Experimental research of particle size and size dispersity on the explosibility
characteristics of coal dust

Qingzhao Li, Ke Wang, Yuannan Zheng, Maliang Ruan, Xiaoning Mei,

Baiquan Lin

PII: S0032-5910(16)30034-1
DOI: doi: 10.1016/j.powtec.2016.01.035
Reference: PTEC 11472

To appear in: Powder Technology

Received date: 15 July 2015

Revised date: 29 December 2015
Accepted date: 27 January 2016

Please cite this article as: Qingzhao Li, Ke Wang, Yuannan Zheng, Maliang Ruan,
Xiaoning Mei, Baiquan Lin, Experimental research of particle size and size disper-
sity on the explosibility characteristics of coal dust, Powder Technology (2016), doi:
10.1016/j.powtec.2016.01.035

This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.
 ACCEPTED MANUSCRIPT

Experimental research of particle size and size dispersity on the

explosibility characteristics of coal dust

Qingzhao Li*, Ke Wang, Yuannan Zheng, Maliang Ruan, Xiaoning Mei, Baiquan Lin

T
School of Safety Engineering, China University of Mining and Technology, State Key Laboratory of Coal

IP
Resources and Safe Mining, Key Laboratory of Gas and Fire Control for Coal Mines of Ministry of

R
Education, Xuzhou 221116, Jiangsu Province, PR China.

SC
Abstract: In the present research work, the effects of particle size (D50) and size dispersity (σD) on the

explosion severity of coal dust were investigated using 20 L spherical explosion vessel. Coal dust samples

NU
with similar median diameter (D50) & different size dispersity (σD) and similar size dispersity (σD) &

different size diameter (D50) were specially prepared by mixing original coal samples having a narrow size
MA
distribution. The explosion severity of each sample was evaluated by measuring the maximum pressure

(Pmax) and the maximum rate of pressure rise ((dP/dt)max). Interestingly, results show that almost all of the
D

selected coal dusts present the same optimum explosion concentration of 250 g/m3. The parameters of Pmax
TE

and (dP/dt)max show an increasing trend in explosion severity as size dispersity (σD) and particle size (D50)

decreases, which indicate a more notable effect due to the contribution of finer coal particles on the dust
P

cloud deflagration dynamics. The higher the volatile matter content, the more severe is the explosion. The
CE

presence of the finer particles would greatly increase the total effective specific surface area and speed up
AC

the devolatilization rate, which would cause acceleration of dust explosion process. Furthermore, the effects

of dust concentration on the explosion process were analyzed by comparing the defined time span (τ1) from

ignition point to reach (dP/dt)max point and time span (τ2) from (dP/dt)max point to reach Pmax point. For the

coal dust with lower D50 and lower σD, τ1 and τ2 are always short. Under lower coal dust concentration

condition (< 250 g/m3), τ1 and τ2 present a stronger dependence on σD and dust concentration. However, the

effect of σD and dust concentration on τ1 and τ2 becomes weak for the higher coal dust clouds (> 250 g/m3).

Results indicated that the evaluation of coal dust explosion hazard should be considered in terms of not

only dust concentration, but also particle size (D50) and size dispersity (σD).

Keyword: Coal dust; Particle size distribution; Dispersity; Dust concentration; Explosion characteristics

* Corresponding author. Tel.: +86 516 83884401; fax: +86 516 83884401
E-mail address: qingzhaolee@163.com (Q.Z. Li)

1
 ACCEPTED MANUSCRIPT

1. Introduction

Coal dust explosion in the coal mine is one of the serious hazards for the coal production industry,

which has attracted extensive interest for many researchers and government managers for a long period.

T
Junshi Fang，the general of Coal division of China Energy Bureau, points out that Chinese coal production

IP
accounts for about 37% of the total world production, but the death toll accounts for about 70% of the

R
world. The severity of coal mine accidents should be paid more attention [1], especially the accident of coal

SC
dust explosion, which is one of the major malignant coal mine accident. However, compared with the

developed countries, Chinese researches on coal dust explosion started relatively later and coal mine

NU
workers had not paid sufficient attention to the harm caused by coal dust explosions. During coal

processing, transport and storage, large amounts of coal dust can be produced and suspended in the air. If
MA
these coal dust & air mixtures meet high temperature, electric spark or flame in the environment, it is very

likely to cause a coal dust explosion. The explosion shock wave can disperse a deposit of coal dust,

subsequently causing an explosion by the flame propagation and heat radiation [2, 3].
D

As for the evaluation and prevention of coal dust explosion, explosion parameters are an important
TE

basis for explosion risk assessment and explosion protection, mainly containing the minimum ignition
P

temperature of a cloud (MITC), the lower explosive limit (LEL), the maximum explosion pressure (Pmax),
CE

the maximum rate of pressure rise ((dP/dt)max) and the explosion index (Kst) [4-6], which have been studied

by many researchers.
AC

Since full-scale mine explosion tests are very expensive and time-consuming, few full-scale coal dust

explosion experiments have been performed, of which the characteristics of the dust explosion in the coal

mine has been obtained in US [7]. And now, many scholars turn to study the gas and coal dust explosion

characteristics using laboratory-scale apparatuses. Coal dust explosions have been conducted in laboratory

chambers, which mostly include 20 L explosion spheres and 1 m3 explosion chambers. Pineau and

Ronchail [8, 9] conducted coal dust/air explosions in a duct which connected to a vessel. In their

experiments, dusts were layered initially at the bottom of the duct. They were dispersed and ignited by the

powerful explosion. In the experiments of Gardner [10], the dust/air mixtures were formed by blowing air

and coal dust through the experimental tube which connected to an ignition chamber. Wolanski [11] and

Kauffman [12] performed dust explosion using a vertical experimental tube.

The effects of coal volatility and particle size on dust explosion were evaluated by Cashdollar [13] and

2
 ACCEPTED MANUSCRIPT

Going [14]. In their experiments, the minimum explosive dust concentrations (MEC) and limiting oxygen

concentrations (LOC) of different dusts were also measured and compared in 20 L and 1 m3 explosion

chambers. Cashdollar [13] studied the characteristics of coal dust explosion using the 20 L explosion vessel

and pointed out a conclusion that coal particle size and the volatile contents had the same important

T
decisive effect on the explosion severity of coal dust, which were consistent with results from Bi [15].

IP
Using two experimental tubes with different diameter and length, coal dust/air explosion experiments were

R
performed by Bartknecht [16, 17] and results showed that small size coal dust produced great maximum

SC
explosion flame speed. Using the 20 L explosion sphere, Gao [18] studied the lower explosion limit, the

maximum explosion pressure and the maximum explosion pressure rising rate of coal dust samples. It was

NU
found that the lower explosion limit decreased with the decrease of particle size. For a given concentration,

finer coal particles always caused higher explosion severity. Considering the effect of large particles, Man’s
MA
[19] research showed that coarse, medium, and pulverized Pittsburgh coal had similar Pmax values, but the

MEC would increase with increasing particle size, which indicated that the explosion was always
D

controlled by the proportion of finer size components to a great extent, probably 150 µm or smaller. Due to
TE

the difference of experimental apparatus and environment condition, these results vary from Amyotte’s [20].

Consulting the analysis of the coal dust explosion results in 20 L standard explosion vessel, Cao [21] found
P

that the participation of effective volume of coal dust had significant influence on explosion parameters.
CE

And, Castellanos [22] reported that particle size dispersity (σD) had great effect on the explosion severity of

aluminum dust. Taking surface area into consideration, Harris [23] proved that dust particle size had the
AC

greatest influence on the propagation of dust explosions. Using FLUENT software to simulate the spraying

and explosion of coal dust in a spherical explosion chamber, Wang [24] found that coal dusts of 1 um were

distributed uniformly in the chamber and coal dusts with larger diameters were concentrated in the center

and boundary of the chamber, which caused a difference in ignition of coal dust.

From previous research results, it can be concluded that dust explosion is a surface-area dependent

process and can be affected not only by the mean diameter but also by the proportion of finer particles.

Therefore, dust explosions should be evaluated not only by the mean diameter but also by the proportion of

finer size components, which is named as size dispersity (σD). In other words, for coal dust with a given

median particle diameter (D50), the values of Pmax and (dP/dt)max will be affected greatly by the systematic

variation of small and large particle size fraction contained in the mixture i.e. different σD. However, most

3
 ACCEPTED MANUSCRIPT

of the previous experimental researches were carried out using samples with a certain median diameter and

did not consider particle size dispersity (σD) [8-12, 15-17]. Furthermore, it is difficult to compare

experimental data from different researchers whose results are reported in relation to different definitions of

average particle size. To understand the effect of particle size and size dispersity (σD) on the explosibility

T
characteristics of coal dust, special designed experiments considering the effect of coal dust size and size

IP
dispersity (σD) for coal dust explosion have been conducted in the present research work.

R
2. Methodology

SC
2.1 Apparatus and methods

NU
In the present research paper, experiments of coal dust explosion were performed in a standard 20 L

stainless steel spherical vessel (Fig.1) according to the international standard ISO6184-1. Before dust
MA
explosion test, a pre-weighted amount of coal dust was first placed in dust container (volume: 0.6 L), the

centrally mounted chemical igniter was connected with the ignition leads, and the explosion chamber was
D

closed safely. The explosion chamber was partially vacuumed to 0.06 MPa (Absolute) firstly and the
TE

dispersing air pressure was set to 2 MPa (Gauge). When the solenoid valve (Made by Kuhner AG

Company, Switzerland) between the dust storage container and the test chamber opened automatically, the
P

air and coal dust were dispersed into the explosion chamber and the chemical igniter was energized after a
CE

60 ms time delay. After the test was finished, the explosion chamber and dust container were thoroughly

cleaned with compressed air for the next test [25, 26]. During the experiments, the coal dust cloud in the
AC

vessel is ignited by electrically activated pyrotechnical igniter, which is prepared in accordance with the

principle of zero-oxygen balance. The igniter consists of zirconium, barium nitrate and barium peroxide by

the weight ratio of 4:3:3, and the energy release of 1.2 g mixture is corresponding to 5 kJ. Based on the

pre-experiments results, coal dust concentration is varied among 60 g/m3-500 g/m3 ranges

2.2. Determination of Pmax, (dP/dt) max, and τ of coal dust samples

By the standard 20 L stainless steel spherical explosion vessel, experiments of coal dust explosion

were performed at least in three replications and the explosion pressure evolutions were measured by a

pressure sensor (produced by Dytran Company, America) installed in the vessel wall and recorded by a

data acquisition system for each run. These data yielded values of the maximum explosion pressure (Pmax)

and maximum rate of pressure rise (dP/dt) max.

4
 ACCEPTED MANUSCRIPT

Fig.2 shows a typical pressure profile as a function of time during a coal dust explosion test, where

(dP/dt)max, Pmax, and τ (coal dust burning time) are obtained for a specific dust concentration, which is

corrected considering the cooling effects of the vessel walls and the pressure effects caused by the igniter

[27]. As for τ1 and τ2 (shown in Fig.2), they are used to describe the flame propagation velocity.

T
IP
2.3. Coal sample preparation and size characterizations

R
Six kinds of coal in different ranks are selected among Chinese typical coal fields. All samples are

SC
crushed and screened separately according to the experimental design. Proximate analyses of original coal

samples (before drying process) are determined by MAC-800A of automatic proximate analyzer according

NU
to standard method of GB/T212-2001 [28] of China and the results are shown in Table 1. Before the

explosion tests, the screened coal samples are dried overnight at about 60 °C under nitrogen gas stream (2
MA
ml/min) environment for moisture removal.

In order to understand the effect of particle size dispersity (σD) at a similar particle size (D50) on dust

explosion property, coal samples are specially combined with different mean diameters particles in certain
D

proportions and manually blend for about 30 minutes to ensure uniform mixing of each coal samples.
TE

Particle size distribution of the samples is determined using an LS13-320 Coulter multi-wavelength laser
P

diffraction particle size analyzer (Beckman Coulter, Inc. Brea, CA) and particle size distributions on a
CE

volume basis and the statistical diameters, D10, D50, and D90 are provided. Dxx refers to the particle size for

which xx% (by volume) of the particles by volume is finer. Measured results of coal samples are
AC

summarized in Table 2. And, particle size dispersity (σD) used for the characterization of size distribution

span is calculated using the following equation [22].

 D   D90  D10  / D50 (1)

3. Results and discussion

3.1. Effect of particle size on maximum explosion pressure and maximum pressure rise

In order to analyze the effect of D50 on Pmax and (dP/dt)max at a relatively same dispersity, the

explosion severity of two samples with similar D (1.63, 1.67) and different D50 (55.94 µm, 37.33 µm) were

tested by the 20 L dust explosion vessel. Sample size and size distribution are measured and results are

shown in Fig.3. The explosion parameters of the two samples are presented in Fig.4 as a function of coal

5
 ACCEPTED MANUSCRIPT

dust concentrations (60, 125, 250, and 500 g/m3). It can be seen that maximum explosion pressure (Pmax)

and the maximum rate of pressure rise (dP/dt)max both undergo an initial growth and then decrease with the

increase of coal dust concentration. However, the finer coal particles (D50 =33.73 µm) always produce

higher explosion pressure (Pmax). Generally, the effective specific surface area of coal dust would greatly

T
increase with particle size reduces. In addition, decreasing of particle size would greatly increase the

IP
number of coal particles under the same dust concentration, which would also increase the effective

R
reaction surface of coal dusts. From these results, it can be deduced that coal particle combustion reactions

SC
are directly related to effective surface of dust particles per unit volume [21]. Interestingly, two samples’

optimum concentrations corresponding to the maximum explosion pressure (Pmax) are almost at the same

NU
concentration of 250 g/m3. This is mainly because of the limitation of oxygen. Compared with the rule of

particle size on Pmax, particle size has much more effect on the maximum rate of pressure rise (dP/dt)max.
MA
On the one hand, under low dust concentration (＜400 g/m3 ) coal dusts with smaller particle size are

heated easily and release much more volatile matter at the same condition, which would cause a higher
D

burning rate and produce a larger explosion pressure and a pressure rise rate [29]. On the other hand, due to
TE

particles’ gravity effects, too many large particles would fail to be entrained in the air and do not participate

in the explosion reactions [27]. Therefore, decreasing of coal particle size (D50) would accelerate coal dust
P

particles combustion rate and produce a significant increase in the explosion pressure rise rate ((dP/dt)max).
CE

However, the behaviors of Pmax and (dP/dt)max in higher dust concentration (such as 500 g/m3 in Fig.4) are
AC

different from those in low concentration conditions, which can be explained by the fact that the dispersity

of small coal dust will be affected due to too many number of particles. In addition, particles with larger

diameters consume less oxygen during devolatilization process due to less number of particles under high

dust concentration. Therefore, more oxygen takes part in oxidation reaction, which accounts for the

behaviors of Pmax and (dP/dt)max.

3.2. Effect of size dispersity on maximum explosion pressure and maximum pressure rise

To study the effect of size dispersity (σD) on Pmax and (dP/dt)max at a near equal D50, two special

prepared samples (D50=51.13 µm & σD=1.84; D50=52.36 µm & σD= 3.60) were tested following the same

test procedure. Particle size distributions of the two prepared coal dust samples are presented in Fig.5. All

the experiments are conducted at the nominal dust concentration of 60, 125, 250, 400, and 500 g/m3.

Experimental results of Pmax and (dP/dt)max of all coal dust explosion tests are shown in Fig.6. Interestingly,

6
 ACCEPTED MANUSCRIPT

the samples characterized by similar D50 but various coal particle size dispersity (σD) present different

explosibility. From the results of Fig.6, it can be seen that coal dust with larger size dispersity (σD) always

presents less explosibility than that of coal dust with smaller size dispersity (σD). Especially for the

maximum rate of pressure rise (dP/dt)max at higher dust concentration ranges (> 250 g/m3), where a

T
significant enhancement for (dP/dt)max values (Fig.6b). It can be concluded that size dispersity (σD) may be

IP
one of the important factors influencing on coal dust deflagration dynamics. Based on the particle size

R
distribution pattern of different coal dust samples (shown in Fig.5), it can be seen that particle size shows a

SC
dual-peak distribution for the coal dust sample with higher dispersity (σD=3.60). This suggests that there

are a number of larger coal particles present in the sample with higher σD value. In other words, a

NU
considerable number of larger coal particles are dispersed into the coal dust mixtures. Present research

shows that the gradual increase in Pmax and (dP/dt)max value is closely related to the fraction of fine
MA
particles suspended in the dust cloud. Therefore, more coarse coal particles presented in the dust cloud

would decrease the total effective surface areas and slow down the deflagration rate of coal dust cloud [30].
D

In addition, as σD increases, too many deposited large coal particles would cause a less efficient heat
TE

transfer to the neighboring coal particles and produce more heat loss at the same explosion conditions [29].

Similarly, in terms of (dP/dt)max, the rate reduces and then declines as σD increases, especially when the
P

concentration of coal dust exceeds 250 g/m3. Generally, it is well known that large particles are more
CE

difficult to be heated and less combustible gas generates in result, which will cause the decrease of

explosion pressure rise rate. Furthermore, when coal dust concentration exceeds 250 g/m3, it may be
AC

unfavorable to the combustion and pyrolyzation of particles because more larger coal particles could reduce

the dispersed characteristics of coal dusts, which would lead to a decrease of explosion rise rate (dP/dt)max.

Therefore, it can be concluded that as particle size dispersity (σD) decreases, the increased fraction of finer

particles would provide much more effective surface areas for reaction and cause a higher devolatilization

rate, which would accelerate the dust explosion propagation.

However, in this paper (As shown in Fig.5b), the large σD is due to the increase of fraction of large

particles (right peak), which attributes to the increase of D90 based on eq.(1). Meanwhile, decrease of D10

and keep the constant of D90 can also increase σD. In this case, much more volatile matters would be

emitted and combusted at higher σD conditions. The Pmax and (dP/dt)max would be deeply controlled by

oxygen supplement in the explosion environment.

7
 ACCEPTED MANUSCRIPT

In all, coal explosion properties are dominated by both particle size (D50) and dust concentration.

Particle size dispersity (σD) is not a direct factor but an significant one because size dispersity (σD) is

closely dependent on particle size. With the reduction of particle size, the total effective surface areas of

particles increases, which would improve the explosion properties of coal dust. Therefore, it is suggested

T
that the evaluation of coal dust explosion hazard should be reported in terms of not only dust concentration,

IP
but also considering the particle size (D50) and size dispersity (σD).

R
3.3. Effect of volatile content on maximum explosion pressure and maximum pressure rise

SC
Influence of coal sample volatile content on the Pmax and (dP/dt)max values were also analyzed.

Samples with relatively similar size dispersity (σD) and similar particle diameter (D50) were special

NU
prepared in terms of experimental purposes. Proximate analysis results of coal dust samples are listed in
MA
Table 1 and particle size distributions of the selected coal dust samples are presented in Fig.7. All the

experiments are conducted at the nominal dust concentration of 60, 125, 250, 400, and 500 g/m3 with the

same procedures by the 20 L explosion vessel. The results of the explosion parameters as a function of coal
D

dust concentration are shown in Fig.8. Results show that significant differences could be observed for the
TE

selected coal samples, especially for (dP/dt)max values, which agrees with the results of Li [26].

Interestingly, the special prepared coal dusts present the similar maximum explosion concentrations of 250
P

g/m3.
CE

Generally, coal with more volatile contents always produce a higher explosion pressure (Pmax) and
AC

pressure rise rate (dP/dt)max. From Fig.8, it can be seen that the higher volatile matter contents a coal

sample has, the larger explosion parameters of Pmax and (dP/dt)max are. The reason for this phenomenon can

be explained as coal dusts with higher volatile content may produce more combustible gas at the same

conditions, which contributes to the gas phase deflagration propagations. In addition, from analysis on the

scanning electron microscope (SEM) images (shown in Fig.9) of the coal dust explosion solid products

under the same concentration (250 g/m3), it can be found that coal dust with more volatile matter content is

more reactive and the pore shape in the surface of the post-explosion solid particles tending to be more

circular and bigger. It can be concluded and verified that coal dust with higher volatile matter contents

would present better deflagration properties and produce a higher explosion severity. However, after

optimum concentration, the explosion pressure (Pmax) would decrease slightly for the coal dusts with higher

volatile matter content. This is because excessive volatile content prevents oxygen reaching the coal

8
 ACCEPTED MANUSCRIPT

particle surface during devolatilization process [31], especially for the coal dust with higher volatile matter

contents. Present results confirm that volatile matter boosts the explosion severity. However, the

deflagration reaction would also be hindered by the devolatilization process due to the excessive

consumption of oxygen at the condition of coal dust concentration much higher than the optimum

T
concentration.

IP
3.4. Effect of particle size and size dispersity on flame propagation velocity of coal dust

R
In the present work, τ1 is defined as the time span from ignition point to point of maximum rate of

SC
pressure rise, which indicates the fastest burning rate of most volatile matters pyrolyzed by coal dust. τ2 is

NU
defined as the time span from the point of maximum rate of pressure rise to the point of maximum

explosion pressure, which means the burning stage of part of coal char. τ1 and τ2 are shown in Fig.2 and
MA
they are used to discuss the effects of D50, σD, and dust concentration on the dust deflagration flame

propagation velocity for the special prepared coal samples. Based on the previous experiments, the

determined τ1 and τ2 as a function of nominal dust concentrations are shown in Fig.10 respectively. From
D

the results of Fig.10a, it can be seen that coal dust samples with similar size dispersity (σD) show an
TE

apparent reduction of τ1, τ2 as D50 decreases at the same dust concentration. On the other hand, as dust

concentration increases, τ2 presents a gradual decreasing rule correspondingly which reveals the
P
CE

acceleration of coal dust combustion due to the distance between coal particles becomes small. However, as

a measure of the gas phase explosion, τ1 increases gradually when coal dust concentration is below 125
AC

g/m3 and then follow the same law of τ2 at relatively high dust concentrations. Under low coal dust

concentration, τ1 is short because particles to be heated are few in number. With the increase of coal dust

concentration, more particles need to be heated which leads to the rise of τ1. As further increase of coal dust

concentration, distance between particles becomes smaller which is beneficial to the heating of coal dust

particles, causing the decrease of τ1 in result. Furthermore, the effects of dust concentration on τ1 and τ2 can

be approximately neglected in the rich dust mixtures. The results confirm the fact that small coal particles

are easy to be heated and ignited, which would generate a higher explosion pressure (Pmax) and pressure

rise rate (dP/dt)max [25]. Taking the longer combustion time of larger coal particles into consideration, it can

be assumed that longstanding combustion of large particles would compensate for their energy loss to a

certain extent [32-35].

In the case of coal samples with similar D50 and different σD, τ1 and τ2 present a reducing trend as σD

9
 ACCEPTED MANUSCRIPT

decreases especially in the poor coal dust cloud (dust concentration lower than 250 g/m3), which is shown

in Fig.10b. In other words, τ1 and τ2 present a stronger dependence on σD at relatively low dust

concentration conditions. From the results, it can be seen that as σD decreases from 3.60 to 1.84 with the

same dust concentration of 60 g/m3, there is a dramatic reduction of τ1 and τ2, where τ1 drops from 80.3 ms

T
to 60.2 ms and τ2 drops from 110.0 ms to 50.1 ms. This phenomenon can be explained from the role played

IP
by the fine particles on coal particle combustion process. Under the same dust concentration, coal dusts

R
with low σD contain more fine particles compared to the coal dusts with high σD. Therefore, coal dusts with

SC
low σD are easier to be heated and ignited. In addition, under the same dust concentration, rising of fine

particles proportion would increase the total numbers of coal dust, which would cause a enhanced

NU
combustion process because of the reduction of the inter-particle spacing [36-38].
MA
4. Conclusion

Using 20 L explosion spherical vessel, the effects of particle size (D50) and size dispersity (σD) on the
D

explosion severity of coal dust were investigated. Coal dust samples with similar median diameter (D50) &

different size dispersity (σD) and similar size dispersity (σD) & different size diameter (D50) were special
TE

prepared by mixing original coal samples having narrow size distributions. The explosion severity of each
P

sample was evaluated by measuring the maximum pressure (Pmax) and the maximum rate of pressure rise
CE

((dP/dt)max). Interestingly, results show that almost all the selected coal dusts present the same optimum

explosion concentrations of 250 g/m3. Parameters of Pmax and (dP/dt)max show an increasing trend in
AC

explosion severity parameters as size dispersity (σD) and particle size (D50) decreasing, which indicate a

more notable effect due to the contribution of finer coal particles on the dust cloud deflagration dynamics.

The higher volatile matter a coal sample has, the more explosion severity is. Presence of the finer particles

would greatly increase the total effective specific surface area and speed up the devolatilization rate, which

would cause acceleration of dust explosion process. Furthermore, the effect of dust concentration on the

explosion process is analyzed by comparing the defined time span (τ1) from ignition point to reach

(dP/dt)max point and time span (τ2) from (dP/dt)max point to reach Pmax point. For the coal dust with lower

D50 and lower σD, τ1 and τ2 are always short. When σD decreases from 3.60 to 1.80 with the same dust

concentration of 60 g/m3, there is a dramatic reduction of τ1 and τ2, where τ1 drops from 80.3 ms to 60.2 ms

and τ2 drops from 110.0 ms to 50.1 ms. Under lower coal dust condition (< 250 g/m3), τ1 and τ2 present a

10
 ACCEPTED MANUSCRIPT

stronger dependence on σD and dust concentration. However, the effect of σD and dust concentration on τ1

and τ2 is weak for the higher coal dust clouds (> 250 g/m3). The results indicate that the evaluation of coal

dust explosion hazard should be considered in terms of not only dust concentration, but also particle size

(D50) and size dispersity (σD).

T
IP
Acknowledgements

R
Financial supports from the National Natural Science Foundation of China (Grant No. 51574230), the

SC
Fundamental Research Funds for the Central Universities (Grant No. 2015XKMS010), the Natural Science

Foundation of Jiangsu Province (Grant No. BK20131115), the Fund for Innovation Team of CUMT

NU
(2014QN001), A project funded by the priority academic program development of Jiangsu higher education

institutions, and the Program for Changjiang Scholars and Innovative Research Team in University (Grant
MA
No. IRT13098) are sincerely acknowledged.

References
D

[1]. Hao, Q.M., Li, X.M., 2011. Discussion on Coal Mining Accidents & Explosions. Coal 20, p. 43
TE

[2]. Zhao, H.Y., 1996. The theory of gas and dust explosion. Beijing Institute of Technology Press, Beijing, China.

[3]. Hu, S.Q., Yu, C.J., Tan, Y.X., 2010. Experimental research on secondary explosion coal dust detonated by gas
P

explosion in pipeline. Journal of Basic Science and Engineering 18, p. 895.

CE

[4]. ASTM, 2006. E1491 Standard Test Method for Minimum Autoignition Temperature of Dust Clouds. Pennsylvania:

Standards Press of America.

AC

[5]. ASTM, 2007. E1515 Standard Test Method for Minimum Explosible Concentration of Combustible Dusts.

Pennsylvania: Standards Press of America

[6]. ASTM, 2005. E1226 Standard Test Method for Pressure and Rate of Pressure Rise for Combustible Dusts.

Pennsylvania: Standards Press of America.

[7]. Sapko, M.J., Weiss, E.S., Cashdollar, K.L., Zlochower, I.A., 2000. Experimental mine and laboratory dust explosion

research at NIOSH. Journal of Loss Prevention in the Process Industries 13, 229–242.

[8]. Pineau JP, Ronchail G, Propagation of dust explosions in Ducts. The international symposium on control and

prevention of dust explosions, Basle, Switzerland, 1982.

[9]. Pineau JP, Ronchail G, Propagation of coal dust explosions in pipes. The symposium on industrial explosions, Pittsburg,

USA, 1986.

[10]. Gardner BR, Winter RJ, Moore MJ, Explosion development and deflagration to detonation in coal dust/air suspensions.

In: Proceedings of the 21st symposium (international) on combustion, Pittsburg, The combustion institute, (1986)

11
 ACCEPTED MANUSCRIPT

335-343.

[11]. Wolanski P, Sacha W, Zelesinski M, Effect of dust concentration on detonation parameter parameters in grain dust-air

mixtures, In: Proceedings of the 4th international colloquium on dust explosion, Porabka-Kozubnik, Poland, (1990)

355-370.

T
[12]. Kauffman CW, Wolanski P, Ural E, Nicholls JA, Detonation waves in confined dust clouds, In: Proceedings of the 19th

IP
international symposium on combustion, Israel, 1982.

[13]. Cashdollar, K. L., 2000. Overview of dust explosibility characteristics. Journal of Loss Prevention in the Process

R
Industries 13, 183-199.

SC
[14]. Going JE, Chatrathi K, Cashdollar KL, Flammability limit measurements for dusts in 20L and 1m3 vessels. Loss

Prevent. Process Ind., 13 (2000) 209-219.

NU
[15]. Bi, M.S., Wang, H.Y., 2008. Experiment on methane-coal dust explosions. Journal of China Coal Society 33(7),

784-788.
MA
[16]. Bartknecht W., Explosions: Course, prevention, protection, New York, (1981) 251-259.

[17]. Bartknecht W., Dust explosions, Springer-Verlag, Berlin, 1989.

[18]. Gao, C., Li, H., Su, D., 2010. Explosion characteristics of coal dust in a sealed vessel. Explosion and Shock Waves 30,
D

p. 164.
TE

[19]. Man, C.K., Harris, M.L., 2014. Participation of large particles in coal dust explosions. Journal of Loss Prevention in the

Process Industries 27, 49-54.

P

[20]. Amyotte, P.R., Mintz, K.J., Pegg, M.J., Sun, Y.H., 1993. The ignitability of coal dust-air and methane-coal dust-air
CE

mixtures. Fuel 72(5), 671-679.

[21]. Cao, W.G., Huang, L.Y., Zhang, J.X., Xu, S., Qiu, S.S., Pan, F., 2012. Research on characteristic parameters of

coal-dust explosion. Procedia Engineering 45, 442-447.

AC

[22]. Castellanos, D., Carreto-Vazquez, V.H., Mashuga, C.V., Trottier, R., Mejia, A.F., Mannan, M.S., 2014. The effect of

particle size dispersity on the explosibility characteristics of aluminum dust. Powder Technology 254, 331-337.

[23]. Harris M.L., Sapko M.J., Zlochower I.A., Perera I.E., Weiss E.S., 2015. Particle Size and Surface Area Effects on

Explosibility Using a 20-L Chamber, Journal of Loss Prevention in the Process Industries, Vol. 37, pp 33-38.

[24]. Wang C., Dong X.Z., Ding J.X., Nie B.S., 2015. Numerical investigation on the spraying and explosibility

characteristics of coal dust. Mining, Reclamation and Environment 28, 5, 287-296.

[25]. Li, Q.Z., Zhai, C., Wu, H.J., Lin, B.Q., Zhu, C.J., 2011. Investigation on coal dust explosion characteristics using 20 L

explosion sphere vessels. Journal of China Coal Society 36, 119-124.

[26]. Li, Q.Z., Lin, B.Q., Dai, H.M., Zhao, S., 2012. Explosion characteristics of H 2/CH 4/air and CH 4/coal dust/air

mixtures. Powder Technology 229, 222-228.

[27]. ASTM, E1226–00, Standard test method for pressure and rate of pressure rise for combustible dusts, 14.02,

2005,334–345.

12
 ACCEPTED MANUSCRIPT

[28]. GB / T212 - 2001. Proximate analysis of coal. Chinese state administration of quality supervision, inspection and

quarantine.

[29]. Kuai, N.S., Li, J.M., Chen, Z., Huang, W.X., Yuan, J.J., Xu, W.Q., 2011. Experiment-based investigations of

magnesium dust explosion characteristics. Journal of Loss Prevention in the Process Industries 24(4), 302-313.

T
[30]. Xu, H.L., Wang, X.S., Gu, R., Zhang, H.P., 2012. Experimental Study on Characteristics of Methane–Coal-Dust

IP
Mixture Explosion and Its Mitigation by Ultra-Fine Water Mist. Journal of Engineering for Gas Turbines and Power

134(6), 061401-061406.

R
[31]. Shaddix, C.R., Molina, A., 2009. Particle imaging of ignition and devolatilization of pulverized coal during oxy-fuel

SC
combustion. Proceedings of the Combustion Institute 32(2), 2091-2098.

[32]. Dufaud, O., 2010. Experimental investigation and modelling of aluminum dusts explosions in the 20 L sphere. Journal

NU
of Loss Prevention in the Process Industries 23(2), 226-236.

[33]. Castellanos, D., Carreto, V., Mashuga, C., Mannan, S., 2010. Application of 2k factorial experimental design to
MA
determine the optimum test parameters for a 36-L dust explosion equipment. Hazards XXII Symposium, Liverpool,

[34]. Eckhoff, R.K., 2009. Understanding dust explosions: The role of powder science and technology. Journal of Loss

Prevention in the Process Industries 22(1), 105-116.

D

[35]. Kuai, N.S., Huang, W.X., Yuan, J.J., Du, B., Li, Z.S., Wu, Y., 2012. Influence of ignition energy on dust explosion
TE

behavior. Explosion and Shock Waves 4, 432-438.

[36]. Ning, J.G., Wang, C., Lu, J., 2006. Explosion characteristics of coal gas under various initial temperatures and
P

pressures. Shock Waves 15(6), 461-472.

CE

[37]. Proust, C., 2006. A few fundamental aspects about ignition and flame propagation in dust clouds. Journal of Loss

Prevention in the Process Industries 19(2), 104-120.

[38]. Pua, Y.K., Jiaa, F., Wanga, S.F., Skjoldb, T., 2007. Determination of the maximum effective burning velocity of
AC

dust–air mixtures in constant volume combustion. Journal of Loss Prevention in the Process Industries 20(4), 462-469.

13
 ACCEPTED MANUSCRIPT

Table caption list

Table 1 proximate analysis of different coal samples

T
Table 2 Particle size characterization of the coal dust samples

R IP
SC
Figure caption list

Fig.1 20L explosion sphere

NU
MA
Fig.2 Typical pressure profile during a dust explosion test

Fig.3 Coal dust samples with different particle size and similar dispersity

Fig.4 Explosion results of coal dust samples with different particle size and similar dispersity
D

Fig.5 Coal dust samples with different size dispersity and similar particle diameter
TE

Fig.6 Explosion results of coal dust samples with different size dispersity and similar particle diameter
P

Fig.7 Coal dust samples with different volatile contents and similar particle diameter & size dispersity
CE

Fig.8 Explosion results of coal dust samples with different volatile contents and similar particle diameter &

size dispersity
AC

Fig.9 SEM pictures of original coal dust

Fig.10 SEM pictures of dust explosion solid products

Fig.11 Determined τ1, τ2 values as a function of nominal dust concentrations

14
 ACCEPTED MANUSCRIPT

Table 1 proximate analysis of different coal samples

Proximate analysis (w/w %)

Samples
Mad Ad Vdaf FCd

T
Tiefa 9.76 28.12 40.97 42.44

IP
Neimeng 1.48 14.25 14.03 73.72
Lvliang 1.03 6.92 23.38 71.32

R
Ningxia 1.58 28.54 33.51 47.51

SC
Huaibei 2.16 20.55 35.77 51.03
Mad: moisture content; Vdaf: volatile matters; Ad: ash; FCd: fixed carbon.

NU
Table 2 Particle size characterization of the coal dust samples
MA
Samples D90(µm) D50(µm) D10(µm) σD
Tiefa-1 103.4 55.94 12.37 1.63
Tiefa-2 71.95 37.33 9.63 1.67
Lvliang-1 111.2 51.13 17.11 1.84
D

Lvliang-2 205.4 52.36 16.67 3.60

TE

Neimeng 101.4 45.37 9.54 2.02

Ningxia 91.62 43.69 7.14 1.93
P

Huaibei 91.66 45.20 7.92 1.85

CE
AC

15
 ACCEPTED MANUSCRIPT

1
2 15
3
4 16 14

13

T
IP
8
9 11 12
10

R
5
6

SC
7

NU
1-Sealing cover; 2-Outer jacket; 3-Inner jacket; 4-Vacuum gauge; 5-Water Inlet; 6-Outlet Valve; 7-Base;

8-Peep hole; 9-Exhaust port; 10-Rebound Nozzle; 11-Dust Container; 12-Pressure gauge; 13-Pressure
MA
Sensor; 14-Water Outlet; 15-Safety limit switch; 16-Spark rod.

Fig.1 20L explosion sphere

D
P TE
CE
AC

16
 ACCEPTED MANUSCRIPT

Maximum explosion pressure (Pmax)

Explosion pressure / MPa

Maximum rate of
pressure rise(dP/dt)max

T
IP



R
Ignition point

Atmospheric pressure

SC
Time (ms)

NU
Fig.2 Typical pressure profile during a dust explosion test
MA
D
P TE
CE
AC

17
 ACCEPTED MANUSCRIPT

Particle Size Distribution Percentage / %

Particle Size Distribution Percentage / %

20 100 20 100

Cumulative PSD Percentage / %

Cumulative PSD Percentage / %
dp10=12.37μm 16 dp10=9.63μm 80
16 80
dp50=55.94μm Tiefa-1 Coal dp50=37.33μm Tiefa-2 Coal
dp90=103.4μm dp90=71.95μm

T
12 60 12 60

IP
8 40 8 40

R
4 20 4 20

SC
0 0 0 0
1 10 100 1000 1 10 100 1000
Particle Size / μm Particle Size / μm

NU
(a) D50 = 55.94µm & D = 1.63 (b) D50 = 37.33µm & D = 1.67

Fig.3 Coal dust samples with different particle size and similar dispersity
MA
D
P TE
CE
AC

18
 ACCEPTED MANUSCRIPT

0.8 80
Maximum explosion pressure / MPa

70

Maximum pressure rise / MPas )

0.7

-1
60
0.6
50

T
0.5
40

IP
Tiefa coal
0.4
D50=55.94m D 30
D50=37.33m D Tiefa coal
0.3
D50=55.94m D

R
20
D50=37.33m D
0.2
10

SC
0.1 0
0 100 200 300 400 500 0 100 200 300 400 500
3 3
Coal dust concentration/(g/m ) Coal dust concentration/(g/m )

NU
(a) Pmax vs. dust concentration (b) (dP/dt)max vs. dust concentration

Fig.4 Explosion results of coal dust samples with different particle size and similar dispersity
MA
D
P TE
CE
AC

19
 ACCEPTED MANUSCRIPT

Particle Size Distribution Percentage / %

Particle Size Distribution Percentage / %

20 100 20 100

Cumulative PSD Percentage / %

Cumulative PSD Percentage / %
dp10=17.11μm 16 dp10=16.67μm 80
16 80
Lvliang-1 Coal dp50=52.36μm Lvliang-2 Coal
dp50=51.13μm
dp90=111.2μm dp90=205.4μm

T
12 60 12 60

IP
8 40 8 40

R
4 20 4 20

SC
0 0 0 0
1 10 100 1000 1 10 100 1000
Paiticle Size / μm Paiticle Size / μm

NU
(a) D50 = 51.13µm & D = 1.84 (b) D50 = 52.36µm & D = 3.60

Fig.5 Coal dust samples with different size dispersity and similar particle diameter
MA
D
P TE
CE
AC

20
 ACCEPTED MANUSCRIPT

0.9 50
Maximum explosion pressure / MPa

Maximum pressure rise / MPas )

-1
0.8
40

T
0.7
30

IP
0.6

20
Lvliang coal

R
0.5 Lvliang coal
D50=51.13m D
D50=51.13m D
10

SC
0.4 D50=52.36m D
D50=52.36m D

0.3 0
0 100 200 300 400 500 0 100 200 300 400 500
3 3
Coal dust concentration/(g/m ) Coal dust concentration/(g/m )

(a) Pmax vs. dust concentration

NU
Fig.6 Explosion results of coal dust samples with different size dispersity and similar particle diameter
(b) (dP/dt)max vs. dust concentration
MA
D
P TE
CE
AC

21
 ACCEPTED MANUSCRIPT

Particle Size Distribution Percentage / %

20 100

Cumulative PSD Percentage / %

16
dp10=9.54μm 80
dp50=45.37μm Neimeng Coal
（ Vdaf=14.03%）
dp90=101.4μm

T
12 60

IP
8 40

R
4 20

SC
0 0
1 10 100 1000
Paiticle Size / μm

NU
(a) D50 = 45.37µm & D = 2.02
Particle Size Distribution Percentage / %

20 100
MA

Cumulative PSD Percentage / %

16
dp10=7.14μm 80
dp50=43.69μm Ningxia Coal
(Vdaf=33.51%)
dp90=91.62μm
12 60
D

8 40
TE

4 20
P

0 0
1 10 100 1000
Paiticle Size / μm
CE

(b) D50 = 43.69µm & D = 1.93

Particle Size Distribution Percentage / %
AC

20 100
Cumulative PSD Percentage / %

16
dp10=7.92μm 80
dp50=45.2μm Huaibei Coal
(Vdaf=35.77%)
dp90=91.66μm
12 60

8 40

4 20

0 0
1 10 100 1000
Paiticle Size / μm

(c) D50 = 45.20µm & D = 1.85

Fig.7 Coal dust samples with different volatile contents and similar particle diameter & size dispersity

22
 ACCEPTED MANUSCRIPT

0.9 90
Maximum explosion pressure / MPa

Neimeng coal (Vdaf=14.03%)

0.8 80

Maximum pressure rise / MPas )

D50=45.37μm & σD=2.02

-1
0.7 70 Ningxia coal (Vdaf=33.51%)
D50=43.69μm & σD=1.93
0.6 60
Huaibei coal (Vdaf=35.77%)

T
0.5 50 D50=45.20μm & σD=1.85

IP
0.4 Neimeng coal (Vdaf=14.03%)
40
D50=45.37μm & σD=2.02
0.3 Ningxia coal (Vdaf=33.51%) 30
D50=43.69μm & σD=1.93

R
0.2 20
Huaibei coal (Vdaf=35.77%)
0.1 D50=45.20μm & σD=1.85 10

SC
0.0 0
0 100 200 300 400 500 0 100 200 300 400 500
3 3
Coal dust concentration/(g/m ) Coal dust concentration/(g/m )

NU
(a) Pmax vs. dust concentration (b) (dP/dt)max vs. dust concentration

Fig.8 Explosion results of coal dust samples with different volatile contents and similar particle diameter &
MA
size dispersity
D
P TE
CE
AC

23
 ACCEPTED MANUSCRIPT

T
R IP
SC
(a) Neimeng coal Vdaf=14.03% (b) Ningxia coal Vdaf=33.51% (c) Huaibei coal Vdaf=35.77%

Fig.9 SEM pictures of dust explosion solid products

NU
MA
D
P TE
CE
AC

24
 ACCEPTED MANUSCRIPT

160 120

140 Tiefa coal Lvliang coal

1 (D50= 55.94μm & σD = 1.63) 100 1(D=1.84 & D50=51.13μm)
120 2 (D50= 55.94μm & σD = 1.63) 2(D=1.84 & D50=51.13μm)
1 (D50= 37.33μm & σD = 1.67)
Time / ms

Time / ms
80 1(D=3.60 & D50=52.36μm)
100

T
2 (D50= 37.33μm & σD = 1.67) 2(D=3.60 & D50=52.36μm)
80 60

IP
60
40

R
40
20
20

SC
0 0
0 100 200 300 400 500 0 100 200 300 400 500
3 3
Coal dust concentration/(g/m ) Coal dust concentration/(g/m )

NU
(a) Coal dust with similar σD and different D50 (b) Coal dust with similar D50 and different σD

Fig.10 Determined τ1, τ2 values as a function of nominal dust concentrations

MA
D
P TE
CE
AC

25
 ACCEPTED MANUSCRIPT

Graphical abstract:

160

140 Tiefa coal

1 (D50= 55.94μm & σD = 1.63)
120 2 (D50= 55.94μm & σD = 1.63)
1 (D50= 37.33μm & σD = 1.67)

Time / ms
100

T
2 (D50= 37.33μm & σD = 1.67)

80

IP
60

R
40

20

SC
0
0 100 200 300 400 500
3
Coal dust concentration/(g/m )

NU
MA
D
P TE
CE
AC

26
 ACCEPTED MANUSCRIPT

Highlights

► Coal dust explosibility is controlled by particle size (D50), size dispersity (σD) and dust concentrations.

► Proportions of fine particles are playing an important role in coal dust explosion.

► Coal dust with lower σD, smaller D50 and higher volatile contents cause a serious explosion severity.

T
IP
► Effects of σD and D50 is greater on poor coal dust clouds than on the rich coal dust clouds.

R
SC
NU
MA
D
P TE
CE
AC

27