Journal of Hazardous Materials 403 (2021) 123843

Contents lists available at ScienceDirect

Journal of Hazardous Materials

journal homepage: www.elsevier.com/locate/jhazmat

Effects of N2 and CO2 dilution on the explosion behavior of liquefied

petroleum gas (LPG)-air mixtures
Zhenmin Luo a, b, c, *, Chengcai Wei a, Tao Wang d, **, Bin Su a, Fangming Cheng a, b, c,
Changchun Liu a, b, c, Yachao Wang a, b, c
a
School of Safety Science & Engineering, Xi’an University of Science and Technology, 58, Yanta Mid. Rd., Xi’an, 710054, Shaanxi, PR China
b
Shaanxi Key Laboratory of Prevention and Control of Coal Fire, 58, Yanta Mid. Rd, Xi’an, 710054, Shaanxi, PR China
c
Shaanxi Engineering Research Center for Industrial Process Safety & Emergency Rescue, 58, Yanta Mid. Rd., Xi’an, 710054, Shaanxi, PR China
d
Postdoctoral Program, Xi’an University of Science and Technology, 58, Yanta Mid. Rd., Xi’an, 710054, Shaanxi, PR China

A R T I C L E I N F O A B S T R A C T

Editor: G. Lyberatos The present work aims to provide theoretical support to prevent LPG explosion accidents and reduce its haz­
ardous effect. The explosion of LPG-air mixtures (4.0 % – 9.0 % in volume) under N2 and CO2 dilution (0.0%–
Keywords: 30.0% in volume) was experimentally investigated by using a 20-L vessel at ambient temperature and pressure.
Liquefied petroleum gas The parameters of the maximum explosion pressure (pmax ), time to reach the maximum explosion pressure (tc ),
Explosion pressure characteristic
maximum rate of the pressure rise ((dp/dt)max ), time to reach the maximum rate of the pressure rise (tb ), and
Flame propagation velocity
average velocity of flame propagation (v) were investigated. In addition, the inerting efficiencies of N2 and CO2
Inerting efficiency
were compared and analyzed. The results show that there exists a cubic function relation among pmax , (dp/dt)max
x, tc , tb and LPG volume fraction during the LPG explosion process. At the volume fraction of LPG was 5.7 %, all
parameters reached peak values. The addition of N2 and CO2 effectively suppressed the explosion of LPG,
reduced pmax , (dp/dt)max , and v and extended tc and tb . At a low LPG volume fraction, the inerting efficiency of
CO2 was higher than that of N2. At a high LPG volume fraction, the inerting efficiency of N2 was higher than that
of CO2.

1. Introduction Lee et al., 2012; Chen et al., 2019b). On the other hand, to prevent the
accidental release of energy and reduce the destructiveness of explo­
With the rapid growth of the global economy, the human demand for sions, researchers are considering how to reduce the explosion intensity.
energy is increasing (Yao et al., 2019; Gaies et al., 2019). As a fossil fuel, As a common combustible gas, LPG has explosive characteristics similar
liquefied petroleum gas (LPG) is clean, efficient and inexpensive and will to those of other combustible gases. Gas explosions usually occur in a
be of great importance in the future (Melikoglu, 2014; Troncoso et al., relatively enclosed space (Shebeko et al., 1995; Zhang et al., 2017);
2019; Sharma et al., 2019). However, due to unsafe production, storage, therefore, related experiments have been performed in pipelines (Huo
transportation and usage factors, LPG explosions occur frequently and Chow, 2017; Razus et al., 2009; Luo et al., 2020d) and storage tanks
(Turgut et al., 2013; Kumar, 2013; Leal and Santiago, 2004; Stawczyk, (Tschirschwitz et al., 2018; Shebeko et al., 2000). To characterize the
2003). Therefore, to prevent the occurrence of LPG explosion accidents LPG explosion properties, related parameters have been studied, such as
and reduce the damage degree of explosions, it is necessary to study the the explosion limit (Luo et al., 2017, 2019a; Wang et al., 2020b), limit
LPG explosion characteristics and inerting regularity. oxygen concentration (Van den Schoor et al., 2009; Su et al., 2020a),
On the one hand, LPG, as a fuel, is expected to enhance its com­ explosion pressure (Wang et al., 2017; Su et al., 2020b), explosion index
bustion performance (such as adding dimethyl ether (DME)) so that LPG (Luo et al., 2019b), and flame propagation velocity (Huzayyin et al.,
can release more energy in the reaction process and improve its utili­ 2008; Luo et al., 2020a). The explosion characteristics of combustible
zation rate (Bariha et al., 2016; Chen et al., 2019a; Zhang et al., 2019; gases are related to a variety of external factors. Previous studies have

* Corresponding author at: School of Safety Science & Engineering, Xi’an University of Science and Technology, 58, Yanta Mid. Rd., Xi’an, 710054, Shaanxi, PR
China.
** Corresponding author.
E-mail addresses: [email protected] (Z. Luo), [email protected] (T. Wang).

https://doi.org/10.1016/j.jhazmat.2020.123843
Received 17 May 2020; Received in revised form 7 August 2020; Accepted 25 August 2020
Available online 2 September 2020
0304-3894/© 2020 Elsevier B.V. All rights reserved.
Z. Luo et al. Journal of Hazardous Materials 403 (2021) 123843

examined the influences of the external temperature (Li et al., 2019a), 2. Experimental
external pressure (Huang et al., 2019), combustible gas concentration
(Wang et al., 2017), and additional ignition energy (Wang et al., 2020a) 2.1. Experimental system and operating procedures
on the explosion characteristics. Generally, the higher the external
temperature and pressure are, the higher the ignition energy is; hence, The experimental system is mainly composed of an explosion reac­
the closer the volume fraction of combustible oxygen is to the chemical tion chamber, an automatic control unit, a gas distribution unit, an
equivalent ratio, the stronger the explosion reaction is. Shimao Wang ignition device, a powder spray device, a data acquisition system, a high-
et al. (Wang et al., 2020a) tested gasoline-air mixtures in a 907.5-L tank speed camera, and a main control computer (as shown in Fig. 1). The
and studied the effects of the concentration, temperature, ignition en­ explosion reaction chamber is a 20-L spherical tank consisting of a
ergy and humidity on the explosion overpressure. Their results showed double-layer stainless steel structure, with a pressure resistance of 4
that the maximum explosion pressure (pmax ) and maximum rate of the MPa. Before each experiment, the vacuum pump valve needs to be
pressure rise ((dp/dt)max ) first increase and then decrease with opened to start vacuuming. When the pressure in the spherical tank is
increasing gasoline concentration, and with increasing ignition energy, lower than 30.0 kPa, the vacuum pump is deactivated. Since a small
pmax and (dp/dt)max exhibit linear and logarithmic increase trends, amount of water vapor will be generated during an explosion, it is
respectively. With increasing initial humidity, pmax and (dp/dt)max show necessary to vacuum at least 3 times after each explosion experiment
a linear downward trend. until no water mist is observed through the glass window. The working
In the aspect of inerting, scholars aim to develop a kind of clean, principle of the gas distribution system is the partial pressure ratio
environmentally friendly inert medium with a good inerting effect and a method, and the gas distribution accuracy is 0.1 %. The system auto­
low cost. At present, common inert media are dry powder (Luo et al., matically adds a gas sample according to the set intake volume. When
2016; Zhou et al., 2019), water (Shirvill, 2004; Gan et al., 2018), inert the gas sample volume is completely added, the system opens the air
gas, etc. N2 and CO2, as the most commonly used inert gases, attain good valve and automatically provides air. After the sample gas and air vol­
inerting effects (Hamidi and Ilminnafik, 2014; Nair et al., 2016; Mishra umes are completely added, they are stirred for 300 s so that the sample
and Kumar, 2010; Kumar and Mishra, 2008). However, the inerting gas and air are evenly mixed. The ignition device adopts high-voltage
effects of N2 and CO2 are related to the types of combustible gases. In pulse ignition, and the ignition energy is 1 J. When the set ignition
previous research, the inerting mechanism of N2 and CO2 for delay time is reached, the system will automatically ignite the gas
single-component gases (such as methane) has been extensively studied, mixture. The data acquisition system mainly includes high-precision
and remarkable research results have been achieved (Li et al., 2019b; pressure sensors and high-speed cameras. The measurement range of
Cui et al., 2018; Zeng et al., 2015). For example, ZR Wang et al. (Wang the pressure sensor is − 0.1 to 2 MPa, the resolution is 0.001 MPa, and
et al. (2014) performed a series of experiments in a closed container to the sampling interval is 0.2 ms. The high-speed camera was designed
analyze the effect of N2/CO2 on the explosion intensity, limiting oxygen and produced by the American Vision Research company, and its highest
concentration, and explosion limit of CH4-air mixtures. Their results shooting rate is 600000 frames per second and the minimum exposure
indicated that the explosion intensity decreases with increasing N2/CO2 time is 1 μs. Zhenmin Luo et al. (Luo et al. (2019b) has also described the
volume fraction, the limiting oxygen concentration decreases linearly details of this experimental equipment in previous studies. In this
with increasing N2 volume fraction, and the explosion limit varies lin­ experiment, the shooting frequency is set to 2000 frames per second, the
early with the N2/CO2 volume fraction in the mixture, while the higher diameter range of the circular window is 0~110 mm, and the resolution
the CO2 volume fraction in the N2/CO2 mixture is, the better the inerting of the images is 512 × 512. All tests were performed in an environment
effect is. at a temperature of 15 ◦ C–25 ◦ C and a relative humidity ranging from 30
At present, there are few studies on hydrocarbon mixtures above C2. % to 65 %.
It remains unclear how a fuel decomposes when it reacts, which
important components exist in the reaction system, what the main re­ 2.2. Experimental materials
action path is, and what the key steps are in the reaction process. LPG, as
a mixed gas, mainly consists of propane and butane, and the reaction The gas selected in this experiment is LPG, which is commonly used
process and path are complex. Theoretically, it is difficult to quantita­ in industrial production (provided by the Xi’an Hengli Gas Co., Ltd.,
tively analyze the influence of inert gas on its inerting efficiency. Shaanxi Province, China). The main components are as follows: 5.0 %
Therefore, in this study, LPG was selected as the research object, and hydrogen, 5.0 % methane, 4.0 % ethane, 73.0 % propane, 4.0 % butane,
N2 and CO2 were selected as inert media, while the explosion pressure propylene, butane and hydrocarbon gas containing more than 5 carbon
characteristics and inerting regulations of LPG were studied in this atoms (accounting for approximately 9% of the composition). In this
paper. First, the explosion pressure characteristics and flame propaga­ experiment, LPG mixtures with volume fractions of 4.0 %, 5.0 %, 6.0 %,
tion velocity parameters of LPG with different volume fractions in air 7.0 %, 8.0 % and 9.0 % were tested, and the optimal explosive volume
were measured by using a visual spherical gas and dust explosion sup­ fraction was measured by the gradual approximation method. In the
pression comprehensive experimental system. The explosion charac­ inerting research part, two commonly used inert gases, N2 and CO2, were
teristics included five parameters: maximum explosion pressure, selected. First, N2 at volume fractions of 5.0 %, 10.0 %, 15.0 %, 20.0 %,
maximum rate of the pressure rise, time to reach the maximum explo­ 25.0 % and 30.0 % was selected to test its inerting efficiency for LPG
sion pressure, time to reach the maximum rate of the pressure rise and mixtures at volume fractions of 4.0 %, 5.0 %, 6.0 %, 7.0 %, 8.0 % and 9.0
flame propagation velocity. Hence, the characteristics of the explosion %, respectively. Then, N2 was replaced by CO2, and the explosion
pressure and flame propagation velocity of LPG/N2 and LPG/CO2 mix­ pressure parameters and flame propagation velocity of the LPG/CO2
tures at different ratios were tested. Finally, the suppression effects of N2 mixtures were tested under the same experimental conditions. The pu­
and CO2 on LPG explosion were comprehensively compared and rity of N2 and CO2 used in the experiment both reached 99.9 %.
analyzed under the same conditions. The explosion characteristic pa­
rameters and inerting regulations of LPG obtained in this experiment can 3. Results and analysis
provide a theoretical reference for the design of LPG inerting explosion-
proof media with a great reference value for enterprise managers to 3.1. Characteristics of the explosion pressure
formulate safety management measures, assess engineering projects and
formulate emergency rescue plans. To characterize the explosion pressure characteristics of LPG, the
studied explosion pressure characteristics included four parameters:
maximum explosion pressure (pmax ), time to reach the maximum

2
Z. Luo et al. Journal of Hazardous Materials 403 (2021) 123843

Fig. 1. Schematic diagram of a 20L spherical explosion test device.

explosion pressure (tc ), maximum rate of the pressure rise ((dp/dt)max ), According to Fig. 3, with increasing LPG volume fraction, pmax and (dp/
and time to reach the maximum rate of the pressure rise (tb ). The rela­ dt)max show a trend of first increasing and then decreasing, and the
tionship between these parameters is shown in Fig. 2. After the explosion corresponding tc and tb values exhibit a trend of first declining and then
reaction starts, a chain reaction is initiated to produce free radicals, rising.
which in turn leads to a new chain reaction. The rate of free radical At an LPG volume fraction of 5.7 %, pmax and (dp/dt)max reached their
generation is intensified, and the rate of chemical reactions gradually maximum values, tc and tb reached their minimum values, and the ex­
increases. With continuous chemical reactions, the concentrations of the plosion reaction at this volume fraction was the most sufficient. When
reactants and key free radicals in the explosion system decrease. At the the LPG volume fraction was between 5.0 % and 6.50 % (close to the
same time, heat conduction occurs in the reaction system and vessel optimal explosive volume fraction), the numerical variation range of the
walls, which consumes part of the heat energy, resulting in a decrease in characteristic parameters of the explosive pressure was not large, and
the chemical reaction rate and explosion pressure. The variation in LPG the explosion pressure fluctuated around 0.775 MPa. At an LPG volume
explosion pressure over time is similar to that of other hydrocarbon fraction of 4.0 % (close to the lower explosion limit) and 9.0 % (close to
gases, such as methane (Luo et al., 2017, 2019b). the upper explosion limit), pmax and (dp/dt)max were relatively low,
Fig. 3 shows a curve of the characteristic parameters of the LPG while tc and tb were relatively large. When the volume fraction of
explosion pressure with the volume fraction. The relationship between flammable gas was lower than the equivalence ratio, this is called the
the explosion pressure characteristic parameters and LPG volume frac­ oxygen-rich and fuel-lean state; when the volume fraction of flammable
tion was fitted with a polynomial function, and it was found that pmax , gas is higher than the equivalence ratio, this is called the fuel-rich and
(dp/dt)max , tc , tb and LPG volume fraction conformed to a cubic func­ oxygen-lean state (Sun and Li, 2017). In this paper, LPG volume frac­
tion, while the degree of fitting exceeded 0.95. Shimao Wang et al. tions between 4.0 % and 5.7 % can be regarded as the oxygen-rich and
(Wang et al., 2020a) studied the change rule of the overpressure of fuel-lean state, and the LPG volume fractions between 5.7 % and 9.0 %
gasoline-air mixture explosion in a 907.5-L fuel tank and reached a can be regarded as the fuel-rich and oxygen-lean state. Therefore, the
similar conclusion; in particular, the maximum explosion pressure and following conclusions can be drawn from the above variation rules: the
maximum rate of the pressure rise first increased and then decreased closer the LPG volume fraction is to the chemical equivalent ratio, the
with increasing gasoline concentration, and the mathematical relation­ higher pmax and (dp/dt)max are and the smaller the corresponding tc and
ship between the explosion pressure and gasoline concentration was tb values are. The reason is that when the LPG volume fraction is low, the
approximately in accordance with a cubic polynomial function. LPG/air mixture is in the oxygen-rich and fuel-lean state, and the O free
radicals are sufficient. With increasing LPG volume fraction, the con­
centration of combustibles increases, the concentration of H, OH and
other key free radicals in the chain reaction process increases, and the
explosion reaction is more sufficient. However, when the LPG volume
fraction is high, the LPG/air mixture is in the fuel-rich and oxygen-lean
state, the O2 concentration is low, the O free radicals are insufficient, the
chain reaction is weakened, and the quantity of H, OH and other key free
radicals decreases, and the reaction cannot be fully completed (Hu and
Wei, 2020).
Fig. 4 shows the change in the characteristic parameters of the ex­
plosion pressure after gradually increasing the N2 volume fraction. With
the constant increase in the N2 volume fraction, pmax and (dp/dt)max of
the LPG mixtures with different volume fractions all showed a down­
ward trend, while tc and tb both showed an upward trend. The addition
of N2 can reduce the concentration of combustibles and O2, absorb the
heat released in the reaction process, collide with the key free radicals
(O, H, and OH), reduce the activity of these free radicals, accelerate the
extinction of these free radicals, and decrease the chemical reaction rate.
Therefore, N2 can reduce pmax and (dp/dt)max of the LPG mixtures and
Fig. 2. Schematic diagrams of parameters. extend tc and tb . At an LPG volume fraction of 4.0 %, the volume fraction

3
Z. Luo et al. Journal of Hazardous Materials 403 (2021) 123843

Fig. 3. Relationship between LPG volume fraction and characteristic parameters of explosion pressure.

Fig. 4. Relationship between N2 volume fraction and characteristic parameters of explosion pressure.

4
Z. Luo et al. Journal of Hazardous Materials 403 (2021) 123843

of N2 gradually increased from 5.0% to 30.0%, and pmax decreased from explosion interval was exited. CO2 also has the function of dilution and
0.68 MPa to 0.46 MPa, a decrease of 32.4 %, while tc increased from 184 cooling, reduces the concentration of combustibles and O2 in the reac­
ms to 481.8 ms, an increase of 161.8 %. Moreover, (dp/dt)max decreased tion system, participates in partial chain reactions, such as CO2+H ⇄CO
from 39.1 MPa.s− 1 to 5.3 MPa.s− 1, a decrease of 86.4 %, and tb increased + OH, reduces the concentration of H radicals, and inhibits chemical
from 128.8 ms to 299.6 ms, an increase of 132.6 %. At an LPG volume reactions.
fraction of 9.0 %, N2 was added at a volume fraction of 15.0 %, and pmax
reached 0.131 MPa (close to atmospheric pressure). When the N2 vol­
ume fraction was increased to 20.0 %, LPG explosion was completely 3.2. Characteristics of flame propagation
suppressed (the explosion interval was exceeded).
Fig. 5 shows the relationship between the CO2 volume fraction and Fig. 6 shows a continuous picture acquired by the high-speed camera
the characteristic parameters of the explosion pressure. As the CO2 at the initial stage of the explosion reaction in which the LPG volume
volume fraction increased, pmax and (dp/dt)max gradually decreased, fraction is 6.0 %. As shown in Fig. 6, when ignition was started, the
while tc and tb gradually increased. The obtained inerting inhibition flame spread from the ignition center to the surroundings as a circle.
regulations were similar to those of N2. Among them, at an LPG volume During diffusion, the brightness increased. Considering the influence of
fraction of 6.0 %, pmax , (dp/dt)max x, tc and tb attained the largest change buoyancy on the vertical direction of flame propagation (Yongzhen and
range, where pmax declined from 0.66 MPa to 0.156 MPa, a decrease of Guitang, 2013), the trajectory and velocity of flame propagation along
76.36 %; tc increased from 185.2 ms to 1028.6 ms, an increase of 455.39 the horizontal direction were studied with the latter as the reference
%; (dp/dt)max decreased from 35.4 MPa.s− 1 to 0.95 MPa.s− 1, a decrease coordinate.
of 97.32 %; and tb increased from 122.2 ms to 640.6 ms, an increase of Based on the pictures acquired by the high-speed camera, the tra­
424.22 %. When the LPG volume fraction was 8.0 % and a CO2 volume jectory of the flame front was extracted with a canny operator combined
fraction of 30.0 % was added, the LPG mixture was completely inerted. with an algorithm. At LPG volume fractions of 4.0 % (close to the lower
When the volume fraction of LPG was 9.0 % and a CO2 volume fraction explosion limit) and 9.0 % (close to the upper explosion limit), the ex­
of 25.0 % was added, the LPG mixture was completely inerted, and the plosion intensity was low, the flame brightness was extremely low, and
the edges of the image could not be effectively identified. Therefore, this

Fig. 5. Relationship between CO2 volume fraction and characteristic parameters of explosion pressure.

5
Z. Luo et al. Journal of Hazardous Materials 403 (2021) 123843

Fig. 6. 6.0 % LPG-air explosion flame picture.

paper only shows the curve of the flame front trajectory over time in the to the effect of the temperature difference, the less heat the flame
LPG explosion process at volume fractions of 5.0 %, 6.0 %, 7.0 % and 8.0 transfers to the unburned area via heat conduction and molecular
% (Fig. 7). In Fig. 7, the trajectory of the flame front linearly increased diffusion, the less the combustion reaction can be fully completed and
over time. Manhou Li et al. (Li et al., 2019b) studied the characteristics the lower flame propagation velocity is. Therefore, in Fig. 7, compared
of methane flame propagation in a 500 × 500 × 500-mm3 cubic to the other operating conditions, the flame propagation velocity was
container; they also reported a similar change rule in that the flame the lowest at an LPG volume fraction of 8.0 %.
propagation velocity could nearly be regarded as a constant at the initial Fig. 7 shows that the trajectory of the flame front increased
stage of the explosion reaction. Fig. 7 reveals that the flame front tra­ approximately linearly over time. Therefore, the flame propagation
jectory was smaller than that of the previous moment at a certain time, velocity can be approximately regarded as a constant. To facilitate the
which may have occurred because the flame was disturbed by the air study of the influence of N2 and CO2 on the flame propagation velocity,
flow during propagation, and the flame front shape was slightly the average propagation velocity (v) is selected as a characteristic
deformed. The closer the fuel and O2 ratio is to the equivalent concen­ parameter, and this parameter is studied.
tration ratio, the more fully the chemical reaction can be completed, the Fig. 8 shows the change trend of v after gradually adding N2 to LPG.
more chemical energy is released, the more heat is fed back to the When the volume fraction of N2 increased from 5.0% to 30.0%, v of the
mixing system by combustion, and the greater the flame can be accel­ LPG mixture at volume fractions of 4.0 %, 5.0 % and 6.0 % decreased
erated. Therefore, as shown in Fig. 7, at an LPG volume fraction of 6.0 %, 72.45 %, 87.06 % and 95.42 %, respectively. At an LPG volume fraction
the flame propagation velocity was the highest. In addition, the higher of 7.0 %, the addition of 30 % N2 made the LPG mixture completely
the LPG volume fraction is, the higher the ratio of carbon to oxygen in inert; when the LPG volume fraction was 8.0 % and 9.0 %, the addition
the reaction system is and the more carbon smoke particles are produced of 25.0 % N2 drove the LPG mixture out of the explosion interval. During
in the reaction process. Because of the radiation of these carbon smoke the entire inerting process, v of the LPG mixture with volume fractions of
particles, the total radiation damage capacity of the reaction system 7.0 %, 8.0 %, and 9.0 % decreased 95.08 %, 82.08 %, and 72.63 %,
increases, resulting in a decrease in the average flame temperature. Due respectively. Theoretically, N2 does not participate in the chemical

Fig. 7. Curve of the trajectory of flame front with time. Fig. 8. Influence of N2 on average flame propagation velocity of LPG.

6
Z. Luo et al. Journal of Hazardous Materials 403 (2021) 123843

reaction, but as a stable gas molecule, it participates in third-order parameters for the LPG/N2 and LPG/CO2 gas mixtures with the LPG
collision reactions, such as H + O2+M⇄HO2+M, which reduces the volume fraction when the volume fraction of inert gas is 10.0 % and 15.0
concentration of H radicals in the reaction system (Zhang et al., 2012), %. Fig. 10 reveals that after adding N2 and CO2 to LPG, pmax and (dp/
and the chemical reaction rate continuously decreases, resulting in a dt)max decreased and tc and tb increased. When the LPG volume fraction
decrease in flame propagation velocity. was between 4.0 % and 6.0 % (the oxygen-rich and fuel-lean state), the
Fig. 9 shows the change rule of v after adding CO2 to LPG. When CO2 inerting efficiency of CO2 on the characteristic parameters of the LPG
was added to the various LPG mixtures with different volume fractions, v explosion pressure was higher than that of N2 for the same amount of
notably decreased. When the LPG volume fraction was 4.0 %, 7.0 %, and additives, while when the LPG volume fraction was between 7.0 % and
8.0 %, the decreases in v were almost equal, decreasing 77.52 %, 76.63 9.0 % (the oxygen- lean and fuel-rich state), the effect was the opposite.
%, and 71.17 %, respectively. At an LPG volume fraction of 5.0 %, the Therefore, it can be concluded from the above analysis that the CO2
CO2 volume fraction was gradually increased from 5.0% to 30.0%, and v inerting effect is higher when the LPG volume fraction is between 4.0 %
decreased from 0.516 m/s to 0.035 m/s, a decrease of 93.28 %. At an and 6.0 %. However, when the LPG volume fraction is between 7.0 %
LPG volume fraction of 6.0 %, the CO2 volume fraction was gradually and 9.0 %, the inerting effect of N2 is higher. Moreover, through the
increased from 5.0% to 30.0%, and v declined from 0.414 m/s to 0.028 comparative analysis, it is found that when the volume fraction of LPG is
m/s, a decrease of 93.24 %. CO2 participates in partial chain reactions, 4.0 % (close to the lower explosion limit) and 9.0 % (close to the upper
such as CO + OH⇄CO2+H, which is the main reaction path for CO to explosion limit), the inerting effects of N2 and CO2 are more profound.
change into CO2 and the key participation step of CO2 in chemical re­ Fig. 11 shows the effects of N2 and CO2 on the characteristic pa­
actions. The addition of CO2 reduces the reaction rate of CO and de­ rameters of the explosion pressure when the LPG volume fraction is 4.0
creases the generation of H radicals (Zeng et al., 2015), which decreases % and 9.0 %, respectively. At an LPG volume fraction of 4.0 % (close to
the heat of the chemical reaction, leading to a decrease in the flame the lower explosion limit), adding the same volume fraction of inert gas
propagation velocity. and keeping the other conditions unchanged, the inerting efficiency of
Both N2 and CO2 can effectively reduce the flame propagation ve­ CO2 was higher than that of N2. When adding 5.0 %, 10.0 %, 15.0 %,
locity. The addition of N2 and CO2 can reduce the concentration of 20.0 %, 25.0 % and 30.0 % inert gas, N2 reduced the pmax value of LPG by
combustibles and O2, increase the heat capacity at a constant volume of 7.94 %, 13.17 %, 14.60 %, 17.78 %, 19.84 % and 46.67 %, respectively,
the reaction system, and absorb the heat released in the reaction process, while CO2 reduced the pmax value by 10.79 %, 17.14 %, 30.48 %, 45.24
thus leading to a decrease in the heat transferred from the burned area to %, 56.51 % and 58.73 %, respectively. At an LPG volume fraction of 9.0
the unburned area, and the movement rate of the flame to the unburned %, the inerting efficiency of N2 was higher than that of CO2 for the same
area decreases. At an LPG volume fraction of 6.0 %, the concentration of volume fraction of inert gas added. Adding 5.0 %, 10.0 %, and 15.0 %
combustibles and O2 is close to the chemical equivalent ratio, and the inert gas, CO2 reduced the pmax value by 9.30 %, 20.72 %, and 40.80 %,
reaction can sufficiently proceed. Moreover, the concentration of O, OH respectively, while N2 reduced the pmax value by 17.5 %, 35.52 %, and
and H free radicals is high. With increasing N2 and CO2 volume frac­ 72.30 %, respectively. Moreover, at an LPG volume fraction of 9.0 %, the
tions, the concentration of the key free radicals in the reaction system addition of 15.0 % N2 drove the LPG mixture out of the explosion in­
sharply decreases, and the chemical reaction greatly decreases, resulting terval, while the addition of CO2 should at least exceed 20.0 %.
in a great impact on the flame propagation velocity. Fig. 12 shows the influence of N2 and CO2 at the same volume
fraction on v of the LPG mixture with different volume fractions (the N2
and CO2 volume fractions are 10.0 % and 15.0 %). After adding N2 and
3.3. Analysis of the inerting efficiency for explosion suppression
CO2 to LPG, v decreased considerably, and it was lower for the LPG/N2
and LPG/CO2 mixtures than for the LPG/air mixture. When the LPG
Both N2 and CO2 can suppress the explosion of LPG, reduce pmax ,
volume fraction was between 4.0 % and 6.0 % (the oxygen-rich and fuel-
(dp/dt)max , and v of LPG explosion and increase tc and tb . To further
lean state), v of the LPG/CO2 mixture was lower than that of the LPG/N2
analyze the inerting effects of N2 and CO2, the principle of the control
mixture. When the LPG volume fraction was between 7.0 % and 9.0 %
variable method is applied to maintain the inert gas volume fraction
(the fuel-rich and oxygen-lean state), v of the LPG/CO2 mixture was
unchanged. With the addition of 10.0 % and 15.0 % N2 and CO2 as an
higher than that of the LPG/N2 mixture. Moreover, through the
example, the inerting effects are compared and analyzed.
comparative analysis, it was found that when the LPG volume fraction
Fig. 10 shows a curve of the explosion pressure characteristic
was 4.0 % (close to the lower explosion limit) and 9.0 % (close to the
upper explosion limit), the inerting difference between these gases was
the largest.
Fig. 13 shows the influence of N2 and CO2 at different volume frac­
tions on v of the LPG mixture with the same volume fraction. At an LPG
volume fraction of 4.0 %, the inerting efficiency of CO2 was higher than
that of N2. When the volume fraction of inert gas was 5.0 %, 10.0 %, 15.0
%, 20.0 %, 25.0 % and 30.0 %, the v values of the LPG/CO2 and LPG/N2
mixtures differed by 0.141 m/s, 0.114 m/s, 0.057 m/s, 0.033 m/s, 0.047
m/s and 0.045 m/s, respectively. At an LPG volume fraction of 9.0 %,
the inerting efficiency of N2 was higher than that of CO2. When the inert
gas volume fraction was 5.0 %, 10.0 %, 15.0 % and 20.0 %, the v values
of the LPG/N2 and LPG/CO2 mixtures differed by 0.018 m/s, 0.009 m/s,
0.027 m/s and 0.034 m/s, respectively. Therefore, according to the
above analysis, under the condition of the oxygen-rich and fuel-lean
state, the inerting effect of CO2 is higher, while the inerting effect of
N2 is higher under the condition of the fuel-rich and oxygen-lean state.

3.4. Discussion

3.4.1. Theoretical analysis

Fig. 9. Influence of CO2 on average flame propagation velocity of LPG. An LPG explosion produces strong pressure shock waves and high-

7
Z. Luo et al. Journal of Hazardous Materials 403 (2021) 123843

Fig. 10. Effect of N2 and CO2 on LPG explosion characteristic parameters (both volume fractions of N2 and CO2 are 10.0 % and 15.0 %).

temperature combustion waves. Three areas result, namely, the area fluctuates. These fluctuations enhance the transferability of momentum,
that is unaffected by the shock wave, the area that is compressed by the energy, and components. The mass transfer coefficient increases because
shock wave between the shock wave and the combustion wave, and the of fluid mixing that is caused by turbulence disturbance. According to
area where the chemical gas explosion reaction occurs. The state pa­ Newton’s law of cooling in convective heat transfer:
rameters, such as the pressure, temperature, and density of these three
mA = hD (ρA,w + ρA,∞ ) (2)
regions, differ. During the initial explosion stage, the flame propagation
velocity is small, and mixed gas flows in a laminar flow model. In this
where mA is the mass flux (kg/m2̂ .s), hD is the convective mass transfer
experimental device, when the pressure wave propagates to the wall of
the reaction vessel, reflection occurs, which results in flame compression coefficient (m/s), ρA,w is the mass concentration of component a (kg/m3̂)
and wrinkling, and the effective contact area between the flame front in the mainstream at the (a, w) interface, and ρA,∞ is the mass concen­
and the unburned gas increases. Flame combustion depends on heat tration of component A in the mainstream far from the interface
conduction and molecular diffusion to increase the temperature of the (kg/m3̂).
unburned mixed gas, cause a chemical reaction, and promote the com­ As the component transport capacity increases, the mass flux per unit
bustion wave to the unburned gas continuously. According to the ther­ area increases, which accelerates the combustion reaction. Therefore, in
mal conductivity formula: Fig. 2, the explosion pressure increases sharply after 100 ms. When the
volume fraction of LPG is close to the equivalent ratio, the fuel reacts
Q = qA (1)
completely with oxygen, and the free radical concentration is sufficient.
The turbulence that is induced by the explosion shock wave is more
where Q is the heat transferred per unit time (W), q is the heat flux
obvious. Therefore, the explosion reaction is more intense, which leads
(W/m2̂ ), and A is the heat transfer area (m2̂).
to the flame propagation velocity, explosion pressure, and increase in
The heat transferred from the flame to the unexplained area in­
explosion pressure reaching a maximum.
creases. Because of the pressure wave disturbance, the mixed gas
After the explosion, the combustion reaction releases heat, which
changes from a laminar to a turbulent state. The fluid movement in the
results in an increased reaction system temperature. Therefore, a non-
turbulent boundary layer is irregular, and the gas flow velocity
stationary temperature and concentration field exist in the reaction

8
Z. Luo et al. Journal of Hazardous Materials 403 (2021) 123843

Fig. 11. Influence of N2 and CO2 on LPG explosion characteristic parameters (the volume fraction of LPG is 4.0 % and 9.0 % respectively).

transferred from the flame to the unburned area increases, and the
combustion material that is transferred from the unburned area to the
flame increases, which results in a more severe explosion reaction.
However, in the oxygen-lean state, an increase in LPG concentration
results in a reduction in oxygen concentration, explosion pressure, and
flame-propagation velocity.
Both N2 and CO2 could suppress the explosion of LPG, reduce pmax ,
(dp/dt)max and v and extend tc and tb . The larger the amount of added
inert gas is, the more notable the inerting effect is. According to the
inerting mechanism of inert gases, the inerting effect of N2 and CO2 is
reflected in the following three aspects: First, the inert gas dilutes LPG,
reduces the concentration of combustibles and combustion aids, and
reduces the effective collision probability between the active molecules
in the reaction system. Second, the inert gas absorbs part of the heat,
which decreases the temperature of the reaction system, and according
to the Arrhenius equation:

k = Ae− Ea/RT
(3)

where k is the rate constant, R is the molar gas constant, t is the ther­
modynamic temperature, Ea is the apparent activation energy, and A is
Fig. 12. Effects of N2 and CO2 on average LPG flame propagation velocity (the
volume fractions of N2 and CO2 are 10.0 % and 15.0 %, respectively).
the frequency factor, the chemical reaction rate decreases. Third, ac­
cording to the chain reaction mechanism, N2 and CO2, as stable mole­
cules, participate in third-body collision reactions and collide with H, O,
space. An increase in environmental oxygen concentration yields a
OH and CH3 radicals, which decreases the activity of these free radicals,
sufficient chemical reaction and the combustion temperature increases
leading to a decrease in the concentration of the key free radicals in the
accordingly. Because of the temperature gradient in the system and the
reaction system and decreasing the chemical reaction rate. Fourth, N2
uneven distribution of fuel concentration in local space, the heat that is

9
Z. Luo et al. Journal of Hazardous Materials 403 (2021) 123843

Fig. 13. Comparison of inerting effect (LPG volume fractions are 4.0 % and 9.0 %, respectively).

and CO2 addition reduces the gas thermal conductivity and reduces the
m CVρ
heat transfer from the burned to the unburned area. When the amount of n= = (5)
M M
N2 and CO2 increases, the concentration of mixed gas increases, which
leads to a decrease of gas thermal conductivity and heat transfer from pM
the flame to the undeveloped region. The heat cannot maintain normal T= (6)
RCρ
flame propagation, the reaction cannot be carried out fully, and the
flame propagation rate and explosion pressure decrease. Therefore, the where p is the pressure; V is the volume; n is the quantity of the sub­
larger the volume fraction of added inert gas is, the more notable the stance; R is the Planck constant; m is the mass of the gas; M is the average
inhibition effect is. molar mass; ρ is the gas density; and C is the volume fraction of the gas.
According to Eq. (6), the temperature of a gas mixture is inversely
3.4.2. Dilution regime of N2 and CO2 on explosion and flame propagation proportional to the volume fraction and density of the gas and directly
of LPG–air mixtures proportional to the pressure and average molar mass.
For the different volume fractions of LPG, there are differences in the When the LPG volume fraction is between 4.0 % and 6.0 %, the
inerting effect between N2 and CO2, which is caused not only by the mixture is in the oxygen-rich and fuel-lean state, with a low content of
physical and chemical properties of the inert gas but also by the prop­ combustible gas and an adequate O2 content in the spherical tank. The
erties of the combustible gas itself. In the process of CO2 inerting, the combustible materials can all participate in the reaction and emit a large
thermodynamic effect is the decisive factor, the chemical kinetic effect is amount of heat. According to Eq. (6), with the other conditions un­
the secondary factor, and the thermal radiation effect has the least in­ changed, the temperature of the mixture is inversely proportional to the
fluence (Hu and Wei, 2020). In the process of chemical reaction, with volume fraction of the gas. The lower the volume fraction of LPG is, the
the generation of heat, the temperature of the reaction system increases, higher the temperature of the reaction system is. Moreover, according to
and the specific heat capacity of N2 and CO2 changes with the change in the data in Table 1, when the temperature is higher than 300 ◦ C, the heat
temperature. Heat capacity data at a constant volume are listed in capacity at a constant volume of CO2 is higher than that of N2, and the
Table 1. heat absorption capacity of CO2 is high. In addition, a large amount of
When the temperature is below 200 ◦ C, the heat capacity at a con­ CO2 and H2O is produced in the reaction process. When CO2 is added to
stant volume of N2 is higher than that of CO2, and the heat absorption the reaction system for inerting purposes, the CO2 content in the reac­
capacity of CO2 is higher than that of N2 when the temperature is higher tion system sharply increases, which is conducive to the positive reac­
than 300 ◦ C. tion of CO2+H⇄CO + OH and H + O2(+M) ⇄H2O(+M), which
According to the ideal gas state equation: decreases the generation of H free radicals, thus achieving a good in­
hibition effect (Santner et al., 2013). However, no N2 is generated in the
pV = nRT (4)
reaction process, and N2 does not participate in the chain reaction, so N2
does not have this advantage.
If the real gas compressibility is considered, Eq. (6) is modified to
obtain:
Table 1 pM
The heat capacity at constant volume of N2 and CO2 (Yongzhen and Guitang, T= (7)
ZRCρ
2013) (kJ/(kg.K)).
Vreal
temperature(℃) N2 CO2 where Z is the gas compressibility factor, Z = Videal . Z reflects the diffi­
0 0.742 0.626 culty of gas compression. Because N2 and CO 2 are non-polar molecules,
100 0.744 0.677 the SRK equation meets the calculation requirements as:
200 0.747 0.731
300 0.752 0.760 RT a(T)
400 0.760 0.794
p= − (8)
V − b V(V + b)
500 0.769 0.824
600 0.779 0.851

10
Z. Luo et al. Journal of Hazardous Materials 403 (2021) 123843

(RTc )2 so the LPG/N2 mixture is not easily ignited. In addition, the main
a(T) = 0.42748 × a(Tt ) (9) component of LPG is propane, and the ratio of carbon to oxygen in the
pc
reaction system is relatively high. When the LPG volume fraction is high,
RTc more carbon smoke particles are produced in the reaction process.
b = 0.08664 (10)
pc Under high-temperature conditions, the CO2+C = CO reaction occurs,
which consumes CO2 in the reaction system. Therefore, when the LPG
( )
a(Tt )0.5 = 1 + F 1 − T0.5
t (11) volume fraction is high, the inerting efficiency of N2 is higher than that
of CO2.
where a(Tt ) is a function of contrast temperature Tt . For non-polar
molecules: 4. Conclusions

F = 0.48 + 1.574ω − 0.176ω2 (12) This paper presents the pressure characteristics of LPG explosion, the
average velocity of flame propagation, and the inhibition effect of N2
where ω is the acentric factor.
and CO2 on LPG explosion by experimental methods. In addition, the
At room temperature and pressure, the compressibility factors of N2
inerting effects of N2 and CO2 are quantitatively compared and
and CO2 are close to 1, which can be considered as ideal gases. Z de­
analyzed. There are three main conclusions, as summarized below.
creases gradually with a constant increase in temperature and pressure.
This law conforms to the curve of compressibility factors with temper­
(1) The explosion pressure of LPG increases first and then decreases,
ature and pressure measured by Hougen and Watson (Atkins and de
and there exists a cubic function relationship between the ex­
Paula, 2006).
plosion pressure and volume fraction. At an LPG volume fraction
At the initial explosion stage, the reaction system is at ambient
of 5.7 %, the maximum explosion pressure, the maximum rate of
temperature and pressure. The compressibility factor Z remains almost
the pressure rise, and the average flame propagation velocity
constant, and the reaction system volume remains unchanged. Based on
reach maximum values, while the time of the maximum explosion
the calculation results, the compressibility factors of N2 and CO2 are
pressure and time of the maximum rate of the pressure rise reach
close to 0.95 when the reaction system temperature is below 200 ◦ C. The
minimum values.
specific heat capacity of N2 at constant volume is significantly greater
(2) After adding N2 and CO2 to LPG, the maximum explosion pres­
than that of CO2, which results in a stronger heat absorption capacity. At
sure, the maximum rate of the pressure rise and the average flame
the initial stage of the explosive reaction, the compressibility factor has
propagation velocity decrease, while the time of the maximum
no obvious influence on the results compared with the specific heat
explosion pressure and time of the maximum rate of the pressure
capacity at constant volume.
rise increase. The larger the amount added is, the more notable
For the continuous explosion reaction, the chemical reaction rate
the inerting effect will be.
increases sharply, the pressure and temperature increase violently, and
(3) When the LPG volume fraction is between 4.0 % and 6.0 % (the
the gas compressibility factor decreases. Because the flammable gas
oxygen-rich and fuel-lean state), the inerting efficiency of CO2 is
explosion in a constant-volume closed vessel is affected mainly by the
higher than that of N2. The opposite is true when the LPG volume
key free radicals that are generated in the explosion (Luo et al., 2020b, c;
fraction is between 7.0 % and 9.0 % (the oxygen-lean and fuel-
Su et al., 2020c), the gas compressibility factor has little effect on the
rich state).
results and does not affect the final conclusion.
When the LPG volume fraction is between 7.0 % and 9.0 %, the gas
CRediT authorship contribution statement
mixture is in the fuel-rich and oxygen-lean state, the fuel cannot fully
participate in the reaction, the chemical reaction is insufficient, and the
Zhenmin Luo: Conceptualization, Methodology, Visualization.
explosion pressure is low. However, the LPG volume fraction is rela­
Chengcai Wei: Formal analysis, Data curation, Visualization, Writing -
tively high. According to the relationship of Eq. (7), the temperature of
original draft. Tao Wang: Formal analysis, Writing - review & editing.
the whole mixture is relatively low. In addition, the heat capacity at a
Bin Su: Writing - review & editing. Fangming Cheng: Validation.
constant volume of N2 is higher than that of CO2, and the heat absorp­
Changchun Liu: Supervision. Yachao Wang: Supervision.
tion capacity of N2 is high. Since O2 is not sufficient in the fuel-rich and
oxygen-lean state, the concentration of the O, H, OH radicals generated
Declaration of Competing Interest
in the reaction process is low, and the chemical effect of CO2 is weak­
ened with decreasing concentration of the key free radicals, so the chain
The authors declare that they have no known competing financial
reaction involving CO2 does not play a decisive role. Moreover, insuf­
interests or personal relationships that could have appeared to influence
ficient oxygen and reaction occur when the LPG volume fraction is large.
the work reported in this paper.
An increase in LPG volume fraction will decrease the thermal conduc­
tivity. After N2 and CO2 addition with the same volume fraction, the
Acknowledgements
thermal conductivity of the mixed gas will be reduced further. During
combustion, the mass diffusion ability is weak, which leads to a decrease
This work has been supported by the National Natural Science
in material input of combustible and oxygen, and the combustion re­
Foundation of China (Grant No. 5167-4193), the Innovation Capability
action cannot be carried out normally. When the temperature is below
Support Program of Shaanxi (Grant No. 2020-TD-021), the National Key
200 ◦ C, the heat capacity at a constant N2 volume is higher than that of
Research and Development Plan (Grant No. 2017-YFC-0804702), the
CO2. Therefore, the N2 inerting effect is better than CO2.
Joint Foundation of Shaanxi (Grant No. 2019-JLM-9), China Post­
The results in Fig. 11(b) further prove the above viewpoint. At an
doctoral Science Foundation (Grant No. 2019-M-663780) and the
LPG volume fraction of 9.0 %, the mixture can be ignited by adding 20 %
Foundation of Shaanxi Educational Committee (Grant No. 20-JK-0775).
CO2, while the mixture exits the explosion interval by adding 20 % N2.
When the LPG volume fraction is high, N2 and CO2 can be added at the
References
same volume fraction. Before ignition, the temperature of the mixture is
close to room temperature. Moreover, the heat capacity at a constant Atkins, Peter, de Paula, Julio, 2006. Atkins’ Physical Chemistry, 8th ed. Oxford
volume of N2 is higher than that of CO2, and the heat absorption capacity University Press.
is high. However, the thermodynamic effect of the inert gas is dominant, Bariha, Nilambar, Mani Mishra, Indra, Chandra Srivastava, Vimal, 2016. Fire and
explosion hazard analysis during surface transport of liquefied petroleum gas (LPG):

11
Z. Luo et al. Journal of Hazardous Materials 403 (2021) 123843

A case study of LPG truck tanker accident in Kannur, Kerala, India. J. Loss Prev. Razus, Domnina, Brinzea, Venera, Mitu, Maria, Oancea, D., 2009. Explosion
Process Ind. 40, 449–460. characteristics of LPG-air mixtures in closed vessels. J. Hazard. Mater. 165 (1–3),
Chen, Yuying, Zhang, Qi, Li, Mingzhi, Yuan, Mengqi, Wu, Dejian, Qian, Xinming, 2019. 1248–1252.
Experimental study on explosion characteristics of DME-blended LPG mixtures in a Santner, Jeffrey, Dryer, Frederick L., Ju, Yiguang, 2013. The effects of water dilution on
closed vessel. Fuel 248, 232–240. hydrogen, syngas, and ethylene flames at elevated pressure. Proc. Combust. Inst. 34
Chen, Yuying, Qian, Xinming, Zhang, Qi, Fu, Liye, Yuan, Mengqi, 2019. Study on the (1), 719–726.
effects of initial pressure and temperature on the explosion characteristics of DME- Sharma, Ashutosh, Parikh, Jyoti, Singh, Chandrashekhar, 2019. Transition to LPG for
blended LPG mixtures in an obstructed confined pipeline. Fuel 257. https://doi.org/ cooking: a case study from two states of India. Energy Sustain. Dev. 51, 63–72.
10.1016/j.fuel.2019.116047. Shebeko, Yu.N., Smolin, I.M., Korolchenko, A.Ya., Shevchuk, A.P., Borodkin, A.N.,
Cui, Gan, Wang, Shun, Bi, Zhenxiao, Li, Zili, 2018. Minimum ignition energy for the Malkin, V.L., Simonov, O.A., Gurinovich, L.V., Popov, S.A., Kolosov, V.A.,
CH4/CO2/O2 system at low initial temperature. Fuel 233, 159–165. Smirnov, E.V., 1995. Some aspects of fire and explosion hazards of large LPG storage
Gaies, Brahim, Kaabia, Olfa, Ayadi, Rim, Guesmi, Khaled, Abid, Ilyes, 2019. Financial vessels. J. Loss Prev. Process Ind. 8 (3), 163–168.
development and energy consumption: Is the MENA region different? Energy Policy Shebeko, Yu.N., Bolodian, I.A., Filippov, V.N., Navzenya, V.Yu, Kostyuhin, A.K.,
135. https://doi.org/10.1016/j.enpol.2019.111000. Tokarev, P.M., Zamishevski, E.D., 2000. A study of the behaviour of a protected
Gan, Bo, Li, Bei, Jiang, Haipeng, Bi, Mingshu, Gao, Wei, 2018. Suppression of polymethyl vessel containing LPG during pool fire engulfment. J. Hazard. Mater. 77 (1–3),
methacrylate dust explosion by ultrafine water mist/additives. J. Hazard. Mater. 43–56.
351, 346–355. Shirvill, L.C., 2004. Efficacy of water spray protection against propane and butane jet
Hamidi, Nurkholis, Ilminnafik, Nasrul, 2014. Inert effects on flammability limits and fires impinging on LPG storage tanks. J. Loss Prev. Process Ind. 17 (2), 111–118.
flame propagation of LPG by CO2. Appl. Mech. Mater. 664, 226–230. Stawczyk, Jan, 2003. Experimental evaluation of LPG tank explosion hazards. J. Hazard.
Hu, Xianzhong, Wei, Hao, 2020. Experimental investigation of laminar flame speeds of Mater. 96 (2–3), 189–200.
propane in O2/CO2 atmosphere and kinetic simulation. Fuel 268. https://doi.org/ Su, Bin, Luo, Zhenmin, Wang, Tao, Zhang, Jiang, Cheng, Fangming, 2020. Experimental
10.1016/j.fuel.2020.117347. and principal component analysis studies on minimum oxygen concentration of
Huang, Lijuan, Pei, Shufeng, Wang, Yu, Zhang, Liang, Ren, Shaoran, Zhang, Zhe, methane explosion. Int. J. Hydrogen Energy 45 (21), 12225–12235.
Xiao, Yu, 2019. Assessment of flammability and explosion risks of natural gas-air Su, Bin, Luo, Zhenmin, Tao, Wang, et al., 2020. Coupling analysis of the flame emission
mixtures at high pressure and high temperature. Fuel 247, 47–56. spectra and explosion characteristics of CH4/C2H6/Air mixtures. Energy Fuels 34,
Huo, Y., Chow, W.K., 2017. Flame propagation of premixed liquefied petroleum gas 920–928. https://doi.org/10.1021/acs.energyfuels.9b03242.
explosion in a tube. Appl. Therm. Eng. 113, 891–901. Su, Bin, et al., 2020. Chemical kinetic behaviors at the chain initiation stage of CH4/H2/
Huzayyin, A.S., Moneib, H.A., Shehatta, M.S., Attia, A.M.A., 2008. Laminar burning air mixture. J. Hazard. Mater. https://doi.org/10.1016/j.jhazmat.2020.123680.
velocity and explosion index of LPG-air and propane-air mixtures. Fuel 87 (1), Sun, Zuo-Yu, Li, Guo-Xiu, 2017. Turbulence influence on explosion characteristics of
39–57. stoichiometric and rich hydrogen/air mixtures in a spherical closed vessel. Energy
Kumar, Pramod, 2013. Fire disaster following LPG tanker explosion at Chala in Kannur Convers. Manage. 149, 526–535.
(Kerala, India): August 27, 2012. Burns 39 (7), 1479–1487. Troncoso, Karin, Segurado, Patricia, Aguilar, Margarita, da Silva, Agnes Soares, 2019.
Kumar, P., Mishra, D.P., 2008. Characterization of bluff-body stabilized LPG jet diffusion Adoption of LPG for cooking in two rural communities of Chiapas, Mexico. Energy
flame with N2 dilution. Energy Convers. Manage. 49 (10), 2698–2703. Policy 133, 0301–4215.
Leal, Ce’sar A., Santiago, Gustavo F., 2004. Do tree belts increase risk of explosion for Tschirschwitz, Rico, Krentel, Daniel, Kluge, Martin, Askar, Enis, Habib, Karim,
LPG spheres? J. Loss Prev. Process Ind. 17 (3), 217–224. Kohlhoff, Harald, Krüger, Simone, Neumann, Patrick P., Storm, Sven-Uwe,
Lee, Tae Sung, Sung, Jae Yong, Park, Dal Jae, 2012. Experimental investigations on the Rudolph, Michael, Schoppa, André, Szczepaniak, Mariusz, 2018. Experimental
deflagration explosion characteristics of different DME-LPG mixtures. Fire Saf. J. 49, investigation of consequences of LPG vehicle tank failure under fire conditions.
62–66. J. Loss Prev. Process Ind. 56, 278–288.
Li, Qianqian, Yan, Zhiyu, Zhang, Yemiao, Wang, Liangchen, Liu, Hu, Huang, Zuohua, Turgut, Paki, Arif Gurel, M., Kadir Pekgokgoz, R., 2013. LPG explosion damage of a
2019. Experimental study on the explosion characteristics of methylcyclohexane/ reinforced concrete building: a case study in Sanliurfa, Turkey. Eng. Fail. Anal. 32,
toluene-air mixtures with methanol addition at elevated temperatures. Process. Saf. 220–235.
Environ. Prot. 132, 126–133. Van den Schoor, F., Norman, F., Vandermeiren, K., Verplaetsen, F., Berghmans, J., Van
Li, Manhou, Xu, Jingchao, Wang, Changjian, Wang, Baozhen, 2019b. Thermal and den Bulck, E., 2009. Flammability limits, limiting oxygen concentration and
kinetics mechanism of explosion mitigation of methane-air mixture by N2/CO2 in a minimum inert gas/combustible ratio of H2/CO/N2/air mixtures. Int. J. Hydrogen
closed compartment. Fuel 255. https://doi.org/10.1016/j.fuel.2019.115747. Energy 34 (4), 2069–2075.
Luo, Zhen-Min, Cheng, Fang-Ming, Wang, Tao, 2016. Suppressive effects of silicon Wang, Z.R., Ni, L., Liu, X., Jiang, J.C., Wang, R., 2014. Effects of N2/CO2 on explosion
dioxide and diatomite powder aerosols on coal mine gas explosions in highlands. characteristics of methane and air mixture. J. Loss Prev. Process Ind. 31, 10–15.
Aerosol Air Qual. Res. 16, 2119–2128. Wang, Tao, Luo, Zhenmin, Wen, Hu, Cheng, Fangming, Deng, Jun, Zhao, Jingyu,
Luo, Zhenmin, Wang, Tao, Ren, Junying, Deng, Jun, Shu, Chimin, Huang, Anqi, Guo, Zhengchao, Lin, Jingjing, Kang, Kai, Wang, Weifeng, 2017. Effects of
Cheng, Fangming, Wen, Zhenyi, 2017. Effects of ammonia on the explosion and flammable gases on the explosion characteristics of CH4 in air. J. Loss Prev. Process
flame propagation characteristics of methane-air mixtures. J. Loss Prev. Process Ind. Ind. 49 (Part B), 183–190.
47, 120–128. Wang, Shimao, Wu, Dejian, Guo, Hai, Li, Xiangdong, Pu, Xuyang, Yan, Zhihui,
Luo, Z.M., Su, B., Wang, T., et al., 2019a. Effects of propane on the flammability limits Zhang, Peili, 2020. Effects of concentration, temperature, ignition energy and
and chemical kinetics of methane-air explosions. Combust. Sci. Technol. 0, 1–17. relative humidity on the overpressure transients of fuel-air explosion in a medium-
Luo, Zhenmin, Liu, Litao, Cheng, Fangming, Wang, Tao, Su, Bin, Zhang, Jiang, scale fuel tank. Fuel 259. https://doi.org/10.1016/j.fuel.2019.116265.
Gao, Shuaishuai, Wang, Chao, 2019. Effects of a carbon monoxide-dominant gas Wang, Tao, et al., 2020. Flammability limit behavior of methane with the addition of
mixture on the explosion and flame propagation behaviors of methane in air. J. Loss gaseous fuel at various relative humidities. Process Saf. Environ. Prot. https://doi.
Prev. Process Ind. 58, 8–16. org/10.1016/j.psep.2020.05.005.
Luo, Zhenmin, Li, Ruikang, Wang, Tao, Cheng, Fangming, Liu, Yin, Yu, Zhijin, Yao, Yao, Ivanovski, Kris, Inekwe, John, Smyth, Russell, 2019. Human capital and energy
Fan, Shixing, Zhu, Xiaochun, 2020. Explosion pressure and flame characteristics of consumption: evidence from OECD countries. Energy Econ. 84 https://doi.org/
CO/CH4/air mixtures at elevated initial temperatures. Fuel 268. https://doi.org/ 10.1016/j.eneco.2019.104534.
10.1016/j.fuel.2020.117377. Yongzhen, Wang, Guitang, Chen, 2013. Advanced Engineering Thermodynamics [M].
Luo, Zhenmin, Li, Dafang, Su, Bin, et al., 2020. On the time coupling analysis of Tsinghua University Press, China.
explosion pressure and intermediate generation for multiple flammable gases. Zeng, Wen, Ma, Hongan, Liang, Yuntao, Hu, Erjiang, 2015. Experimental and modeling
Energy 198, 117329. https://doi.org/10.1016/j.energy.2020.117329. study on effects of N2 and CO2 on ignition characteristics of methane/air mixture.
Luo, Zhenmin, Li, Dafang, Su, Bin, et al., 2020. Thermodynamic effects of the generation J. Adv. Res. 6 (2), 189–201.
of H*/OH*/CH2O* on flammable gas explosion. Fuel 280, 118679. https://doi.org/ Zhang, Weikuo, Chen, Zheng, Kong, Wenjun, 2012. Effects of diluents on the ignition of
10.1016/j.fuel.2020.118679. premixed H2/air mixtures. Combust. Flame 159 (1), 151–160.
Luo, Zhenmin, et al., 2020. Experimental study on the deflagration characteristics of Zhang, Qi, Wang, Yaxing, Lian, Zhen, 2017. Explosion hazards of LPG-air mixtures in
methane-ethane mixtures in a closed duct. Fuel. https://doi.org/10.1016/j. vented enclosure with obstacles. J. Hazard. Mater. 334, 59–67.
fuel.2019.116295. Zhang, Qi, Chen, Yuying, Fan, Tao, Yuan, Mengqi, Liu, Zhenyi, Huang, Ping,
Melikoglu, Mehmet, 2014. Demand forecast for road transportation fuels including Qian, Xinming, 2019. Flame dynamics and flammability limit of DME(30%)/LPG
gasoline, diesel, LPG, bioethanol and biodiesel for Turkey between 2013 and 2023. blended clean fuel in elongated closed pipeline under multi-factors. Fuel 254.
Renew. Energy 64, 164–171. https://doi.org/10.1016/j.fuel.2019.115731.
Mishra, D.P., Kumar, P., 2010. Effects of N2 gas on preheated laminar LPG jet diffusion Zhou, Jianhua, Li, Bei, Ma, Daqing, Jiang, Haipeng, Gan, Bo, Bi, Mingshu, Gao, Wei,
flame. Energy Convers. Manage. 51 (11), 2144–2149. 2019. Suppression of nano-polymethyl methacrylate dust explosions by ABC powder.
Nair, Aswathy, Velamati, Ratna Kishore, Kumar, Sudarshan, 2016. Effect of CO2/N2 Process. Saf. Environ. Prot. 122, 144–152.
dilution on laminar burning velocity of liquid petroleum gas-air mixtures at elevated
temperatures. Energy 100, 145–153.

12