per unit area ranges from 0.0075 kg/(m 2 s) to 0.029 kg/(m 2 s). In Tibet, the time duration at this stage is larger than that in Sichuan, and the mass variation rate per unit area is smaller due to lower oxygen content. The mass variation rate per unit area and time duration in Yunnan is between that in Sichuan and Tibet; 3. Attenuation stage: the fuel is burning out at this stage, and the mass variation rate per unit area of gasoline is decreasing gradually. In Sichuan, the mass variation rate per unit area decreases from 0.011 kg/(m 2 s) to 0 kg/(m 2 s), and this stage lasts for approximately 7 s. As for Tibet, the mass variation rate per unit area decreases from 0.005 kg/(m 2 s) to 0 kg/(m 2 s), and the time duration is approximately 15 s. The results in Yunnan are between that in Sichuan and Tibet.
After the Karman filtering method, the mass loss rate per unit area and HRR were obtained for the burning experiments in three sites, as shown in Table 5. In Sichuan, the fuel mass loss rate per unit area at the stable stage is approximately 0.0125 kg/(m 2 s), and the HRR is 68 kW. The fuel mass loss rate per unit area at the stable stage and HRR are 0.009 kg/(m 2 s) and 49 kW in Yunnan, respectively; meanwhile, they decrease to 0.0063 kg/(m 2 s) and 34 kW in Tibet, respectively. Under the experimental conditions with 0.22 kg of No. 93 gasoline burning in a plate of 30 cm × 40 cm, the relationship between the ambient pressure and HRR follows the equation below: where, ̇ represents the HRR (kW); represents an experiment constant that can be determined through experiment results; is the ambient pressure (kPa) and is 1.3. In this experiment, equals 0.172 as determined by non-linear regression. Figure 8 shows the relationship between ambient pressure and HRR. The experiment results correspond well with Equation (3). It can be concluded that the HRR decreases with the increase in ambient pressure due to lower oxygen content. Apart from the decrease in HRR, higher flame height is observed in high-altitude areas during the burning process. The highest flame occurred in Tibet among the three experimental sites. The higher flame height allows the burning fuel to absorb more oxygen and complete the combustion reaction due to lower oxygen content in high-altitude areas. Flame height is an important sign in the initial stage of a fire, and a higher flame is easier to detect and identify the existence of a fire, thereby reducing economic losses and casualties.
Smoke is the leading cause of death for stranded people in a fire, while the toxicity of the smoke is related to the smoke release rate. When the smoke release rate is low, it is more likely for people to survive. According to the experimental results in this paper, the smoke release rate decreases with the increase in the altitude due to the lower HRR, which means the severity of a fire in high-altitude areas is lower compared with plain areas.
In summary, it is easier to detect a fire in high-altitude areas. This finding will earn more time for authorities to handle fires and decrease economic losses and casualties. Additionally, the smoke release rate is small in high-altitude areas due to lower HRR, which means that the possibility for people surviving a fire increases.

Conclusions
In this paper, a series of field experiments in plain and high-altitude areas were conducted to explore the effects of ambient pressure on burning characteristics of fuel, including HRR, flame height, and smoke release rate. The major findings are as follows: 1. The whole combustion process can be divided into three stages based on trends of the mass variation rate of gasoline in this experiment: initial stage, stable stage, and attenuation stage. The initial stage has a gradually increasing mass loss rate. The stable stage is the longest stage with a fluctuating mass loss rate. The attenuation stage has a gradually decreasing mass loss rate. The time duration of all three stages increases with the increase in altitude due to lower oxygen content in high-altitude areas; 2. The HRR decreases with the decrease in ambient pressure due to lower oxygen content, while ambient pressure and HRR follow an exponential relation; 3. The flame height increases to absorb more oxygen and complete the combustion reaction in high-altitude areas due to lower oxygen content. The smoke release rate is less due to the sharp decrease in HRR and lower oxygen content in high-altitude areas.